développement de la compréhension des inférences causales
TRANSCRIPT
Développement de la compréhension
des inférences causales en contexte de récit
chez les enfants âgés de 3 à 6 ans
Thèse
Paméla Filiatrault-Veilleux
Doctorat en médecine expérimentale
Philosophiae doctor (Ph.D.)
Québec, Canada
Paméla Filiatrault-Veilleux, 2016
ii
Développement de la compréhension
des inférences causales en contexte de récit
chez les enfants âgés de 3 à 6 ans
Thèse
Paméla Filiatrault-Veilleux
Sous la direction de :
Chantal Desmarais, Ph.D., directrice de recherche
Caroline Bouchard, Ph.D., codirectrice de recherche
iii
Résumé
INTRODUCTION : La compréhension inférentielle joue un rôle clé en lecture à l’âge
scolaire et émergerait dès l’âge de 3 ou 4 ans. Or, malgré sa pertinence pendant la période
préscolaire, elle est relativement peu documentée chez les enfants de 3 à 6 ans. Une
meilleure description de ses jalons développementaux contribuerait aux connaissances
nécessaires pour mieux la mesurer et la stimuler chez les jeunes enfants.
OBJECTIFS : La thèse vise à décrire le développement de la compréhension de six types
d’inférences causales liées à la grammaire de récit d’une histoire (les réponses internes, le
problème, le but, les tentatives de résolution du problème, les prédictions et la solution) en
contexte de lecture partagée chez les enfants de 3 à 6 ans. L’objectif 1 consiste à dégager
une proposition théorique de la séquence développementale par le biais d’une recension des
écrits. À partir des jalons théoriques proposés, l’objectif 2 est de concevoir un outil
d’évaluation et d’en déterminer les qualités psychométriques. Avec l’outil, l’objectif 3 est
d’en décrire le développement auprès d’enfants âgés de 3 à 6 ans.
MÉTHODES : Suivant la recension des écrits, l’outil de lecture partagée Un abri pour
Pinson est conçu sur iPad. Pour répondre aux objectifs 2 et 3, les participants sont 121
enfants au développement typique âgés de 3 à 6 ans divisés en trois groupes selon l’âge et
le milieu éducatif fréquenté (milieu de garde ou maternelle).
RÉSULTATS : Une séquence développementale de la compréhension des inférences
causales chez les enfants de 3 à 6 ans est proposée. Dès 3 ans, les enfants démontrent des
habiletés inférentielles émergentes, qui s’améliorent significativement avec l’âge. Dès 4
ans, les enfants peuvent inférer les réponses internes, le problème et les prédictions. Vers 5
et 6 ans, les enfants infèrent plus facilement le but, les tentatives de résolution du problème
et la solution. Également, avec l’âge, les enfants se rapprochent des cibles attendues en
termes de qualité de réponses produites.
CONCLUSION : Cette thèse contribue à une meilleure caractérisation d’un pilier
important à la compréhension en lecture et ouvre sur des pistes de réflexion quant à la
stimulation de cette habileté au préscolaire.
iv
Abstract
INTRODUCTION: Inferential abilities play a critical role in reading comprehension at
school age and emerge between 3 or 4 years of age. Despite its relevance during the
preschool period, it is relatively poorly documented in children aged 3-6. A better
description of its developmental milestones would contribute to the knowledge needed to
better measure and stimulate it in young children.
OBJECTIVES: This thesis aims to describe the development of comprehension of six
causal inferences in children between 3-6 years old. These inferences target elements of
story grammar (i.e. internal responses, problem, goal, attempts to solve the problem,
predictions and solution) in the context of a narrative. The first aim is to propose milestones
in the development of this ability through a scoping review. The second, based on the
theoretical sequence proposed, is to create a dialogic reading task application for iPad and
to evaluate its psychometric properties. The third aim is to describe the inferential
comprehension development of 3 to 6-year-old by using the dialogic reading task on iPad.
METHODS: Following the scoping review, the task Un abri pour Pinson (“A shelter for
Pinson”) was created for use on the iPad. To answer objectives 2 and 3, 121 typically
developing children from 3-6 years of age are divided into three groups according to age
range and program attended (daycare centers or kindergarten).
RESULTS: A developmental sequence of the comprehension of six causal inferences
targeting elements of story grammar is proposed for children from 3-6 years old. Inferential
comprehension emerged early, from 3-4 years old, with considerable inter-individual
variability. The ability to infer the problem of the story, the internal response of a character,
and predictions were easier starting at age 4. Then, the 5-6 years old were better able to
infer the goal, the attempt to solve the problem, and the resolution. Finally, between the
ages of 3 and 6, children improved in terms of the quality of response they provided.
CONCLUSION: This thesis contributes to a better characterization of an important pillar
of reading success. This knowledge will lead to better support educational and intervention
strategies with regards to this ability in preschool facilities.
[Tape
z une
citati
on
v
Table des matières
Résumé .................................................................................................................................. iii
Abstract .................................................................................................................................. iv
Table des matières .................................................................................................................. v
Liste des tableaux ................................................................................................................. vii
Liste des figures ................................................................................................................... viii
Dédicaces ............................................................................................................................... ix
Remerciements ....................................................................................................................... x
Avant-propos ........................................................................................................................ xii
Chapitre 1. Introduction générale ........................................................................................... 1
1.1 Problématique ............................................................................................................. 1
1.2 But et objectifs ............................................................................................................ 6
Chapitre 2. Recension des écrits ............................................................................................. 8
Résumé ............................................................................................................................... 8
Abstract ............................................................................................................................... 9
Introduction ...................................................................................................................... 11
Methods ............................................................................................................................ 14
Results .............................................................................................................................. 15
Discussion ......................................................................................................................... 27
Conclusion ........................................................................................................................ 33
Acknowledgements .......................................................................................................... 35
References ........................................................................................................................ 36
Chapitre 3. Conception et qualités psychométriques de l’outil Un abri pour Pinson .......... 39
3.1 Mise en contexte sur la conception de l’outil ............................................................. 39
3.2 Collaborateurs à la conception de l’outil d’évaluation Un abri pour Pinson ............. 40
3.2.1 Graphiste .......................................................................................................... 40
3.2.2 Programmeurs informatiques ........................................................................... 41
3.2.3 Narrateurs ........................................................................................................ 42
3.3 Conception et qualités psychométriques de l’outil Un abri pour Pinson.................. 43
Résumé ..................................................................................................................... 43
Introduction ............................................................................................................... 44
Méthode .................................................................................................................... 48
Résultats .................................................................................................................... 56
Discussion ................................................................................................................. 61
Conclusion ................................................................................................................ 65
Note des auteurs ........................................................................................................ 66
Remerciements ......................................................................................................... 66
Annexe ...................................................................................................................... 67
Références ................................................................................................................. 68
vi
Chapitre 4. Développement de la compréhension des inférences causales chez les enfants
âgés de 3 à 6 ans ................................................................................................................... 73
Résumé ............................................................................................................................. 73
Abstract ............................................................................................................................. 75
Introduction ...................................................................................................................... 76
Method .............................................................................................................................. 80
Results .............................................................................................................................. 84
Discussion ......................................................................................................................... 90
Limitations of the study and future directions .................................................................. 95
Conclusion ........................................................................................................................ 96
Acknowledgment .............................................................................................................. 97
References ...................................................................................................................... 100
Chapitre 5. Autres résultats et apports à la compréhension inférentielle ........................... 105
5.1 Effet de genre............................................................................................................ 105
5.2 Effet de de la structure répétitive de l’histoire sur la compréhension des tentatives de
résolution du problème. .................................................................................................. 108
Références ...................................................................................................................... 114
Chapitre 6. Discussion générale ......................................................................................... 115
6.1 Synthèse des résultats ............................................................................................. 115
6.2 Forces et limites ...................................................................................................... 120
6.3 Retombées .............................................................................................................. 122
6.4 Perspectives futures ................................................................................................ 124
6.5 Conclusion .............................................................................................................. 125
Bibliographie ...................................................................................................................... 126
Annexes .............................................................................................................................. 134
vii
Liste des tableaux
Tableau 1. Classification of causal inferences targeting elements of story grammar.* ....... 13
Tableau 2. Inference types and age of the participants......................................................... 18
Tableau 3. Inference types and methodological characteristics of the studies .................... 21
Tableau 4. Milestones in the development of children’s comprehension of causal inferences
targeting elements of the story grammar ...................................................................... 30
Tableau 5. Promising methods to assess inferential comprehension in young children. ..... 33
Tableau 6. Classification des six inférences ciblant les éléments de la grammaire de récit
ainsi que le nombre, le contexte de l’histoire et les exemples de questions inférentielles
élaborées pour l’outil Un abri pour Pinson .................................................................. 50
Tableau 7. Définition des quatre catégories de réponses sur un continuum de qualité,
exemples de réponses obtenues à la question « Il veut quoi Pinson, tu penses ? » et
points accordés. ............................................................................................................ 52
Tableau 8. But et résumé des éléments évalués par les experts cliniciens lors des deux focus
groupes. ........................................................................................................................ 53
Tableau 9. Caractéristiques des trois groupes de participants : moyenne (et écart-type) pour
l’âge, pourcentage de mères ayant terminé leurs études secondaires et pourcentage de
familles ayant un revenu supérieur au seuil de faible revenu. ...................................... 54
Tableau 10. Résultats de l’analyse factorielle après rotation varimax de l’outil Un abri pour
Pinson. .......................................................................................................................... 58
Tableau 11. Résultats des trois groupes de participants à l’outil Un abri pour Pinson, à la
tâche Petit cochon s’est perdu et à l’ÉVIP. .................................................................. 58
Tableau 12. Statistiques AC1 de Gwet et intervalle de confiance à 95% pour chaque
question lors des deux analyses inter-juges. ................................................................. 60
Tableau 13. Résultats obtenus par les 34 enfants sélectionnés au hasard au test-retest à 2
semaines d’intervalle. ................................................................................................... 61
Tableau 14. Participants’ characteristics. ............................................................................. 81
Tableau 15. Inference types, number of questions asked, story context, examples from the
task and scores .............................................................................................................. 83
Tableau 16. Subscores in percentage for the six inference types of each group. ................. 86
Tableau 17. Les trois questions inférentielles et réponses attendues portant sur les
prédictions des tentatives de résolution du problème. ................................................ 109
Tableau 18. Les trois questions inférentielles et réponses attendues portant sur les
évaluations des tentatives de résolution du problème. ............................................... 111
Tableau 19. Synthèse du développement de la compréhension des inférences causales en
contexte de récit chez les enfants de 3 à 6 ans ............................................................ 119
viii
Liste des figures
Figure 1. Selection of the articles after applying inclusion and exclusion criteria. .............. 14
Figure 2. Processus de conception des illustrations de l’outil Un abri pour Pinson ............ 41
Figure 3. Saisie d’écran des boutons de fonctionnalités d’une page de l’application .......... 42
Figure 4. Scatter plot by groups of the means and standard deviation for the total score of
inferential comprehension. ........................................................................................... 85
Figure 5. Scatter plot by groups of the mean and standard deviation of the inferences
targeting the problem, the internal response and the prediction. .................................. 87
Figure 6. Scatter plot by groups of the mean and standard deviation of the inferences
targeting the goal, the attempt to solve the problem, and resolution. ........................... 88
Figure 7. Cumulative frequency histogram of mean number of responses per category. .... 89
Figure 8. Comparaison des performances obtenues par les filles et les garçons quant au
score inférentiel total à la tâche Un abri pour Pinson. ............................................... 107
Figure 9. Moyennes et écarts-types obtenus par les trois groupes d’enfants pour les
prédictions des trois tentatives de résolution du problème. ........................................ 110
Figure 10. Moyennes et écarts-types obtenus par les trois groupes d’enfants pour les
évaluations des trois tentatives de résolution du problème. ....................................... 112
ix
À Thomas, Nathan, Arnaud, Olivier, Audrey,
Anthony, Charles, Dylanne, Gabriel, Elizabeth,
Jérémie, Philippe, Samuel, William, Léo, Logan,
Xavier, Liam et tous mes autres petits cocos.
À Rachel Filiatrault-Veilleux, notre ange.
[Tape
z une
citati
on
prise
x
Remerciements
Ce projet doctoral n’aurait pu être réalisé sans la présence de nombreuses personnes. De par
leur collaboration au projet ou leur soutien à différents niveaux, je tiens à leur exprimer mes
plus sincères remerciements.
D’abord et avant tout, je tiens à remercier du fond du cœur ma directrice de thèse, Chantal
Desmarais. Chantal, je ne pourrais jamais te remercier assez pour ton écoute, ta présence, ta
sensibilité, tes conseils toujours judicieux, ton support et ton professionnalisme. Tu
représentes pour moi un réel modèle de réussite. Tu es une chercheuse qui non seulement se
distingue par l’excellence de ses travaux, mais qui se soucis également constamment de la
retombée clinique, sur le terrain, des projets que tu mènes. Chantal, merci mille fois de
m’avoir permis de collaborer à ce projet de recherche. Je n’aurais pu être mieux guidée.
Merci pour ces quatre si belles années. Je te souhaite une longue lignée d’étudiants au
doctorat.
Je tiens également sincèrement à remercier Caroline Bouchard, ma codirectrice de thèse.
Bien que nos rencontres aient été réalisées sur une base moins régulière, sache Caroline que
j’en suis toujours ressortie grandie et encore plus inspirée. Ton apport à cette thèse
doctorale est indéniable. Merci beaucoup ! Bien entendu, un merci également à Natacha
Trudeau, co-chercheuse dans le projet, qui a toujours su apporter un excellent regard
critique. Ses commentaires et suggestions ont toujours été très judicieux et ont sans
contredit contribué à l’amélioration significative des manuscrits insérés à la présente thèse.
Ensuite, un grand merci à Jean Leblond, statisticien, qui nous a grandement aidées à la
réalisation des analyses statistiques de cette thèse. Finalement, je tiens à remercier les
membres de mon jury de thèse : Paola Colozzo, Ginette Dionne et Laura Monetta, merci
pour ce très bel échange, vos commentaires constructifs et votre intérêt envers ma thèse !
Ensuite, je remercie les auxiliaires de recherche et les étudiantes à la maîtrise en
orthophonie ayant contribué à la réalisation de ce projet de recherche. Gabrielle Francoeur,
Florence Dion Morin, Catherine Boucher, Carole-Anne Lavoie-Allard, Catherine Julien :
merci mille fois pour votre temps, votre soutien, vos sourires et votre énergie positive !
xi
Je dois également exprimer mes remerciements envers la Faculté de médecine de
l’Université Laval, les bourses doctorales des Fonds Québécois de la Recherche sur la
Société et la Culture (FQRSC) et du Conseil de Recherches en Sciences Humaines (CRSH)
pour leur appui financier tout au long de ces quatre dernières années.
Merci ensuite aux excellents collaborateurs à la conception de l’outil d’évaluation Un abri
pour Pinson ! Nous sommes aujourd’hui très fières de cet outil, qui n’aurait pu prendre vie
de cette façon sans votre précieuse contribution. Merci particulièrement à Simon Genest et
Jean-Nicolas Giguère de la compagnie Logiciels Skyglow Inc., les concepteurs de
l’application informatique. Malgré nos champs d’expertise diamétralement opposés, vous
avez su capter avec beaucoup de génie l’essence même de ce projet, merci ! Merci
également à Édith Julien, qui a su grâce à son formidable talent de graphiste rendre Pinson
vivant et attrayant pour les jeunes enfants ! Merci aux personnes qui ont prêté leur voix aux
personnages de l’application. Merci à Simon Kinkead, le narrateur, Suzie Gobeil,
l’orthophoniste et la famille Pinson (le brillant Antoine et ses parents). Finalement, un
merci sincère aux orthophonistes ayant participé à l’étape de validation de contenu de
l’outil. Ce processus avec vous a été excessivement enrichissant. Je suis certaine qu’Un
abri pour Pinson n’aurait pas pu être ce qu’il est aujourd’hui sans votre apport. Votre
expertise clinique a été énormément appréciée : Roseline Boyer, Mireille Delisle, Ulysse
Hubert, Anne-Marie Lanoue, Catherine Maxès-Fournier et Myriam Tremblay : merci !
Je termine cette section avec des remerciements de nature plus personnelle. Merci tout
d’abord à mes trois folles d’amies, Lisou, Jojo et Véro. Merci d’égayer ma vie de folie
comme vous le faites toujours si bien. Chanelle, malgré la distance, merci de continuer de
cultiver cette belle complicité qui nous a si bien servi dans notre « ancienne vie ». Ensuite,
je ne pourrais jamais assez remercier ma collègue et amie Geneviève. Ge, merci tout
simplement de faire partie de ma vie. Sans toi, je suis certaine que je ne serais pas la
personne que je suis aujourd’hui. Je ne pourrais jamais assez remercier la vie de t’avoir
mise sur mon chemin ! Merci ensuite à mon frère Pascal (et Marie !) et mes parents, qui ont
toujours été là pour me supporter, et ce, peu importe mes projets et choix de vie.
Finalement, merci à mon mari, l’amour de ma vie. Tu es mon ancrage et ma stabilité. Merci
d’avoir toujours cru en moi, des premières aux dernières secondes de cette belle aventure...
xii
Avant-propos
La présente thèse est composée de trois articles principaux :
1. Filiatrault-Veilleux, P., Bouchard, C., Trudeau, N. and Desmarais, C. (2015).
Inferential comprehension of 3-6 year-olds within the context of story grammar: a
scoping review. International Journal of Language & Communication Disorders,
50(6), 737-49. doi: 10.1111/1460-6984.12175
2. Filiatrault-Veilleux, P., Desmarais, C., Bouchard, C., Trudeau, N. et Leblond, J.
(Sous presse). Conception et qualités psychométriques d’un outil d’évaluation de la
compréhension d’inférences en contexte de récit chez des enfants âgés de 3 à 6 ans.
Revue Canadienne d’Orthophonie et d’Audiologie.
3. Filiatrault-Veilleux, P., Desmarais, C., Bouchard, C. and Trudeau, N. (Re-soumis le
30-11-2015 au Journal of Speech Language and Hearing Research).
Comprehension of inferences in a narrative in 3 to 6 year old children.
Le premier article est inséré au chapitre 2 de cette thèse. Il est inséré tel que publié en
2015 dans l’International Journal of Language & Communication Disorders. Le deuxième
article est inséré au chapitre 3 de cette thèse. Il est inséré tel qu’accepté le 9 septembre 2015
dans la Revue Canadienne d’Orthophonie et d’Audiologie. Le manuscrit est actuellement
sous presse et sera publié lors d’une prochaine parution du périodique. Le troisième article
est inséré au chapitre 4 de cette thèse. Il est actuellement en révision par le Journal of
Speech Language and Hearing Research et est inséré tel que re-soumis le 30 novembre
2015 suite aux demandes de corrections des réviseurs. Seuls les numéros des tableaux et
figures des trois articles ont été ajustés afin qu’ils concordent à l’ordre de présentation de la
thèse.
Les trois articles ont été rédigés par Paméla Filiatrault-Veilleux, qui en est la première
auteure. Elle était ainsi responsable de l’analyse des résultats, de la rédaction et de la
soumission aux fins de publication. Les coauteures sont Chantal Desmarais, directrice de la
thèse, Caroline Bouchard, codirectrice de la thèse et Natacha Trudeau, co-chercheuse dans
xiii
le cadre du projet de recherche subventionné par le CRSH1. Cette subvention CRSH de
laquelle proviennent les données analysées de la présente thèse a été octroyée en 2011 à
Chantal Desmarais, responsable du projet de recherche. Dre Chantal Desmarais a ainsi
participé activement à l’élaboration et au bon déroulement de ce projet doctoral. Les trois
co-chercheuses ont participé à la révision, la correction et la bonification des différentes
versions des trois manuscrits. Jean Leblond, statisticien au Centre interdisciplinaire de
recherche en réadaptation et intégration sociale (CIRRIS), a également aidé à l’analyse
statistique des articles 2 et 3. Compte tenu de sa contribution significative à l’analyse
statistique des données de l’article 2, il en est également coauteur.
1 Desmarais, C., Bouchard, C. et Trudeau, N. (2011-2013). Élaboration et validation d'un outil de mesure de la
compréhension des inférences chez les enfants de 3 à 6 ans., Subvention, Conseil de recherches en sciences
humaines du Canada, Programme Savoir : Subventions de développement Savoir, du 2011-06-01 au 2013-05-
31.
1
Chapitre 1. Introduction générale
1.1 Problématique
La réussite scolaire des enfants constitue une priorité dans notre société (ministère de
l'Éducation, de l'Enseignement supérieur et de la Recherche, 2015; ministère de la Famille,
2015). Elle joue effectivement un rôle plus que déterminant dans la vie d’une personne, qui peut
devenir libre de ses choix, de ses rêves et de ses aspirations. Ainsi, maintes études se sont
penchées sur les facteurs pouvant favoriser ce succès, et ce, dès le plus jeune âge des enfants
(Duval et Bouchard, 2013). Les apprentissages réalisés pendant la période de la petite enfance,
avant l’entrée à l’école, constituent sans contredit des éléments clés de la réussite éducative de
l’enfant, et par conséquent, de sa réussite scolaire ultérieure (Duval et Bouchard, 2013;
McKenzie, 2009).
Les récentes données portant sur les prédicteurs à la réussite scolaire future d’un enfant
pointent dans la même direction : le développement des habiletés en compréhension en lecture y
joue un rôle de premier plan (Conseil canadien sur l’apprentissage, 2007; Conseil des statistiques
canadiennes de l’éducation, 2009). C'est pourquoi les réflexions sur l’importance des divers
moyens pour développer cette compétence chez les jeunes enfants sont au cœur des
préoccupations en recherche. Puisque les prérequis à l’apprentissage de la lecture émergent tôt
dans le développement des petits, il importe de s’intéresser précocement à leur développement et
leur mise en place, avant l’entrée à l’école des enfants. Le dépistage et l'intervention précoces
consistent en effet en nos meilleures chances de prévenir et/ou à nettement atténuer les difficultés
en lecture des jeunes enfants et, par la même occasion, d’améliorer leurs chances de réussite
scolaire ultérieure.
Parmi les prédicteurs de la réussite en lecture, la compréhension des inférences, soit
l’habileté à « lire entre les lignes », connaît un intérêt grandissant en milieu éducatif. Avec raison,
elle occuperait le premier rang des compétences à développer pour favoriser la réussite en lecture
à l’âge scolaire (Cain, Oakhill, & Elbro, 2003; Kintsch & Kintsch, 2005; Trabasso & Wiley,
2005; van den Broek et al., 2005; Zwaan & Singer, 2003). Effectivement, au courant des
dernières années, il a été démontré que la compréhension des inférences est une habileté
2
essentielle, voire critique à la compréhension en lecture à l’âge scolaire (Bianco et al., 2010;
Bianco, Pellenq, Lambert, Bressoux, Lima & Doyen, 2012; Botting & Adams, 2005; Ford &
Milosky, 2003; Kendeou, van den Broek, White, & Lynch, 2009; NICHD, Early Child Care
research Network, 2005; Spackman, Fujiki, & Brinton, 2006). Dans les écrits scientifiques, il a
été longtemps considéré que la compréhension inférentielle apparaissait lors de l’entrée des
enfants dans l’univers de la lecture. Cet état de fait explique donc que les données de recherche
actuellement disponibles documentent majoritairement les habiletés inférentielles des enfants
d’âge scolaire jusqu’à l’âge adulte (Ackerman, Silver, & Glickman, 1990; Bishop & Adams,
1992; Letts & Leinonen, 2001; van den Broek, Tzeng, Risden, Trabasso, & Basche, 2001).
Or, les récentes recherches tendent à démontrer que la compréhension des inférences émerge
beaucoup plus précocement dans le développement de l’enfant, bien avant son entrée à l’école.
Par le biais d’observations anecdotiques des habiletés de compréhension des jeunes enfants, il est
effectivement plausible d’émettre l’hypothèse selon laquelle la compréhension inférentielle
émerge dès l’âge de 2 ou 3 ans (Botting & Adams, 2005). Dans une optique de prévention des
difficultés en lecture, de plus en plus d’auteurs recommandent donc d’en soutenir le
développement à l’âge préscolaire en contexte de langage oral (Blanc, 2010; Florit, Roch, &
Levorato, 2011; Kendeou et al., 2009; Makdissi & Boisclair, 2006; van den Broek et al., 2005).
Compte tenu de son effet sur les habiletés en lecture à l’âge scolaire, la présente thèse doctorale
porte sur le développement de la compréhension des inférences des enfants de 3 à 6 ans.
La compréhension inférentielle est définie comme l’habileté à comprendre un message dont
l’interprétation repose en partie sur des éléments implicites (Bianco & Coda, 2002; Bishop, 1997;
Gineste & Le Ny, 2002; van Kleeck, 2008). Un exemple d’inférence pouvant être réalisée par un
enfant est de comprendre qu’il pleut à l’extérieur lorsqu’il entend « Oh! Il va falloir mettre un
imperméable », et ce, sans même qu’il ait regardé par la fenêtre. Pour comprendre un tel énoncé,
l’enfant doit établir des liens entre ce qu’il a entendu, le contexte dans lequel il se trouve et ses
connaissances du monde (Paul, 2000). Les habiletés inférentielles sont démontrées essentielles à
la compréhension de nombreux messages au cours de la période préscolaire, particulièrement en
contexte de récit, où elles contribuent à la création d’une représentation mentale complète et
cohérente d’une histoire (Makdissi & Boisclair, 2006; Potocki, Ecalle, & Magnan, 2012; van
Kleeck, 2008). En effet, afin de tirer le maximum de la lecture d’une histoire, l’enfant doit être
3
capable d’aller plus loin que l’information mentionnée explicitement dans le livre grâce à ses
habiletés inférentielles.
Ainsi, la méthode actuellement privilégiée permettant d’évaluer les habiletés des enfants en
bas âge à comprendre des inférences en contexte de langage oral est la lecture partagée (Blanc,
2010; van Kleeck, 2008). La lecture partagée consiste à poser des questions à l’enfant pendant la
lecture d’une histoire comportant une structure narrative (Makdissi & Boisclair, 2006; Price, van
Kleeck, & Huberty, 2009; van Kleeck, Vander Woude, & Hammett, 2006; Wenner, 2004). Les
réponses de l’enfant aux questions permettent de vérifier s’il est capable d’aller au-delà de
l’information présentée. Le schéma de la structure de récit comporte les éléments suivants : une
situation initiale; un événement déclencheur, soit le problème qui engendre une réponse interne;
un but qui motive les actions du personnage pendant le récit; des tentatives de résolution du
problème et leurs conséquences; et finalement, une solution (Mandler & Johnson, 1977; Stein &
Glenn, 1979).
Le conte « Les Trois petits cochons » (popularisé par Walt Disney, 1993), bien connu des
enfants, est un bon exemple de structure narrative classique. En se basant sur les éléments de
cette structure narrative, différentes questions inférentielles peuvent leur être posées. À titre
d’exemple, dans ce récit, six différents types de questions inférentielles pourraient être dégagées :
la réponse interne, soit l’émotion du personnage (p.ex., « Comment se sent le petit cochon
lorsque le loup veut détruire sa maison ? »); le problème de l’histoire (p.ex., « Quel est le
problème des trois petits cochons ? »), le but de l’histoire (« Qu’est-ce qu’ils veulent? »), une
prédiction (« Que va-t-il se passer lorsque le loup va souffler sur la maison en paille ? »), une
tentative de résolution du problème (« Qu’est-ce que les petits cochons pourraient faire pour
déjouer le loup ? ») et la solution de l’histoire (« Qu’est-ce qui a fonctionné finalement? »). Ces
types d’inférences, qui ciblent spécifiquement les éléments de la grammaire de récit, sont
appelées « inférences causales ». Elles sont reconnues pour être comprises plus précocement par
les jeunes enfants, soit dès l’âge de 3 ou 4 ans (Bowyer-Crane & Snowling, 2005; Das Gupta &
Bryant, 1989; Isbell, Sobol, Lindauer, & Lowrance, 2004; Makdissi & Boisclair, 2006; van
Kleeck, 2008).
4
Cependant, chez les enfants qui éprouvent des difficultés langagières, ceci serait toute autre
chose. Il a été à maintes reprises soulevé que ces enfants présentent des difficultés à comprendre
des inférences (Adams, Clarke, & Haynes, 2009; Bishop, 1997; Botting & Adams, 2005;
Bowyer-Crane & Snowling, 2005; Dodwell & Bavin, 2008; Ryder, Leinonen, & Schulz, 2008;
Skarakis-Doyle & Dempsey, 2008). En effet, lors de tâches impliquant la compréhension
d’inférences, les enfants âgés de 5 à 11 ans ayant un trouble de langage obtiennent des résultats
inférieurs à leurs pairs du même âge (Adams et al., 2009; Bishop, 1997; Botting & Adams,
2005). Comme la compréhension inférentielle est une habileté centrale, voire critique, dans le
développement subséquent de la lecture (Bishop & Adams, 1990; Catts, 2009; Kendeou et al.,
2009; Serpell, Baker, & Sonnenschein, 2005), ces difficultés constituent ainsi un important
facteur de risque à prendre en considération chez ces enfants.
À ce jour, il est donc démontré que la compréhension inférentielle est une habileté critique
en compréhension en lecture, et par conséquent, à la réussite scolaire d’un enfant. Bien que les
écrits scientifiques aient à ce jour davantage documenté cette habileté chez l’enfant d’âge
scolaire, les récentes études rapportent qu’elle émerge plus tôt, avant l’entrée à l’école. Elle peut
être entre autres mesurée par le biais de la lecture partagée. Dans ce contexte, les inférences
causales ciblant les éléments de la structure narrative seraient plus précocement comprises par les
enfants. Finalement, il est également documenté que la compréhension inférentielle est
fréquemment déficitaire chez les enfants présentant des difficultés langagières, pouvant par le fait
même entraîner des conséquences néfastes sur leurs apprentissages tout au long de leur parcours
scolaire.
À la lumière des différentes études réalisées dans le domaine, un besoin de connaissances se
dégage : comment la compréhension des inférences se développe-t-elle chez les enfants d’âge
préscolaire ? Les écrits scientifiques ne fournissent actuellement aucune indication claire quant
aux étapes de développement de cette habileté chez le jeune enfant. En effet, bien que nous
sachions que la compréhension inférentielle puisse émerger dès l’âge de 3 ou 4 ans, sa mise en
place jusqu’à l’apprentissage de la lecture en 1re année du primaire au Québec demeure
méconnue. Pourtant, ces connaissances sont indispensables dans une optique de réussite
éducative, puisqu’elles nous permettraient d’en soutenir plus efficacement le développement dès
le plus jeune âge, et ce, particulièrement chez les enfants en difficultés langagières.
5
Une autre lacune occasionnant également un effet considérable sur la pratique clinique
actuelle des orthophonistes est l’absence d’un outil d’évaluation permettant d’évaluer cette
habileté chez les jeunes enfants. À ce jour, les orthophonistes utilisent donc une variété de tâches
informelles qui s’appuient malheureusement sur très peu de fondements scientifiques. Par
ricochet, l’intervention et la prévention des habiletés de compréhension inférentielle en bas âge
s’avèrent très ardues. En effet, considérant la méconnaissance des étapes développementales et
l’absence d’un outil d’évaluation chez les petits, comment les orthophonistes peuvent-elles agir
de front sur le développement de cette habileté essentielle en lecture fréquemment déficitaire
chez les enfants en suivi orthophonique ? La question demeure entière.
Ainsi, une meilleure description des habiletés des jeunes enfants à faire des inférences et des
attentes développementales qui en découlent s’avèrent essentielles. Ces connaissances
permettraient aux orthophonistes de s’appuyer sur des assises théoriques solides, permettant de
soutenir et de promouvoir plus efficacement le développement de la compréhension des
inférences chez les jeunes enfants. Pour y arriver, la conception d’un outil d’évaluation s’avère
également essentielle afin de collecter ces informations. Dans le but de pallier à manque de
connaissances, la présente thèse de doctorat vise à proposer une séquence développementale de la
compréhension des six types d’inférences causales des enfants âgés de 3 à 6 ans, tranche d’âge
présumée de l’émergence de cette habileté en contexte de langage oral. Les six types d’inférences
causales basées sur la structure de récit d’une histoire sont ciblés, soit la réponse interne, le but,
les tentatives de résolution du problème, les prédictions et la solution.
La première étape de la démarche doctorale vise à dégager des écrits scientifiques les
éléments soutenant une proposition théorique de la façon dont se met en place la compréhension
inférentielle à l’âge préscolaire (article 1). La deuxième étape est de concevoir un outil
d’évaluation à partir des jalons développementaux mis de l’avant à l’étape 1 (article 2). Enfin, la
troisième étape consiste à valider de façon empirique et préciser le cas échéant cette proposition
théorique quant au développement des inférences auprès d’enfants d’âge préscolaire (article 3).
Les résultats d’une telle démarche serviront à mettre de l’avant une proposition scientifiquement
appuyée d’une séquence de développement de la compréhension des inférences causales chez le
jeune enfant, qui pourra servir d’assises à une démarche pédagogique visant la stimulation
efficace du développement de cette habileté chez le jeune enfant.
6
1.2 But et objectifs
Le but de cette thèse est de décrire le développement de la compréhension de six types
d’inférences causales liées à la grammaire de récit d’une histoire (c.-à-d. les réponses internes,
l’événement déclencheur, le but, les tentatives de résolution du problème, les prédictions et la
solution) en contexte de lecture partagée chez les enfants âgés de 3 à 6 ans. Pour ce faire, trois
objectifs généraux sont poursuivis et ils se déclinent en trois articles distincts abordés dans les
chapitres 2 à 5 de cette thèse.
Le premier objectif général (chapitre 2) consiste à dégager une proposition théorique de la
séquence développementale de la compréhension des inférences causales en contexte de récit
chez les enfants âgés de 3 à 6 ans par le biais d’une recension des écrits (de type « scoping
review »). Plus spécifiquement et à partir des écrits scientifiques cette étude vise à :
1.1 Décrire le développement de la compréhension de six types d’inférences causales
ciblant les éléments de la grammaire de récit afin de proposer des jalons
développementaux chez les enfants âgés de 3 à 6 ans.
1.2 Mettre en lumière les éléments méthodologiques clés utilisés pour recueillir cette
information auprès des jeunes enfants.
À partir des jalons théoriques proposés, le deuxième objectif général (chapitre 3) consiste à
concevoir un outil d’évaluation de la compréhension des six inférences causales impliquées en
contexte de récit, à en déterminer la validité de contenu, de même qu’à en mesurer les qualités
psychométriques. Plus spécifiquement, cette étude vise à :
2.1 Concevoir un outil d’évaluation des habiletés de compréhension de six types
inférences causales en contexte de lecture partagée, sous la forme d’un livre
d’histoire interactif sur iPad.
2.2 Évaluer la validité de contenu, via une consultation d’experts cliniques, des
éléments suivants : (a) le matériel de l’outil que constitue la structure narrative du
récit, les questions posées et la qualité des illustrations qui accompagnent le récit et
(b) la fonctionnalité de l’outil via le médium utilisé, soit un iPad.
7
2.3 Déterminer les qualités psychométriques de l’outil créé que sont : la validité de
construit et la validité concurrente de l’outil de même que sa fidélité inter-juge et
test-retest.
Le troisième objectif général (chapitres 4 et 5) de la thèse est de décrire le développement
de la compréhension des six types inférences causales chez des enfants âgés de 3 à 6 ans, à l’aide
de l’outil de lecture partagée créé sur iPad afin de confirmer et/ou modifier la séquence
développementale mise précédemment de l’avant. Plus spécifiquement, cette étude vise à :
3.1 Comparer les performances des enfants des différentes tranches d’âge quant au
score total de compréhension inférentielle et quant aux types d’inférences causales
comprises ciblant les éléments de la grammaire de récit.
3.2 Comparer les performances des enfants des différentes tranches d’âge en fonction
de la qualité des réponses produites.
3.3 Vérifier la présence d’un effet de genre.
3.4 Vérifier l’effet de la structure répétée de l’histoire sur la compréhension des
tentatives de résolution du problème.
8
Chapitre 2. Recension des écrits
Le présent chapitre est le résultat d’une recension des écrits portant sur le développement
de la compréhension de six types d’inférences causales ciblant les éléments de la grammaire de
récit chez les enfants âgés de 3 à 6 ans. Cet article est intitulé « Inferential comprehension of 3-6
year-olds within the context of story grammar: a scoping review” et a été publié en 2015 dans
l’International Journal of Language and Communication Disorders.
Filiatrault-Veilleux, P., Bouchard, C., Trudeau, N. and Desmarais, C. (2015). Inferential
comprehension of 3-6 year-olds within the context of story grammar: a scoping review.
International Journal of Language & Communication Disorders, 50(6), 737-49.
Résumé
Contexte : La compréhension des inférences joue un rôle crucial en lecture à l’âge scolaire et
émergerait précocement dans le développement de l’enfant. Il serait donc pertinent pour les
cliniciens et chercheurs dans le domaine d'avoir accès à une séquence développementale au
préscolaire appuyée scientifiquement.
Objectifs : Cette étude vise à combler les lacunes scientifiques actuelles quant à la compréhension
inférentielle chez les jeunes enfants en procédant à une recension des écrits. À partir des écrits
scientifiques, les objectifs spécifiques sont de : (1) décrire le développement de la compréhension
des inférences causales ciblant des éléments de la grammaire de récit dans le but de proposer des
jalons développementaux et (2) mettre en évidence les éléments méthodologiques clés utilisés
pour recueillir cette information chez les jeunes enfants.
Méthode : Seize études provenant de six bases de données répondant aux critères d'inclusion ont
été analysées qualitativement dans le cadre d’une recension des écrits. Contributions principales :
Des éléments clés dans le développement des inférences causales impliquées en contexte de récit
chez les jeunes enfants ainsi que des procédures prometteuses pour évaluer cette habileté sont
soulevés.
Conclusion : Davantage d'études sont nécessaires afin de mieux cerner le développement de cette
habileté au préscolaire.
9
Abstract
Background: The ability to make inferences plays a crucial role in reading comprehension and
the educational success of school-aged children. However, it starts to unfold much earlier than
school entry and literacy. Given that it is likely to be targeted in speech language therapy, it
would be useful to clinicians to have access to information about a developmental sequence of
inferential comprehension. Yet, at this time, there is no clear proposition of the way in which this
ability develops in young children prior to school entry.
Aims: To reduce the knowledge gap with regards to inferential comprehension in young children
by conducting a scoping review of the literature. The two objectives of this research are: (1) to
describe typically developing children’s comprehension of causal inferences targeting elements
of story grammar, with the goal of proposing milestones in the development of this ability and (2)
to highlight key elements of the methodology used to gather this information in a pediatric
population.
Methods: A total of 16 studies from six databases that met the inclusion criteria were
qualitatively analyzed in the context of a scoping review. This methodological approach was used
to identify common themes and gaps in the knowledge base to achieve the intended objectives.
Main contribution: Results permit the description of key elements in the development of six
types of causal inference targeting elements of story grammar in children between 3 to 6 years
old. Results also demonstrate the various methods used to assess this ability in young children
and highlight particularly interesting procedures for use with this younger population.
Conclusions: These findings point to the need for additional studies to better understand this
ability and to develop strategies to stimulate an evidence-based developmental sequence in
children from an early age.
Keywords: inferential comprehension, inference types, story grammar, young children, scoping
review.
10
What this paper adds
What is already known in relation to this subject?
The ability to make inferences is crucial to reading comprehension throughout the school years.
Yet, this ability emerges in younger children and there are indications in the literature that, in a
narrative context, 3-6 year olds with typical language development are already able to answer
questions that require them to make different inference types.
What this study adds:
This scoping review summarizes research findings and identifies gaps in the existing literature by
highlighting two key elements: the primary results obtained in studies of typically developing
children aged 3-6 based on six causal inference types targeting the elements of story grammar
and the promising methodological avenues which could be used to evaluate this ability in young
children.
11
Introduction
Language comprehension supports learning and social integration in young children. It is a
complex process that involves more than understanding each word, morpheme, and sentence
(Paul 2000). Indeed, to comprehend a message, children must take into account both their own
knowledge of the world as well as that of the speaker’s, the speaker’s motivations and intentions,
and certain expectations as to how people generally communicate. In addition, they must develop
the ability to read between the lines (Bishop 1997) and to fill in their representation of a message
with certain elements that are not explicitly spoken by the other person (Gineste and Le Ny 2002,
van Kleeck 2008). We refer to these abilities as inferential language comprehension. An example
of an inference that can be made by a child during dialogic reading is when he predicts what the
main character will do next in the story. If the main character of the story has lost his favorite toy,
a child may infer that on the next page the main character would search for the toy or ask his
parents to help him. To make this inference, the child must understand words and sentences and
further connect them with his own knowledge and experience to interpret the motivation and
intention of the main character.
According to experts in the field (Cain and Oakhill 2007), inferential comprehension
deserves to be further explored in young children because of its crucial importance for reading
comprehension once they reach school age (Botting and Adams 2005, Ford and Milosky 2003,
Makdissi and Boisclair 2006, Spackman et al. 2006). In a similar fashion to what happens in oral
language, when learning to read one must create a coherent mental representation of the message
to grasp the meaning (van Kleeck 2008). Indeed, some studies show significant positive links
between inferential comprehension in young children and their future performance in reading
comprehension at school-age (Ferreiro and Taberoski 1982, Kendeou et al. 2009, Kontos and
Wells 1986).
Recent research suggests that methods used to measure inferential comprehension must be
different for young children than those used with older children who are learning to read or adults
who have already acquired this skill. The currently preferred method to assess inferential
comprehension in early childhood is by way of dialogic reading, whereby an adult asks the child
questions while reading a story constructed with a classical narrative structure (Ford and Milosky
12
2008, Makdissi and Boisclair 2006, Price et al. 2009, van Kleeck et al. 2006, Wenner 2004,
Whiterhurst and Lonigan 1998). This method has proved to be effective to promote inferential
comprehension development (Wenner 2004), in addition to contributing to the child’s overall
language development (Botting and Adams 2005, Cain and Oakhill 1999, van Kleeck 2008).
Indeed, it allows young children to establish more explicit links between events when they listen
to the story, which leads to the elaboration of a complete representation of its meaning (Makdissi
and Boisclair 2006).
Dialogic reading with children optimally follows the narrative structure of a story that is
comprised of the following story grammar elements: (1) a setting, (2) a triggering event, (3) a
problem that generates an internal response, (4) a goal that motivates the character's actions
during the narrative, (5) attempts to solve the problems as well as their consequences, and finally,
(6) a resolution to the story (Mandler and Johnson 1977, Stein and Glenn 1979). Understanding
inferences can be measured in young children through different types of questions related to the
story (Desmarais et al. 2013, Ford and Milosky 2003, Ford and Milosky 2008, Makdissi and
Boisclair 2006, Trabasso and Nickels 1992, van Kleeck et al. 2006, van Kleeck 2008, Wenner
2004). In this study, we focus on causal inference types that target the elements of story grammar
thus helping to establish links between events in the story (Bowyer-Crane and Snowling 2005,
Makdissi and Boisclair 2006, van Kleeck 2008). As the story grammar elements tie together as a
causal chain, the comprehension of causal inferences between those links could play a critical
role in the comprehension of stories (Isbell et al. 2004, Makdissi and Boisclair 2006, van Kleeck
2008). Since these inferences are crucial to comprehension (Isbell et al. 2004, Makdissi and
Boisclair 2006) and appear early in child development (Das Gupta and Bryant 1989), it seems
appropriate, from a developmental and clinical point of view, to evaluate them at an early age.
Six causal inference types targeting the elements of story grammar are shown in table 1 (Bowyer-
Crane and Snowling 2005, Graesser et al. 1994, Makdissi and Boisclair 2006, van Kleeck 2008).
13
Table 1. Classification of causal inferences targeting elements of story grammar.*
Inference types Examples of question
Triggering event / the protagonist’s problem What is the protagonist’s problem?
Internal Response How does the protagonist feel?
The character’s purpose/goal What does the protagonist want or what is he
looking for?
Attempts to resolve the problem What could he do? Do you think it will work?
Predicting the next step What do you think will happen next?
Consequence or solution Did the protagonist achieve what he wanted?
Sources: *Bowyer-Crane and Snowling (2005), Graesser et al. (1994), Makdissi and Boisclair (2006), van Kleeck
(2008).
The studies investigating causal inferences in children aged 4-6 years have shown
inferences related to story grammar appear early and can be measured in a dialogic reading
context (Das Gupta and Bryant 1989, Desmarais et al. 2013, Kendeou et al. 2009, van Kleeck
2008). Furthermore, it is anecdotally reported that the ability to make inferences emerges as early
as age 2 (Botting and Adam 2005). From an early intervention perspective and given the crucial
role of inferential abilities in reading comprehension and later academic success, a clearer picture
of normal development in this area is necessary. More knowledge regarding the construction of
inferential language skills could serve as the foundation of educational and clinical strategies
aimed at stimulating the development of this ability in young children.
The aim of this study was to review the state of knowledge and hopefully close the
knowledge gap with regards to inferential comprehension in young children. This was
accomplished by conducting a scoping review of the literature on this topic with a focus on
studies that use dialogic reading with children aged 3-6. Given that the development of this
ability in young children has received little attention and that this topic has been analysed from
multiple points of view and domains, i.e. speech and language pathology, psychology, education,
a scoping review method was chosen (Rumrill et al. 2010). This type of review documents the
research to date, with the goal of summarizing research findings and identifying gaps in the
existing literature (Arksey and O'Malley 2005, Drager et al. 2010). The two objectives of this
research was: (1) to describe typically developing children’s comprehension of causal inferences
14
targeting elements of story grammar, in order to propose milestones in the development of this
ability and (2) to highlight key elements of the methodology used to gather this information
among this young population.
Figure 1. Selection of the articles after applying inclusion and exclusion criteria.
Methods
A literature search was conducted in peer-reviewed journals published from 1985 to 2012. Six
databases were targeted: Medline/PubMed, Medline/Ebsco, CINAHL, PsycINFO, ERIC,
Education Full Text. The keywords used were: inferential comprehension (inferences, inference
types, inferential comprehension, inferential level of comprehension), young children (children,
preschool child: 2-6 years, educational level: early childhood education, kindergarten, preschool
education), language development (assessment, language, oral language, language development)
and dialogic reading / narrative context (dialogic reading, story book reading, narrative context).
The total number of studies identified was 461. As can be seen in figure 1, once the duplications
were removed, there was a total of 372 articles.
Databases
n = 372
Remained by title
n = 145
Remained by abstract
n = 48
Remained by article
n = 37
Final
n = 16
ExcludedExcluded
n = 21
Excluded
n =15
Added by references
n = 4
Excluded
n = 97
Excluded
n = 227
15
To select articles for further analysis, the following inclusion criteria were used: (1)
targeting oral language development; (2) involving inferences in story grammar context; (3)
examining young, typically developing children aged 3-6, and (4) being written in English or
French. Thus, studies of children with language or developmental disorders and/or relating to
written language were excluded from the study. At each step of the literature search, the first two
authors surveyed the articles independently and, in the case of a disagreement, discussed the
selection to reach a consensus. The process of selection of the articles is shown in figure 1.
Following the application of inclusion and exclusion criteria, a total of 16 articles were analyzed
qualitatively to reach the intended objectives. This methodological approach is typically used to
examine a broad research topic and to identify themes and patterns in an area of research
(Rumrill et al. 2010). Given the context of this area of research, this scoping review will provide
an important foundation for future studies in this domain, by identifying gaps in the knowledge
base and by highlighting the most common approaches that researchers use to evaluate inferential
comprehension in young children.
Results
The results are presented in two sections as per the two objectives of the study. The first provides
information linked to the description of typically developing children’s comprehension of six
causal inferences within the context of story grammar in relation to the age of children, i.e. the
internal response, the problem of the protagonist, the purpose/goal of the character, the attempts
to solve the problem, the prediction of the next step and the consequence or resolution. The
second section presents results regarding methods used to gather this information with young
children.
Inference types in children aged 3-6 years within the context of story grammar
Overall, out of a total of 16 articles, three studies investigated inferential comprehension
in children aged 3, thirteen studies in 4 year olds, eight in 5 year olds and, finally, fourteen
studies included 6 year olds2. Fifteen studies used a cross-sectional design and one, i.e. Kendeou
et al. (2008), used a longitudinal study design. The column “Mean age” in table 2 presents the
2 Note that in two studies, because of the study design, results from 6 year olds could only be reported in conjunction
with results of 7 and 8 year olds.
16
mean ages of children constituting the sample of each study in detail. With regard to inference
types, our review attempted to provide an overall coherent picture of inferential comprehension
in this population while taking into account the fact that a variety of labels and definitions of
inference types were used across studies. Looking at the definitions and examples of inferences
from different authors, we made the decision to group certain inference types in the same
category3. This was done to simplify the analysis and to present the results in a way that would
paint a clear picture of this ability in young children. In the 16 articles included in this review, the
two inference types that were examined in most studies are the internal response of the character
(9/16) and the goal/purpose of the protagonist (6/16). The inference type least studied is the
problem of the character (i.e. in 2/16 studies). However, it must be noted that the definitions and
examples given by the authors for this type of inference are similar to those provided for the goal
or purpose of the character. This observation helps to illustrate why inferential comprehension
remains difficult to explain and measure. The final classification of inference types is presented
in table 2 in relation with the key results obtained by children in relation to their age. Table 3
presents the definitions and examples given by each author to assess the various inference types
in relation again with the children’s age. The two tables illustrate that the ability to respond to
various causal inferential questions in a story grammar context gradually evolves in children from
age 3 to 6 years. It is clear that inferential comprehension is present and evolving during the
preschool period, long before children enter school and learn how to read. In fact, the ability to
understand the causal connections between events is present during early childhood and has been
evidenced by the 16 studies included in this review (table 2). Results indicate that inferences
constrained by the story or based on the story structure are understood better and earlier than
other types of inferences (Gibbons et al. 1986, Hudson and Slackman, 1990) and play a critical
role in the comprehension of stories beyond other language skills and vocabulary (Kendeou et al.
2008). The analysis presented in table 2 also reveals some consistencies in the association
between age and the type of inferences that children understand. The results suggest a tentative
developmental sequence that we elaborate in the discussion, while also confirming the need for
more investigation and theory-building on this topic.
3 For example, according to Kendeou and al. (2008), the definition of "Action inference" refers to how an agent's
goal is achieved. According to Makdissi and Boisclair (2006), a "Solution" type of inference is an inferential
question related to the solution of the story. Therefore, we included inferences called "Consequence or solution,”
“Causal consequence,” “Resolution of the problem” and “Action inference” in the same category.
17
Methods employed to assess inferential comprehension within the context of story grammar in
young children
First, the qualitative analysis showed that studies used a variety of experimental tasks to
assess inferential comprehension in young children (refer to “Task” column in table 2). Materials
included short stories created specifically for the experiment (10/16), existing storybooks (5/16),
or the subtest of a published assessment tool (2/16). These were presented as short texts read
aloud or via audio output (4/16), short videos or films (3/16), storybooks (5/16), pictures (2/16),
or objects to be manipulated (1/16). Second, table 3 illustrates that the number of questions asked
to the children varied between studies, ranging from four (Makdissi and Boisclair 2006) to 96
(Crais and Chapman 1987). Furthermore, rather than asking questions, three studies analyzed
what children said during or after the presentation of a story (verbatim transcription and coding).
Third, the timing of the questions varied across studies. All studies used off-line questions, i.e.
after the presentation of the story, while only two studies also focused on the ability to make
inferences on-line by asking questions during the presentation of the story (Makdissi and
Boisclair 2006, Lynch and van den Broek 2007). Finally, different scoring procedures of the
answers provided by the children were reported and some authors used more than one procedure.
The types of scoring used were pre-set categories, e.g. correct, partial or incorrect (5/16);
multiple choice, i.e. select the correct picture from a choice of two, three or four (4/16), closed
questions or a dichotomised variable, i.e. “yes or no,” or “true or false” judgment statements
(4/16), scored responses as correct/incorrect (3/16) and, finally, coded transcription of the
children’s spontaneous productions (4/16). Overall, the impressive variety of tasks and methods
related to the story itself touched on the modality of the presentation, the number of questions to
test inferential comprehension, the moment at which they are presented to the children, and the
scoring of children’s responses.
18
Table 2. Inference types and age of the participants
References Number of
participants
Age
(years)
Inferences types
examined Task Key results
Deconti and
Dickerson
(1994)
36
(12 per age
group)
3- 5
Internal response Situation stories and drawings with
implicit/explicit goal,
positive/negative emotion.
3 year olds (y.o.): make inferences about
internal responses, but 3 < 4 and 5 y.o.
Condition implicit = explicit.
Makdissi and
Boisclair
(2006)
12
(four per
age group)
3- 5
Problem, purpose/goal,
theme, solution
Story read aloud with on-line and off-
line questions.
3 y.o.: Awareness of the internal response,
would appear first. 3-5 y.o.: problem >
purpose > solution > theme of the story. On-
line > off-line questions.
Wenner
(2004)
64
4 Character goal Stories with causal connections
manipulated: goal success occurring
early or late. Long/short version.
Pictures/props supporting story.
4 y.o.: make inferences about the purpose of
story and the causal connections. Are
sensitive to the structure of story: goal
success occurring late > early.
Gibbons et al.
(1986)
96 4:5 Narrative convention;
constrained by story
(non-specific).
Stories of 30 seconds. Modality:
audio/audiovisual form & output:
verbal/dolls-toys manipulation.
Recall/reconstruction.
4 y.o.: Inferences constrained by the story>
other types of inferences. Method effect:
audiovisual> audio.
19
Hudson and
Slackman
(1990)
42 4:7 Based on story structure
(non-specific)
Reading aloud stories about familiar
events. Story recall & 2 inferential
off-line questions.
4 and 6 y.o.: inferences based on story
structure> not based on story structure.
Ford and
Milosky
(2008)
16a 4: 9 Internal response Stories presented audio-visually about
an emotional state.
4 y.o.: make inferences on internal response.
Making inference predict social competence.
Kendeou et
al. (2008)
113 4 and 6 Goals/purposes, attempts
resolution of problem,
causal antecedent,
causal consequence,
state of characters,
internal response
Narratives presented orally or audio-
visually. Story recall and questions.
4 and 6 y.o.: ability to make inferences
predicted understanding of narrative beyond
other language skills and vocabulary. 4-6
and 8 y.o.: increase of the ability to infer the
purpose of the character.
Florit et al.
(2011)
221 4-6 Local or global
coherenceb, based on
world knowledge or
text.
2 tasks: Reading aloud two texts &
experimental tool: short paragraphs.
4-6 y.o.: ability to understand explicit and
implicit information develops from 4 to 6
years of age. Condition explicit> implicit
and this difference is stable with the increase
in age.
20
Gnepp and
Gould
(1985)
48 5:5
Prediction and
explanation of the
internal response
Short stories with prediction and
explanation of the internal response
in reaction to event.
5 y.o.: Internal response situational>
personalized. Increase of personalized type.
Negative emotions > positives.
Spackman et
al. (2006)
43a 5 and 6 Internal response Presentation of short scenarios (similar
to Ford & Milosky 2003
methodology).
5 and 6 y.o.: can make inferences about the
internal responses of characters. Happiness
> sadness, fear and anger.
Blanc (2010) 28
(14 per age
group)
5-6
Internal response
(different types),
character (who?),
time (when?), space
(where?), causality
(why?).
Reading aloud stories.
5 and 6 y.o.: Space and Causality dimensions >
Internal response in story comprehension
5-7 y.o.: Story comprehension precision with
age. Relation between Causality & Internal
response. Are sensitive to the types of
emotional information transmitted. Gender
effect: girl> boy on Internal response &
Causality.
Ford and
Milosky
(2003)
24a 5: 4- 6:
2
Internal responses Stories created to enable a particular
emotion presented in 3 modalities
(visual, verbal, combined)
5 and 6 y.o.: can make inferences on the
internal response. Modality effect:
combined> visual or verbal. Emotion
“happy” > “mad” or “surprised”.
21
Notes: aTypically developing children included in the control group of a study on child development disorders. bLexical, referential, causal, thematic, concerned super-ordinate goals, internal response and internal states.
Blanc (2009) 30 6 4 situational dimensions:
space (where?), time
(when?) , causality
(why?), character
(who?)
Reading aloud stories.
6 y.o.: prefer dimensions Character, Space and
Causality. Compared to older children,
weaker to infer Causality (which improves
with age). Information considered important
in the story changes with age.
Adams et al.
(2009)
64a 6 Causal inference,
internal response,
character motivation
Picture with short text. 6 y.o.: can make causal inferences, inferences
on internal responses and motivations of the
characters.
Crais and
Chapman
(1987)
16 6 Major setting;
antecedent;
consequence
Reading aloud stories (7-11 sentences
per story).
6 y.o.: Can make inferences, but their
performance is lower than older children
aged 9 and 10 years.
Lynch and
van den
Broek
(2007)
25 6 Goals characters Playing two versions of a story (Goal
and non-goal directed) and think-
aloud protocol.
6 y.o.: can make inferences about the goals of
the characters. Goal > non-goal directed
narratives.
22
Table 3. Inference types and methodological characteristics of the studies
Material and procedures Examples from the task
Internal response (Number of studies = 9) / 3-6 year olds
Deconti and
Dickerson
(1994)
Stories of characters in various emotion-provoking situations. Twelve
off-line multiple choice questions (MCQ).
Billy has a toy tractor. He sits on it. A wheel falls off and he
cannot ride it anymore (emotion judgements).
Ford and
Milosky
(2008)
Scenarios about emotion. Following stories, facial expression, matched
or not, appears on screen. Eighteen off-line MCQ.
It was Twinkie’s birthday. He opened a present. It was a big
teddy bear. (Facial expression matched or not, e.g.
happy).
Florit et al.
(2011)
Stories of 3 or 4 sentences, referring to common and familiar events,
followed by two inferential questions (Q).Ten off-line MCQ.
Inferential questions were on character’s emotions and
internal states.
Gnepp and
Gould
(1985)
Stories (drawings) with prediction and explanation of the internal
response in reaction to an event. Twelve off-line Q and answers (A)
classified in categories.
A child picked up his gerbil, and the gerbil bit him, and it
hurt. “How did he feel? Can you tell me why he felt…?”
Spackman et
al. (2006)
Story supported by pictures designed to activate a particular emotion
(anger, fear, happiness, sadness). Sixteen off-line Q and A classified
in categories.
Chris falls asleep. Bob puts a lollipop in Chris’s hair. Chris
tries to pull out the lollipop, but it just gets stuck in more
hair. Chris is… The children indicate what emotion is
experienced.
Blanc (2010) Reading aloud storybooks. Forty-two statements off-line and true or A King finds his wife and his son in a well. Statement :
23
false (T/F). When the king found his wife and son, he was happy.
(T/F)
Ford and
Milosky
(2003)
Stories of two actions and an emotional response (4 emotions). Thirty-
six off-line MCQ.
Twinky, a teddy bear, was bouncing a ball. Bully (his
friend) took the ball. Twinky was… Show or tell how
Twinky feels.
Adams et al.
(2009)
Short text read aloud with a picture. Eleven off-line Q and A classified
in categories.
Picture of a kitchen in the aftermath of a burglary.
Inferential question: “How does the family feel now?”
Kendeou et al.
(2008)
Stories read aloud with pictures. Story recall & Q. Seven, eight or nine
off-line Q and A correct-incorrect & verbatim coding.
Q on character emotion inferences: refer to the feelings
experienced by a character in response to an event.
Goal/purpose of character / Motives (Number of studies = 6) / 4-6 year olds
Wenner (2004) Stories supported by pictures or by using small props. “Goal-late” and
“Goal-early” versions. Four off-line Q and A classified in categories.
The Friendly Pirate: He wants to find a treasure (goal). He
finds it after a series of actions (late) or at the beginning
of the story (early). “What did the pirate want to do in
this story?”
Makdissi and
Boisclair
(2006)
Reading aloud a storybook, Q asked during or after the story. Four on-
line and four off-line Q and verbatim coding.
Franklin in the Dark: a turtle is afraid of the dark inside its
own shell. “Why did Franklin go to see the animals?”
Florit et al.
(2011)
Short stories on common and familiar events & 2 inferential Q. Ten off-
line MCQ.
Inferential Q concerned superordinate goals.
24
Kendeou et al.
(2008)
Stories read aloud with pictures. Story recall & Q. Seven, eight or nine
off-line Q and A correct-incorrect & verbatim coding.
Q on goal inferences: refer to motives of an agent.
Lynch and van
den Broek
(2007)
Stories with pictures. Two versions (goal-directed and non-goal-
directed) of two narratives. Nine off-line Q and correct-incorrect A &
Think-aloud protocol
The little boy (he lost his frog and searches for it): explicit
statement of superordinate goal or intentional action in
connection with the goal.
Adams et al.
(2009)
Short text read aloud with a picture. Eleven off-line Q and A classified
in categories.
Picture: a kitchen in the aftermath of a burglary. Q on
motives for the crime: “Why would someone steal
something?”
Problem of the character / Major setting (Number of studies = 2) / 3-6 year olds.
Makdissi and
Boisclair
(2006)
Reading aloud a storybook, Q asked during or after the story. Four on-
line and four off-line Q and verbatim coding.
Franklin in the Dark: a turtle is afraid of the dark inside its
own shell. “What was Franklin’s big problem?”
Crais and
Chapman
(1987)
Stories incorporating themes taken from international fables. Ninety-six
off-line Q and T/F & story retell.
The king’s son has to find which of two ducks the male is.
He asked a farmer help to find answer. Judgment
“yes/no” on major setting: “Did the boy ask for the
farmer’s help?”
Attempts to solve or resolution of the problem / plot episode / Action inference (Number of studies = 3) / 5-6 year olds.
Blanc (2009),
(2010)
Reading aloud storybooks. Forty-two or 64 off-line Q and T/F. Main character facing difficulties. “Coraline had to go into
the woods in search of food for her brothers” (yes/no).
25
Kendeou et al.
(2008)
Stories read aloud with pictures. Story recall & Q. Seven, eight or nine
off-line Q and A correct-incorrect & verbatim coding.
Action inferences: refer to how an agent’s goal is achieved
(through actions).
Consequence or solution / Causal consequence / Resolution of the problem / Action inference (Number of studies = 3) / 5-6 year olds
Makdissi and
Boisclair
(2006)
Reading aloud a storybook, Q asked during or after the story.
Four on-line and four off-line Q and verbatim coding.
Franklin in the Dark: a turtle is afraid of the dark inside its
own shell. Off-line: “Did Franklin find a solution?”
Crais and
Chapman
(1987)
Stories incorporating themes taken from international fables.
Ninety-six off-line Q and T/F & story retell.
The king’s son has to find which of two ducks the male is.
He asked a farmer help to find answer. Judgment
“yes/no” on major setting: “Did the boy ask for the
farmer’s help?”
Kendeou et al.
(2008)
Stories read aloud with pictures. Story recall & Q. Seven, eight or nine
off-line Q and A correct-incorrect & verbatim coding.
Action inferences: refer to how an agent’s goal is achieved.
Predicting the next step or antecedent / Causality / Causal antecedent (Number of studies = 3) / 6 year olds
Adams et al.
(2009)
Short text read aloud with a picture. Eleven off-line Q and A classified
in categories.
Picture: a kitchen in the aftermath of a burglary. Inferential
Q: “Why was the dog barking?”
Crais and
Chapman
(1987)
Stories incorporating themes taken from international fables.
Ninety-six off-line Q and T/F & story retell.
The king’s son has to find which of two ducks the male is.
He asked a farmer help to find answer Judgment
“yes/no” on causal antecedent: Did the female duck get
in the water first?
26
Kendeou et al.
(2008)
Stories read aloud with pictures. Story recall & Q. Seven, eight or nine
off-line Q and A correct-incorrect & verbatim coding.
Causal antecedent: connection between current event &
previous context. Causal consequence: an event
unfolding from the current event.
27
Discussion
The aim of this study was to review the state of knowledge and attempt to close the knowledge
gap with regards to inferential comprehension in young children by conducting a scoping review
of the literature. This review summarizes research results by structuring two key elements. First,
it describes the performance of typically developing 3-6 year olds in studies targeting causal
inferences involved in story grammar (table 4). Second, it highlights methods used by these
studies to assess the ability to make inferences in young children (table 5). The review provides
insights into the developmental sequence children follow in making inferences related to story
grammar. As early as age 3, the inferences likely to occur are internal response, purpose/goal of
the character and the triggering event. In addition, some inferences such as the attempt to solve
the problem, the prediction of the next step and the consequence or solution were studied only in
5-6 year olds. Taking all of these observations into consideration, it is also important to note that
the interpretation of our results underscores the difficulty in drawing theoretical conclusions
related to this complex ability in young children and shows that further studies are needed to
better understand it.
Inference types targeting elements of story grammar in children aged 3 to 6 years
The inference type that is reported for the youngest children is the internal response of
characters. More precisely, Deconti and Dickerson (1994) reported that 3 year old children were
already able to make inferences about a character’s emotional reaction, but that 4 and 5 year-olds
performed even better. Further, the 3 year olds’ performance was greater than chance level,
suggesting that at this age children have some comprehension of the relation between emotions
and situations. Moreover, according to Ford and Milosky (2008), not all 4-year-old children were
able to generate such inferences. They observed that even if all children in their study were able
to correctly identify emotions by associating them with the appropriate facial expressions, they
were not all able to use their knowledge to independently generate the inference at that age. Other
authors distinguished a) personalized types, i.e. inference based on personal experiences, from b)
situational internal responses, i.e. inference based on the perspective or relevant prior experience
of the character in a story (Gnepp and Gould 1985). According to these authors, there was a
gradual age related increase in children’s ability to interpret an event from another person's
28
perspective in light of his prior experiences. Then, with regards to the types of emotions (either
positive or negative), some authors reported that positive emotions were more easily inferred by
children than those of a negative nature (Ford and Milosky 2003, Spackman et al. 2006). Our
analysis of the available data indicate that children as young as 3 or 4 years of age can make an
internal response of characters inference. Some interesting features emerge from the data linked
to types of emotions transmitted (e.g. either positive or negative) and how they are processed by
children.
Looking at our results chronologically, the next type of inference examined was the
purpose/goal of the character, studied in 4-6 year olds. In Wenner (2004), an interesting feature
of 4 year-old children’s performance was their sensitivity to the structure of the story when
inferring the goal of a character. Indeed, children scored better when the goal’s achievement
occurred late in the story rather than early. In the “late condition”, there would be more causal
connections constituting the narrative, and then actions became consistent with achieving the
goal. Consequently, the “late condition” supported children in making inferences more
accurately. In the same vein, Lynch and van den Broek (2007) evaluated the ability of 6-year-old
children by manipulating the hierarchical structure of the narrative. Children's performance was
better when protagonists in the story had a principal goal linked to several smaller goals, than
when it was non goal-directed. This suggests that 6 year olds are sensitive to the important role of
a character’s goal in a story. At that age, they can engage in complex cognitive processes while
listening to stories, forming coherent mental representations by manipulating hierarchical links
within the narrative. Kendeou et al. (2008) reported that there is a clear developmental pattern
related to the ability to infer the goal of the character between age 4 and 8. Thus, the ability to
infer the purpose/goal of the story occurs early in development and is supported by the causal
relations constituting the narrative
The inference targeting the problem of the protagonist has also been studied in 4-6-year-
old children by two groups of researchers (Makdissi and Boisclair 2006, Crais and Chapman
1987). According to Makdissi and Boisclair (2006), there is a link between children's ability to
correctly infer the goal of the character and the problem to be solved. It is indeed quite logical
that, in order to identify the protagonist’s goal, one needs to be aware that there is a problem to
solve. It is also interesting to note that Crais and Chapman (1987) report that 6 year old children
29
obtained results that were inferior to those of 9 and 10 year olds. Due to this type of inference
only being analysed in two studies, it may be too early to draw conclusions about its
developmental nature. Yet, given the important link between the problem to solve and the goal of
the protagonist, it is possible that some studies may have considered these two types of inferences
as a single one.
The fourth type of inference, studied in 5 and 6 year olds, was the consequence or
resolution of the problem. At age 5, children are able to make this inference, but it appears to be
more difficult to make than an inference about the problem or the goal of the story. That being
said, young children may show some early signs of comprehension of that inference type. They
have demonstrated the ability to make connections between the final emotion of the story
(internal response) and the final outcome, even if they were not able to infer the solution
(Makdissi and Boisclair 2006). Once again, more studies are needed to make a conclusion about
this inference type in young children.
Turning to the ability to infer about attempts to solve the problem in a story, it has been
documented in studies involving 5 and 6 year olds. These inferences are effectively more difficult
for young children but this does not seem to prevent them from grasping the general meaning of
the story. This is in agreement with some authors, not all inferences are critical to the
comprehension of a story (Blanc 2009, 2010, Kendeou et al. 2008). The actions of the characters
to achieve a goal contribute significantly to the comprehension of the story only by 6-8 year olds
but it does not help 4 year olds understand (Kendeou et al. 2008). According to these authors, a
developmental pattern is consistent with developmental hierarchy inferences, showing that
children are gradually able to identify causal relationships involving the goals, emotions, and
desires of the characters, based on the achievement of the goal by way of the narrative. An
interesting fact, mentioned by Morrow (1985), is that the setting, the theme, and the resolution of
the problem are recalled more frequently than the plot episode, or attempts to solve the problem.
Morrow suggests that this may be because the setting, the theme and the resolution of the
problem represent the more meaningful elements of the story. The plot episodes, recalled least
often, present many details that are repeatedly extraneous and are not needed to deduce the most
important ideas from the story. In sum, the ability to infer the attempts to solve the problem has
been measured in older children, 5-6 year olds, and emerges later in the development because of
30
the many links children need to make between goals, emotions, and desires of the characters.
Furthermore, the elements constituting the plot episodes may carry less important details of the
story for children.
Finally, the prediction of the next step of an event has been examined in 6-year-old
children. Blanc (2009) report that they were able to make inferences about causal relationships
while listening to a story. In interpreting this finding, Blanc suggests that there may be a
relationship between the ability to infer the causality of an event and the internal response of a
character (Blanc 2009, 2010). Other studies also argue that children at age 6 are able to make
causal inferences (Adams et al. 2009, Crais and Chapman 1987). Furthermore, Kendeou et al.
(2008) report an improvement in the ability to make inferences about causal relationships
(antecedents and consequences) between the ages of 6 and 8. Thus, the ability to infer the next
step and causality in a story is present in older children, around the age of 6 and appears to
improve in subsequent years.
Table 4. Milestones in the development of children’s comprehension of causal inferences
targeting elements of the story grammar
At age 3 Internal
Response
From an early age, children have some comprehension
of the relation between emotions and situations. May
be sensitive to the types of emotions transmitted (e.g.
either positive or negative).
At age 4 Purpose/goal of
the character
May be sensitive to the structure of the story and causal
relations constituting the narrative.
At age 4 Triggering event
/ problem of
the
protagonist
May be a link between children's ability to correctly
infer the character’s goal and the problem to solve.
More studies are needed.
At age 5-6 Consequence or
solution
Appears to be more difficult than the inference related
to the problem or goal of the story. More studies are
needed.
At age 5-6 Attempts to May be more difficult to infer than other types because
31
resolve the
problem
details constituting the plot episodes may carry less
important information for children. More studies are
needed.
Studied at
age 6
Predicting the
next step
May be a link between the ability to infer the causality
or the prediction of an event and the internal response
of a character. More studies are needed.
Methods used: Promising ways to assess inferential comprehension in young children
As indicated in the results, the variety of tasks used is impressive. No two studies
employed the same methods and instruments to assess this ability. In fact, 25 years later, we are
faced with the conclusion that it is difficult to assess developmental differences of inferential
comprehension since the measurement procedures vary greatly in the literature (Hudson and
Slackman 1990). There are some important aspects identified in studies that seem to support the
evaluation of the inferential comprehension of young children in the context of oral language. Of
note is the fact that the methods used to assess inferential comprehension of young children
should be different from those used for older children or adults, taking into account the aspects of
development, such as working memory, executive functions, level of language development, etc.
If these elements are not addressed, they can negatively influence the results. The methods used
by studies to assess the ability to make inferences in young children deemed promising by the
authors of this paper are shown in table 5.
First, given that the modality of presentation of the story can play a role in the level of
comprehension of children, it appears important to select a modality that supports and promotes
better inferential comprehension to optimize the outcomes for young children. Audio-visual
presentation or dialogic reading appears promising in this regard. This modality appears to
promote focus and attention of the young child on the task (Ford and Milosky 2003) more than
the sole presentation of pictures or text read aloud. Indeed, there is no doubt in the literature of
the importance of visual support, in addition to the oral presentation of the information to
children.
32
Second, in most studies, questions were mostly asked off-line, i.e. after the presentation of
the story which is not characteristic of dialogic reading. However, according to Makdissi and
Boiclair (2006), young children can develop more explicit causal relationships during the
storytelling –on-line, rather than after the story –off-line. In fact, dialogues during the
presentation of a story can increase and complexify the expression of causal relationships. This
finding is also demonstrated by several studies that highlight the importance of interactive
reading in language development in children (Justice and Kaderavek 2004, Price et al. 2009). The
context of interactive reading (on-line) invites the child to dialogue and to play an active role in
the construction of the meaning of the story. In addition, reading helps children to understand the
components of narrative structure and thereby to make the inferences that are tied to the story
(Justice and Kaderavek 2004). In contrast, when questions are asked after the presentation of the
story, it puts more weight on memory skills rather than on comprehension skills (Blanc 2010, van
den Broek et al. 2001). Therefore, a promising research avenue is to further explore the ability of
young children to make inferences on-line, during the presentation of a story.
Finally, the procedures used to analyze children’s responses vary considerably from one
study to another which also complicates comparisons between studies. There are two methods to
measure the inferential ability of young children which appear to be promising. The first is the
classification of the types of responses given by children (Adams et al. 2009, Barnes et al. 1996,
Desmarais et al. 2013, Gnepp and Gould 1985, Spackman et al. 2006). This method leads to a
more nuanced discrimination of the children’s performance, an effective strategy that can be used
in clinical practice (Adams et al. 2009). Such categories can also serve as a means of scaffolding
this complex ability on several levels, instead of a “correct/incorrect” scoring. A child can
actually infer a part of the problem and generate a hypothesis to answer that is not entirely true or
entirely false and therefore falls between the two. The second promising method, used by Lynch
and van den Broek (2007), is the "think-aloud protocol" while performing the task. This
procedure, originally used with older children (Trabasso et al. 1995), was adapted for 6 year olds.
The spontaneous answers provided represent children's thinking while listening to a story, which
would specifically examine the mental processes in which they engage during the task. Thus, 6-
year-old children could perform this type of task and demonstrate acceptable skills to
33
spontaneously produce inferences. In fact the authors went so far as to suggest extending the
approach to even younger children (Lynch and van den Broek 2007).
Table 5. Promising methods to assess inferential comprehension in young children.
Modality of
presentation
Performance may be linked to the number of modalities used to stimulate
comprehension (combining auditory and visual support).
Timing of questions Asking questions during the presentation of a story may increase the
complexity of the expression of causal relationships. In contrast,
questions asked after the presentation of the story may place the
emphasis on memory skills rather than on comprehension skills.
Analysis of
children’s responses
Categorisation of responses evaluates the ability of the child on several
levels, resulting in a more representative description of this complex
skill that a dichotomous response cannot elicit. Think aloud protocol:
Answers provided spontaneously represent children's thinking while
listening to a story, which specifically observes the mental processes
during the execution of the task.
Conclusion
Well before entering school, young children begin to make inferences in the context of the stories
read to them. As early as three years of age, they are good at inferring the internal response, the
problem and the goal of a character in a story. Later on, they develop the ability to infer attempts
to solve the problem, predict the next step in a story and understand a consequence or solution to
an event. The most robust results found in this review are presented in a chronological order that
may eventually serve as a reference to elaborate a developmental trajectory of inferential
comprehension and to confirm milestones in development. With regards to research methodology
in this area, using reading aloud of a picture book that provides auditory and visual input, asking
on-line questions during the story, and scoring children’s responses along a scale of adequacy are
promising avenues for application both in future studies as well as in clinical observations of this
ability. In addition, breaking down inferential comprehension in smaller parts, as we have done in
34
this review, may lead to designing educational material that promote inferential comprehension in
young children by using more explicit and thus more efficient strategies.
35
Acknowledgements
This research was supported by a grant to Chantal Desmarais from the Social Sciences and
Humanities Research Council of Canada. The authors are also grateful to Catherine Boucher and
Gabrielle Francoeur, research assistants in the project.
Declaration of interest: The authors report no conflicts of interest. The authors alone are
responsible for the content and writing of the paper
36
References
ADAMS, C., CLARKE, E. and HAYNES, R. 2009. Inference and sentence comprehension in
children with specific or pragmatic language impairments. International Journal of
Language & Communication Disorders, 44, 301-318.
ARKSEY, H. and O'MALLEY, L. 2005. Scoping studies: towards a methodological framework.
International Journal of Social Research Methodology, 8, 19-32.
BARNES, M. A., DENNIS, M. and HAEFELE-KALVAITIS, J. 1996. The effects of knowledge
availability and knowledge accessibility on coherence and elaborative inferencing in
children from six to fifteen years of age. Journal of Experimental Child Psychology, 61,
216-241.
BISHOP, D. V. M. 1997. Uncommon understanding : development and disorders of language
comprehension in children, Hove, UK, Psychology Press.
BLANC, N. 2009. La compréhension de contes présentés oralement en classes de CP et CE1 :
quelle utilisation des dimensions situationnelles ? L'Année Psychologique, 109, 607-628.
BLANC, N. 2010. La compréhension des contes entre 5 et 7 ans: Quelle représentation des
informations émotionnelles? [The comprehension of the tales between 5 and 7 year-olds:
Which representation of emotional information?]. Canadian Journal of Experimental
Psychology/Revue Canadienne de Psychologie Expérimentale, 64, 256-265.
BOTTING, N. and ADAMS, C. 2005. Semantic and inferencing abilities in children with
communication disorders. International Journal of Language & Communication
Disorders, 40, 49-66.
BOWYER-CRANE, C. and SNOWLING, M. J. 2005. Assessing children's inference generation:
what do tests of reading comprehension measure? The British Journal of Educational
Psychology, 75, 189-201.
CAIN, K. and OAKHILL, J. 2007. Children's comprehension problems in oral and written
language : a cognitive perspective, New York, Guilford Press.
CAIN, K. and OAKHILL, J. V. 1999. Inference making ability and its relation to comprehension
failure in young children. Reading and Writing, 11, 489-503.
CAIN, K., OAKHILL, J. V. and ELBRO, C. 2003. The ability to learn new word meanings from
context by school-age children with and without language comprehension difficulties.
Journal of Child Language, 30, 681-694.
CRAIS, E. R. and CHAPMAN, R. S. 1987. Story Recall and Inferencing Skills in
Language/Learning-Disabled and Nondisabled Children. Journal of Speech and Hearing
Disorders, 52, 50-55.
DAS GUPTA, P. and BRYANT, P. E. 1989. Young children's causal inferences. Child
Development, 60, 1138-46.
DECONTI, K. A. and DICKERSON, D. J. 1994. Preschool children's understanding of the
situational determinants of others' emotions. Cognition & Emotion, 8, 453-472.
DESMARAIS, C., NADEAU, L., TRUDEAU, N., FILIATRAULT-VEILLEUX, P. and
MAXES-FOURNIER, C. 2013. Intervention for improving comprehension in 4-6 year
old children with specific language impairment: practicing inferencing is a good thing.
Clinical Linguistics & Phonetics, 27, 540-52.
DRAGER, K. D., REICHLE, J. and PINKOSKI, C. 2010. Synthesized speech output and
children: a scoping review. American Journal of Speech-Language Pathology , 19, 259-
73.
37
FERREIRO, E. and TABEROSKI, A. 1982. Literacy before schooling, Portsmouth, NH,
Heinemann Educational Books Inc.
FLORIT, E., ROCH, M. and LEVORATO, M. C. 2011. Listening text comprehension of explicit
and implicit information in preschoolers: The role of verbal and inferential skills.
Discourse Processes, 48, 119-138.
FORD, J. A. and MILOSKY, L. M. 2003. Inferring emotional reactions in social situations :
Differences in children with language impairment. Journal of Speech, Language and
Hearing Research, 46, 21-30.
FORD, J. A. and MILOSKY, L. M. 2008. Inference generation during discourse and its relation
to social competence: an online investigation of abilities of children with and without
language impairment. Journal of Speech, Language & Hearing Research, 51, 367-380.
GIBBONS, J. ANDERSON, D.R., SMITH, R., FIELD, D.E. and FISCHER, C. 1986. Young
children's recall and reconstruction of audio and audiovisual narratives. Child
Development, 57, 1014-1023.
GINESTE, M.-D. and LE NY, J. F. 2002. Psychologie cognitive du langage : de la
reconnaissance à la compréhension, Paris, Dunod.
GNEPP, J. and GOULD, M. E. 1985. The Development of Personalized Inferences:
Understanding Other People's Emotional Reactions in Light of Their Prior Experiences.
Child Development, 56, 1455-1464.
GRAESSER, A. C., SINGER, M. and TRABASSO, T. 1994. Constructing inferences during
narrative text comprehension. Psychological Review, 101, 371-95.
HUDSON, J. A. and SLACKMAN, E. A. 1990. Children's use of scripts in inferential text
processing. Discourse Processes, 13, 375-385.
ISBELL, R., SOBOL, J., LINDAUER, L. and LOWRANCE, A. 2004. The Effects of
Storytelling and Story Reading on the Oral Language Complexity and Story
Comprehension of Young Children. Early Childhood Education Journal, 32, 157-163.
JUSTICE, L. M. and KADERAVEK, J. N. 2004. Embedded-explicit emergent literacy
intervention I: Background and description of approach. Language, speech, and hearing
services in schools, 35, 201-11.
KENDEOU, P., BOHN-GETTLER, C., WHITE, M. J. and VAN DEN BROEK, P. 2008.
Children's inference generation across different media. Journal of Research in Reading,
31, 259-272.
KENDEOU, P., VAN DEN BROEK, P., WHITE, M. J. and LYNCH, J. S. 2009. Predicting
reading comprehension in early elementary school : the independent constributions of oral
language and decoding skills. Journal of Educational Psychology, 101, 765-778.
KONTOS, S. and WELLS, W. 1986. What preschool children know about reading and how they
learn it. Young Children, 42, 58–66.
LYNCH, J. S. and VAN DEN BROEK, P. 2007. Understanding the glue of narrative structure:
Children's on- and off-line inferences about characters' goals. Cognitive Development, 22,
323-340.
MAKDISSI, H. and BOISCLAIR, A. 2006. Interactive reading: A context for expanding the
expression of causal relations in preschoolers. Written Language and Literacy, 9, 177-
211.
MANDLER, J. M. and JOHNSON, N. S. 1977. Remembrance of Things Parsed: Story Structure
and Recall. Cognitive Psychology, 9, 111-151.
38
MORROW, L. M. 1985. Retelling Stories: A Strategy for Improving Young Children's
Comprehension, Concept of Story Structure, and Oral Language Complexity. The
Elementary School Journal, 85, 647-661.
PAUL, R. 2000. "Putting things in context": literal and discourse approaches to comprehension
assessment. Seminars in Speech and Language , 21, 247-54; quiz 255.
PRICE, L. H., VAN KLEECK, A. and HUBERTY, C. J. 2009. Talk during Book Sharing
between Parents and Preschool Children: A Comparison between Storybook and
Expository Book Conditions. Reading Research Quarterly, 44, 171-194.
RUMRILL, P. D., FITZGERALD, S. M. and MERCHANT, W. R. 2010. Using scoping literature
reviews as a means of understanding and interpreting existing literature. Work, 35, 399-
404.
SPACKMAN, M. P., FUJIKI, M. and BRINTON, B. 2006. Understanding emotions in context:
the effects of language impairment on children's ability to infer emotional reactions.
International Journal of Language & Communication Disorders, 41, 173-88.
STEIN, N. L. and GLENN, C. G. 1979. An analysis of story comprehension in elementary school
children., Norwood, NJ: Ablex, R. Freedle (Ed.).
TRABASSO, T. and NICKELS, M. 1992. The development of goal plans of action in the
narration of a picture story. Discourse Processes, 15, 249-275.
TRABASSO, T., SUH, S., PAYTON, P. and JAIN, R. 1995. Explanatory inferences and other
strategies during comprehension and their effect on recall. In R.F. Lorch (Ed.), Sources of
coherence in reading (pp. 219-239). Hillsdale, NJ : Lawrence Erlbaum Associates.
VAN DEN BROEK, P., TZENG, Y., RISDEN, K., TRABASSO, T. and BASCHE, P. 2001.
Inferential questioning: Effects on comprehension of narrative texts as a function of grade
and timing. Journal of Educational Psychology, 93, 521-529.
VAN KLEECK, A. 2008. Providing Preschool Foundations for Later Reading Comprehension:
The Importance of and Ideas for Targeting Inferencing in Storybook-Sharing
Interventions. Psychology in the Schools, 45, 627-643.
VAN KLEECK, A., VANDER WOUDE, J. and HAMMETT, L. 2006. Fostering literal and
inferential language skills in Head Start preschoolers with language impairment using
scripted book-sharing discussions. American Journal of Speech-Language Pathology, 15,
85-95.
WENNER, J. A. 2004. Preschoolers' comprehension of goal structure in narratives. Memory, 12,
193-202.
WHITEHURST, G. J. and LONIGAN, C. J. 1998. Child Development and Emergent Literacy.
Child Development, 69 (3), 848-72.
39
Chapitre 3. Conception et qualités psychométriques de l’outil
Un abri pour Pinson
Le présent chapitre porte en premier lieu sur la mise en contexte concernant la conception
de l’outil Un abri pour Pinson. Il présente ensuite les collaborateurs à sa conception en décrivant
leur contribution au projet. Enfin, la suite du chapitre est constituée d’un article accepté pour
publication dans la Revue Canadienne d’Orthophonie et d’Audiologie. Il est intitulé « Conception
et qualités psychométriques d’un outil d’évaluation de la compréhension d’inférences en contexte
de récit chez des enfants âgés de 3 à 6 ans ».
3.1 Mise en contexte sur la conception de l’outil
À partir des résultats mis en évidence par la recension des écrits (Filiatrault-Veilleux,
Bouchard, Trudeau, & Desmarais, 2015), un outil d’évaluation standardisé de la compréhension
des inférences causales a été conçu sur iPad. Chantal Desmarais et Paméla Filiatrault-Veilleux en
sont les principales auteures.
L’outil d’évaluation Un abri pour Pinson a été développé en s’inspirant des critères de van
Kleeck et collaborateurs (2006) qui avaient conçu une tâche de lecture partagée afin d’évaluer la
compréhension inférentielle des participants de leur étude, âgés de 3 à 5 ans. Un précédent projet
dirigé par Dre Chantal Desmarais (Desmarais, Archambault, Filiatrault-Veilleux, & Tarte, 2012)
a également servi d’assise à l’élaboration de ce nouvel outil d’évaluation. Dans leur étude,
Desmarais et collaborateurs avaient développé un outil expérimental, sensible au développement
langagier, élaboré à partir d’un livre d’histoire pour enfant intitulé Petit cochon s’est perdu
(Amery & Cartwright, 2001). Cet outil d’évaluation, en format papier, comportait toutefois
certaines faiblesses méthodologiques, telles que des variations dans la passation et un manque de
standardisation, qui nuisaient à la validité de l'outil. Ces lacunes méthodologiques qui découlaient
entre autres de la version papier ont donc poussé les auteures à vouloir en améliorer la validité, en
se tournant vers les nouvelles technologies. En effet, elles prennent un essor considérable dans le
domaine de l’évaluation en pédagogie et ouvrent vers de nouvelles avenues très intéressantes. En
plus de susciter l’attention et l’intérêt des enfants (Brandone, Michnick Golinkoff, & Hirsh-
Pasek, 2008; Verhallen, Bus, & de Jong, 2006), elles permettent également de combiner dans une
40
même application un ensemble de fonctionnalités permettant d’uniformiser la procédure de
passation des tests. Ainsi, dans le cadre du projet de recherche subventionné par le CRSH4,
Chantal Desmarais, responsable du projet, et Paméla Filiatrault-Veilleux, étudiante au doctorat,
ont développé un outil d’évaluation des habiletés inférentielles, en utilisant le même concept de la
lecture d’une histoire, mais sur iPad cette fois, afin de maximiser la standardisation de la
procédure d’administration. Les auteures ont ainsi conçu une nouvelle histoire pour enfants, en y
intégrant des questions inférentielles à poser aux enfants ainsi que des croquis des illustrations
accompagnant le récit. Une graphiste et des programmeurs informatiques ont ensuite été
embauchés afin de collaborer à la conception de l’application informatique sur iPad.
3.2 Collaborateurs à la conception de l’outil d’évaluation Un abri pour Pinson
3.2.1 Graphiste
Les illustrations qui représentent l’histoire ont été créées en collaboration avec la
graphiste Édith Julien. À partir de croquis de dessins qui accompagnaient l’histoire, la graphiste a
conçu des illustrations sur support informatique (Figure 2). Cette procédure a été employée afin
que les illustrations comportent tous les détails essentiels à l’interprétation de l’histoire et à
l’évaluation précise de la compréhension des inférences sélectionnées. Quelques rencontres et
échanges ont eu lieu entre la graphiste et les auteures de l’outil afin que les dessins représentent
bien les personnages et le contenu souhaité. Elles ne devaient pas contenir de détails non
pertinents à l’interprétation visuelle. En effet, il importait que les illustrations soient claires et
attrayantes pour les jeunes enfants tout en dégageant une interprétation non équivoque du
contenu.
4 Desmarais, C., Bouchard, C. et Trudeau, N. (2011-2013). Élaboration et validation d'un outil de mesure de la
compréhension des inférences chez les enfants de 3 à 6 ans., Subvention, Conseil de recherches en sciences
humaines du Canada, Programme Savoir : Subventions de développement Savoir, du 2011-06-01 au 2013-05-31.
41
Croquis des illustrations Illustrations sur support informatique
Figure 2. Processus de conception des illustrations de l’outil Un abri pour Pinson
3.2.2 Programmeurs informatiques
La conception de l’application iPad a été réalisée en collaboration avec des programmeurs
informatiques de la compagnie Logiciels Skyglow Inc., une équipe composée de Simon Genest,
Jean-Nicolas Giguère et David Jacques. À partir du projet d’histoire soumis par les auteures, les
programmeurs informatiques ont élaboré un ensemble de fonctionnalités informatiques servant de
support à l’histoire et à la codification des réponses des enfants aux questions. L’histoire pour
enfants prend ainsi la forme d’un livre interactif, sur iPad. À chacune des pages de l’histoire, des
boutons permettent d’activer les fonctionnalités suivantes : la narration du récit, la question posée
à l’enfant, l’enregistrement de la réponse de l’enfant et la codification de la réponse (Figure 3). À
chacune des pages, une fois la réponse codifiée, le bouton « terminer » s’active afin de
déverrouiller le changement de page de l’histoire. L’enfant peut donc participer à l’administration
42
de la tâche en tournant lui-même les pages au moment approprié. Cette application iPad permet
donc d’évaluer les habiletés de compréhension des inférences des jeunes enfants, tout en assurant
une procédure de passation systématique et rigoureuse grâce aux différentes fonctionnalités
informatiques.
Figure 3. Saisie d’écran des boutons de fonctionnalités d’une page de l’application
3.2.3 Narrateurs
Finalement, des fichiers audio des différentes voix des personnages de l’histoire ont été
intégrés à l’application informatique. La voix du narrateur de l’histoire est celle de Simon
Kinkead. Le narrateur a conservé une voix neutre tout au long du récit, afin de ne pas indicer la
compréhension des inférences ciblées. L’orthophoniste Suzie Gobeil a prêté sa voix aux
questions inférentielles posées par l’application informatique. Il était important que les deux voix,
celle de la narration et celle des questions posées, soient bien différentes. Cette distinction de
genre entre les deux voix permet d’indiquer à l’enfant quand il doit être attentif à la narration et
quand il doit répondre à une question. Les différents personnages de l’histoire ont été interprétés
par un petit garçon de 7 ans prénommé Antoine (Pinson) et ses parents (mère et père de Pinson),
de même que par Chantal Desmarais (Mme la moufette) et Nicolas Murray (M. le castor, M. le
porc-épic et Tom). Finalement, certains effets sonores (par ex. : bruit de la pluie et chants
Activation de la narration
Activation de la question
inférentielle à poser à l’enfant
Enregistrement de la réponse
de l’enfant
Codification de la réponse de
l’enfant
Le bouton « terminer » s’active
après la codification de la
réponse. Il est ensuite possible
de tourner la page de l’histoire.
Numéro de page de l’histoire
43
d’oiseaux) ont également été intégrés aux fichiers audio afin d’ajouter une touche de réalisme à
l’histoire.
3.3 Conception et qualités psychométriques de l’outil Un abri pour Pinson
La suite du présent chapitre est constituée d’un article accepté le 9 septembre 2015 dans la
Revue Canadienne d’Orthophonie et d’Audiologie. Le manuscrit est actuellement sous presse et
sera publié lors d’une prochaine parution du périodique. Il est intitulé « Conception et qualités
psychométriques d’un outil d’évaluation de la compréhension d’inférences en contexte de récit
chez des enfants âgés de 3 à 6 ans ».
Filiatrault-Veilleux, P., Desmarais, C., Bouchard, C., Trudeau, N. et Leblond, J. (Sous presse).
Conception et qualités psychométriques d’un outil d’évaluation de la compréhension d’inférences
en contexte de récit chez des enfants âgés de 3 à 6 ans. Revue Canadienne d’Orthophonie et
d’Audiologie.
Résumé
L’habileté à comprendre des inférences joue un rôle crucial dans la réussite éducative du jeune
enfant. Or, il s’agit d’un construit complexe et il n’existe pas d’outil standardisé pour le mesurer.
De manière à contribuer à ce domaine de connaissances, un outil d’évaluation intitulé Un abri
pour Pinson a été élaboré sur iPad afin d’évaluer la compréhension d’inférences ciblant les
éléments de la grammaire de récit chez des enfants âgés de 3 à 6 ans. Suite à cette conception, la
présente étude vise à examiner la validité de contenu, évaluée auprès d’orthophonistes experts,
afin d’optimiser l’outil avant son utilisation auprès d’enfants. Puis, la recherche en détermine les
qualités psychométriques à partir d’un échantillon de 121 enfants, âgés de 3 à 6 ans. La validité
de construit, la validité concurrente, la fidélité inter-juge et test-retest sont ainsi mesurées. Les
résultats démontrent qu’Un abri pour Pinson s’avère valide sur le plan du contenu et démontre de
bonnes qualités psychométriques. Ce nouvel outil d’évaluation contribuera à préciser le
développement de la compréhension inférentielle chez les jeunes enfants en contexte de récit et
pourra être utilisé en clinique comme en recherche.
Mots clés : Inférences; grammaire de récit; outil d’évaluation; validité; fidélité
44
Introduction
La compréhension inférentielle constitue un élément important de la compréhension en lecture à
l'âge scolaire (Bianco et al., 2010; Cain & Oakhill, 2007; Joffe, Cain, & Maric, 2007; van den
Broek et al., 2005; van Kleeck, 2008). En effet, pour comprendre un texte, un enfant doit non
seulement décoder et comprendre les informations explicites qu’il lit, comme les mots et les
phrases, mais il doit également interpréter les informations implicites, ce qui l’oblige à réaliser
des inférences afin de bien saisir le sens du texte (Bishop, 1997). Cette habileté permet ainsi à
l’enfant de se créer une représentation mentale cohérente et complète d’un message (Ford &
Milosky, 2003; Joffe et al., 2007). Il n'est donc pas surprenant que plusieurs études aient
démontré des liens positifs significatifs entre la compréhension inférentielle pendant la petite
enfance et le succès ultérieur en compréhension en lecture (Ferreiro & Taberoski, 1982; Kontos
& Wells, 1986), lui-même relié à la réussite scolaire ultérieure de l’enfant. En effet, des récentes
études longitudinales ont soulevé la contribution relativement indépendante du développement
des habiletés inférentielles et de la compréhension orale précoce sur la réussite ultérieure en
compréhension de lecture (Bianco et al., 2010; Kendeou, van den Broek, White, & Lynch, 2009;
NICHD, Early Child Care research Network, 2005).
Des écrits scientifiques ont démontré que les habiletés de compréhension inférentielle
s’inscrivent tôt dans le développement de l’enfant (Bianco et al., 2010; Das Gupta & Bryant,
1989; Desmarais, Nadeau, Trudeau, Filiatrault-Veilleux, & Maxes-Fournier, 2013b; van Kleeck,
2008). Effectivement, les informations dont on dispose jusqu‘à maintenant, quoique parcellaires,
indiquent que les habiletés inférentielles sont présentes dès le jeune âge de l’enfant. Ainsi, dès
l’âge de 3 ou 4 ans, les enfants seraient capables de répondre à des questions inférentielles en lien
avec une histoire qui leur est lue (Deconti & Dickerson, 1994; Makdissi & Boisclair, 2006, van
den Broek, 2005). Par ailleurs, dès l’âge de 4 à 6 ans, il est possible de distinguer, à des tâches de
compréhension inférentielle, la performance d’enfants au développement typique lorsqu’elle est
comparée à celle d’enfants qui présentent un trouble de langage (Adams, Clarke, & Haynes,
2009; Desmarais et al., 2013b; Filiatrault-Veilleux, Tarte, & Desmarais, 2015; Ford & Milosky,
2003, 2008; Spackman, Fujiki, & Brinton, 2006; Trabasso & Nickels, 1992). Ces derniers
présenteraient des difficultés à comprendre des inférences et obtiendraient des résultats
comparables à ceux d’enfants plus jeunes (Bishop, 1997; Botting & Adams, 2005; Bowyer-Crane
45
& Snowling, 2005; Ryder, Leinonen, & Schulz, 2008; Skarakis-Doyle & Dempsey, 2008). Étant
donné que cette habileté apparaît tôt, et aussi compte tenu de son implication en compréhension
de lecture et dans la réussite scolaire notamment, il est pertinent pour les orthophonistes de
s’intéresser précocement à son développement.
Dans une perspective de dépistage précoce des difficultés langagières chez les petits, il
importe de pouvoir évaluer l’habileté à comprendre des inférences en bas âge. Or, l’état actuel
des connaissances ne permet pas de le faire. D’une part, peu d'études ont été menées pour évaluer
cette habileté pendant la période de la petite enfance (Ford & Milosky, 2003, 2008; Trabasso &
Nickels, 1992; van Kleeck, Vander Woude, & Hammett, 2006; Wenner, 2004). D’autre part, les
méthodes utilisées pour mesurer cette habileté sont variées et, à l’heure actuelle, aucun outil n’est
adapté aux enfants qui ne lisent pas encore.
Partageant la préoccupation actuelle concernant la rareté des outils d’évaluation en
orthophonie en contexte franco-québécois (Bouchard, Fitzpatrick, & Olds, 2009), et parce qu’il
n’existe pas encore d’outil clinique évaluant spécifiquement la compréhension des inférences
chez les jeunes enfants, nous avons entrepris le développement de cet outil chez les enfants non-
lecteurs de 3 à 6 ans. Sa conception, de même que sa validation, ont suivi les quatre étapes
proposées par la concertation de l’American Educational Research Association, l’American
Psychological Association et le National Council of Measurement in Education (AERA, APA, &
NCME, 1999) ainsi que par Laveault et Grégoire (2014). La première étape consiste à expliciter
le construit théorique, qui permet de faire le point sur les perspectives pertinentes dans
l’évaluation de la compréhension des inférences chez les enfants d’âge préscolaire. La deuxième
étape repose sur la création d’un outil d’évaluation à partir des éléments conceptuels et
méthodologiques importants mis en évidence par une recension des écrits (Filiatrault-Veilleux,
Bouchard, Trudeau, & Desmarais, 2015). La troisième étape vise à valider le contenu des
variables sélectionnées à l’aide d’experts dans le domaine. Finalement, la quatrième étape
comprend la réalisation d’analyses statistiques pour déterminer les qualités psychométriques de
l’outil.
Dans une perspective développementale, les écrits scientifiques proposent des pistes
intéressantes pour conceptualiser la compréhension des inférences chez les enfants qui n’ont pas
46
encore appris à lire. La méthode actuellement privilégiée pour évaluer cette habileté en contexte
de langage oral chez les jeunes enfants est la lecture partagée (Desmarais et al., 2013b; Makdissi
& Boisclair, 2006; van Kleeck et al., 2006). La lecture partagée consiste en la lecture d’une
histoire où l’adulte interagit avec l’enfant, en lui offrant des modèles et en lui posant des
questions afin, entre autres, de l’aider à aller plus loin dans sa compréhension du récit
(Whitehurst & Lonigan, 1998). Ce type de lecture offre donc de nombreuses occasions pour
amener l’enfant à faire des inférences, ce qui tend à augmenter son intérêt envers l’histoire
(Desmarais et al., 2013b; Justice & Kaderavek, 2004; Whitehurst & Lonigan, 1998).
Le livre utilisé doit être composé d’une structure narrative classique, qui devient
prévisible pour l’enfant. Les éléments de la grammaire de récit qui constituent cette structure
classique sont une situation initiale, un événement déclencheur, un problème qui génère une
réponse interne, un but qui motive les actions du personnage au cours du récit, des tentatives de
résolution du problème et une solution (Mandler & Johnson, 1977; Stein & Glenn, 1979). Dans
ce contexte facilitant et signifiant, où les jeunes enfants peuvent inférer grâce au soutien de
l’adulte qui les guide dans leur compréhension de l’histoire, ils sont capables de répondre à des
questions inférentielles de différents types (Bianco & Coda, 2002; Desmarais et al., 2013b; Ford
& Milosky, 2003, 2008; Makdissi & Boisclair, 2006; Trabasso & Nickels, 1992; van den Broek
et al., 2005; van Kleeck et al., 2006).
Comme il s’agit d’une habileté en développement, il importe de s’intéresser aux types de
questions inférentielles qui seraient comprises précocement par les jeunes enfants. Les questions
ciblant les éléments de la grammaire de récit (c'est à dire le problème, les réponses internes, le
but du personnage, les tentatives pour le résoudre, la prédiction de la prochaine étape et la
solution de l’histoire) sont reconnues pour être comprises tôt par les enfants d’âge préscolaire
(Filiatrault-Veilleux et al., 2015; Makdissi & Boisclair, 2006). Ces types d’inférences sont
appelées « inférences causales », c’est-à-dire qu’elles requièrent la compréhension d’un lien de
causalité entre des événements (Bianco & Coda, 2002; Dupin de Saint-André, 2008; Graesser,
Singer, & Trabasso, 1994; Johnson & Johnson, 1986; Lefebvre, Bruneau, & Desmarais, 2012;
Richards & Anderson, 2003). En plus d’émerger précocement, les inférences causales sont
considérées « nécessaires » à la compréhension d’une histoire, s’avérant essentielle pour bien
saisir le sens d’un récit (Bianco & Coda, 2002). Finalement, selon d’autres auteurs, ces inférences
47
permettent également d’assurer la « cohérence globale » d’un récit, en reliant ensemble les idées
quant aux buts/motivations des personnages qui ne sont pas explicitement mentionnées dans
l’histoire (Currie & Cain, 2015). La classification des inférences causales basées sur les éléments
de la grammaire de récit utilisée dans cette étude est présentée dans le Tableau 1 ( Bianco &
Coda, 2002; Bowyer-Crane & Snowling, 2005; Makdissi & Boisclair, 2006; van Kleeck, 2008).
En contexte de lecture partagée, il apparaît également primordial de concevoir une tâche
qui peut soutenir et optimiser cette habileté en développement chez l’enfant. Les éléments se
dégageant d’une récente recension des écrits (Filiatrault-Veilleux et al., 2015) ont guidé
l’élaboration de l’outil. Premièrement, la modalité employée pour présenter l'histoire à l’enfant
peut jouer un rôle dans sa performance. Comme l'enfant n’est pas encore apte à lire une histoire,
l'information doit être présentée à l’oral et, idéalement, accompagnée d’illustrations, voire de
courtes vidéos servant de support visuel au récit (Gibbons, Anderson, Smith, Field, & Fisher,
1986; Kendeou, Bohn-Gettler, White, & van den Broek, 2008; Wenner, 2004). En effet,
l’utilisation d’une combinaison de ces modalités de présentation optimise l’attention de l’enfant
(Verhallen, Bus, & de Jong, 2006), tout en soutenant ses habiletés inférentielles (Ford &
Milosky, 2003).
Deuxièmement, les enfants peuvent générer plus facilement des inférences au cours de la
lecture du récit, plutôt qu’à la toute fin (Makdissi & Boisclair, 2006). En effet, lors de
l’administration d’une tâche, poser des questions après la présentation de l’histoire semble
surcharger davantage les habiletés mnésiques du jeune enfant plutôt que d’évaluer ses habiletés
de compréhension (Blanc, 2010; van den Broek, Tzeng, Risden, Trabasso, & Basche, 2001).
Ainsi, une piste prometteuse consiste à explorer l’habileté à faire des inférences pendant la
présentation de l’histoire. Cette méthode favoriserait une augmentation de la complexité des
expressions causales produites pendant le récit (Desmarais et al., 2013b; Makdissi & Boisclair,
2006; van Kleeck et al., 2006).
Troisièmement, la méthode sélectionnée pour évaluer le niveau de performance de
l’enfant doit également tenir compte d’un continuum de degré de compréhension inférentielle,
plutôt que d’utiliser un système dichotomique de réponse correcte vs incorrecte. Un enfant peut
effectivement déduire une partie de la réponse qui n'est pas entièrement vraie ou fausse, mais qui
48
se situe entre les deux, sur un continuum de qualité. La classification des types de réponses
données par les enfants en fonction de leur niveau de qualité utilisée dans certaines études permet
une meilleure discrimination des performances des enfants (Adams et al., 2009; Desmarais et al.,
2013b; Gnepp & Gould, 1985; Spackman et al., 2006). Par exemple, à la question “Que va faire
le garçon ?” après avoir cassé son jouet préféré, un enfant qui répond « il va demander à son père
pour de l’aide » plutôt que « le garçon va pleurer » pourrait démontrer un niveau de
compréhension plus élevé.
En s’appuyant sur les recommandations de création d’outils d’évaluation robustes (AERA
et al., 1999; Laveault & Grégoire, 2014), la présente étude vise à concevoir un nouvel outil
d’évaluation de la compréhension des inférences en contexte de récit et à en déterminer la validité
de contenu, de même qu’à en mesurer les qualités psychométriques. Sur le plan de la validité de
contenu, elle vise à évaluer, via une consultation d’experts cliniques (a) le matériel de l’outil que
constitue la structure narrative du récit, les questions posées et la qualité des illustrations qui
accompagnent le récit ; (b) la fonctionnalité via le médium utilisé, soit un iPad et (c) les scores
obtenus par des enfants à la tâche sur un continuum de qualité afin de concevoir une échelle de
codification. En ce qui concerne les qualités psychométriques, l’étude vise à déterminer la
validité de construit et la validité concurrente de l’outil de même que sa fidélité inter-juge et test-
retest.
Méthode
L’étude a reçu l’approbation éthique de l’Institut de Réadaptation en Déficience Physique
de Québec (IRDPQ). Considérant sa nature, elle s’est déroulée en deux phases distinctes et les
éléments de cette section sont donc présentés de la façon suivante : 1) conception de l’outil Un
abri pour Pinson et validité de contenu auprès d’experts (participants, matériel et procédure) et 2)
qualités psychométriques de l’outil (participants, matériel, procédure et analyse des données).
1) Conception de l'outil: Un abri pour Pinson et validité de contenu
Comme les nouvelles technologies prennent un essor considérable en évaluation auprès des
jeunes enfants (Brandone, Michnick Golinkoff, & Hirsh-Pasek, 2008), l’outil d’évaluation a été
49
conçu sur iPad dans le but d’uniformiser la procédure de passation. Ce médium permet de
combiner des illustrations attrayantes à une présentation audio de la narration, des questions et
des effets sonores, ce qui permet une standardisation essentielle dans un contexte d’outil de
mesure. Six types d'inférences causales ciblant les éléments de la grammaire de récit d’une
histoire sont mesurés par l’outil: le problème, les réponses internes, le but du personnage, les
tentatives de résolution du problème, les prédictions et la solution. Le nombre, le contexte de
l’histoire et les exemples de questions posées sont présentés dans le Tableau 6.
50
Tableau 6. Classification des six inférences ciblant les éléments de la grammaire de récit ainsi que le nombre, le contexte de l’histoire
et les exemples de questions inférentielles élaborées pour l’outil Un abri pour Pinson
Types d’inférence N Contexte de l’histoire Questions inférentielles
Événement déclencheur/
problème du personnage
2 L’orage brise le nid de Pinson. Oh oh! Qu’est-ce qui arrive ?
Pourquoi Pinson se cherche un abri ?
Réponse interne 4 Il a peur, car ses parents quittent le nid et
l’orage arrive.
Il est triste, car son nid se brise.
Il est découragé, car il ne trouve pas d’abri.
Il est content, car il retrouve ses parents et il a
une nouvelle maison.
Comment il se sent Pinson ? Pourquoi ?
(justification de la réponse interne) (question
posée à quatre moments différents)
But du personnage 2 Il se cherche un abri. Il veut quoi Pinson?
Tentatives de résolution du
problème
7 Il rencontre trois animaux qui tentent de l’aider,
mais qui n’y arrivent pas (une moufette, un
porc-épic et un castor).
Est-ce que Pinson va être bien dans cette
maison ? Pourquoi ?
(prédiction du succès de la tentative de
résolution, 4 questions)
Est-ce un bon abri pour Pinson ? Pourquoi ?
(évaluation du résultat de la tentative de
résolution, 3 questions)
Prédiction de la prochaine
étape
2 La branche d’arbre va casser en raison de
l’orage.
Il va arriver quoi, tu penses ?
C’est quoi l’idée de Tom ?
51
Tom va lui construire une cabane en bois.
Conséquences ou solution 2 Tom lui a construit une cabane en bois bien
solide.
Est-ce un bon abri pour Pinson ? Pourquoi ?
(évaluation de la solution)
Qui a le plus aidé Pinson selon toi ? Pourquoi ?
52
L’application iPad Un abri pour Pinson apparaît sous la forme d’un livre d’histoire pour
enfant comprenant vingt illustrations réalisées par une graphiste. Le résumé de l’histoire est
présenté en Annexe. Les deux premières pages du livre préparent l’enfant au fonctionnement de
l’outil en l’exposant à cinq questions littérales. La tâche consiste ensuite à répondre à dix-neuf
questions inférentielles, soit une par page (pages 3 à 20), à l’exception d’une page où deux
questions sont posées (page 19). Lors de l’administration, à chacune des pages l’enfant est invité
à regarder l’illustration tout en écoutant la narration, puis à répondre à la question inférentielle
posée. Les questions sont posées pendant la présentation de l’histoire.
Les réponses produites par les enfants sont codifiées selon un continuum de qualité en
quatre niveaux afin de pouvoir obtenir une gradation de la performance de cette habileté en
développement. En s’inspirant du Preschool Language Assessment Instrument–Second Edition
(PLAI-2) (Blank, Rose, & Berlin, 2003) et d’un précédent projet de recherche (Desmarais et al.,
2012; Desmarais et al., 2013b), les réponses produites par les enfants sont classées en quatre
catégories de réponses {A, B, C, D} sur un continuum de qualité. Le Tableau 7 présente le nom
et la définition des catégories de réponses ainsi que des exemples de réponses et le nombre de
points accordés (de 0 à 3 points).
Tableau 7. Définition des quatre catégories de réponses sur un continuum de qualité, exemples
de réponses obtenues à la question « Il veut quoi Pinson, tu penses ? » et points accordés.
Continuum de qualité des réponses
Faible Élevée
Catégorie D - Inadéquate C - Peu contingente B - Incomplète A - Attendue
Définition
Absence de
réponse ou
réponse erronée.
Trop vague ou n’est
pas directement en lien
avec ce qui est attendu
Logique, mais
incomplète ou
imprécise
Correspond
entièrement à la
réponse cible
Exemples
de réponses
Il veut aller dans
l’eau. Il veut manger.
Il veut retrouver
ses parents.
Il veut se trouver
un abri.
Points
accordés 0 1 2 3
53
Participants. Pour évaluer la validité de contenu de l’outil, six orthophonistes œuvrant dans des
contextes différents (centre de réadaptation; CSSS; cabinet privé) ont été invités à participer. Les
orthophonistes devaient: (1) être membres de l'Ordre des Orthophonistes et Audiologistes du
Québec (OOAQ) et (2) présenter un minimum de cinq ans d'expérience auprès d’enfants d’âge
préscolaire avec des difficultés langagières. De plus, un chercheur expert dans le domaine de la
compréhension inférentielle chez les jeunes enfants a été sollicité pour participer à l’étape de la
validation des scores obtenus.
Matériel et procédure. La validité de contenu a été effectuée en deux temps au moyen de deux
rencontres de type focus groupe (Tableau 8). (1) Une version papier de l’outil comprenant les
vingt illustrations, la narration associée et les questions à poser aux enfants a été remise aux six
orthophonistes. Pour chaque élément de cotation envisagé, les orthophonistes ont évalué la
pertinence selon une échelle de Likert {très pertinent, assez, peu, pas pertinent}. Une première
rencontre (focus groupe #1) d’une durée de 2 heures a ensuite permis d’échanger sur le contenu
et a conduit à des ajustements. (2) Après cette rencontre, le contenu de l’outil a été transféré sur
l’application iPad. La même procédure de validation a été reprise pour cette autre version de
l’outil. Avant la deuxième rencontre (focus groupe #2), les orthophonistes devaient utiliser la
version iPad avec trois enfants âgés de 3 à 6 ans. Les deux focus groupes ont été réalisés dans le
but d’obtenir un consensus pour chacun des éléments discutés. Le but et les éléments discutés
lors des deux focus groupes sont présentés dans le Tableau 8. Cette étape de validation du
contenu visait à optimiser l’outil, avant son utilisation auprès d’enfants.
Tableau 8. But et résumé des éléments évalués par les experts cliniciens lors des deux focus
groupes.
But de la rencontre Éléments évalués et discutés
Focus
groupe
#1
Valider le contenu du
matériel employé de
l’outil en version papier.
Pertinence et clarté de l’histoire ; adaptée aux enfants
de 3 à 6 ans ; respect de la structure du récit.
Formulation des questions posées ; qualité et quantité
quant aux six types d’inférences à évaluer.
Clarté et caractère non équivoque des illustrations ;
visuellement attrayantes pour les jeunes enfants.
54
Focus
groupe
#2
Valider la fonctionnalité
de l’outil sur iPad
suivant la passation à
trois enfants.
Facilité d'utilisation de l’outil sur iPad.
Attrait et attention des enfants lors de la passation.
Fonctionnement des paramètres de l’application.
Temps de passation.
2) Qualités psychométriques de l’outil
Participants. Cent-vingt-et-un enfants au développement typique (59 filles et 62 garçons) ont été
recrutés, en collaboration avec des centres de la petite enfance (CPE) et des écoles primaires de la
région de Québec, de même que par le biais d’une invitation par courriel aux étudiants et
employés de l’Université Laval. Les parents des enfants devaient au préalable avoir signé le
formulaire de consentement lié à l’approbation par le comité d’éthique de la recherche de
l’Institut de Réadaptation en Déficience Physique de Québec (IRDPQ) (#2013-309). Les enfants
devaient respecter les critères d’inclusion suivants : (1) avoir entre 3;0 et 6;11 ans et (2) détenir le
français comme langue maternelle. De plus, ceux présentant un retard de langage ou tout autre
problème cognitif ou développemental, rapporté via un questionnaire complété par les parents,
étaient exclus de la recherche. Les 121 enfants ont été divisés en trois groupes en fonction de leur
âge et du milieu fréquenté (milieu de garde ou maternelle). La description de l’échantillon est
présentée dans le Tableau 9.
Tableau 9. Caractéristiques des trois groupes de participants : moyenne (et écart-type) pour l’âge,
pourcentage de mères ayant terminé leurs études secondaires et pourcentage de familles ayant un
revenu supérieur au seuil de faible revenu.
Groupe N
(Filles-Garçons)
Âge moyen
en mois (E-T)
Scolarité mère
> DES (%)
Revenu >
SFR (%)
3 ans fréquentant un
milieu de garde
30
(14-16)
42,16
(3,44)
100 100
4 et 5 ans fréquentant un
milieu de garde
37
(18-19)
56,18
(6,17)
100 91,9
5 et 6 ans fréquentant la
maternelle
54
(27-27)
73,27
(3,67)
100 94,4
55
Matériel. Trois tâches ont été administrées aux enfants. La première tâche est Un abri pour
Pinson, le nouvel outil de la compréhension des inférences. La seconde tâche est un outil
expérimental de la compréhension des inférences, élaboré à partir du livre Petit cochon s’est
perdu (Desmarais, Nadeau, Trudeau, Filiatrault-Veilleux, & Maxes-Fournier, 2013a)5. Comme il
n’existe pas de test de référence absolu permettant de mesurer la compréhension inférentielle en
contexte de lecture partagée chez les enfants, cette tâche expérimentale a été sélectionnée
puisqu’elle s’est avérée sensible au développement langagier dans un précédent projet
(Desmarais et al., 2013a). Dans Petit cochon s’est perdu, dix questions inférentielles ciblant les
éléments du récit et les connaissances du monde sont posées aux enfants pendant la lecture.
Finalement, une mesure de la taille du vocabulaire réceptif, un construit démontré comme étant
relié à la compréhension inférentielle (Currie & Cain, 2015; Florit, Roch, Altoè, & Levorato,
2009; Kendeou et al., 2008), a été réalisée grâce à l’Échelle de vocabulaire en images Peabody
(EVIP) (Dunn, Thériault-Whalen, & Dunn, 1993).
Procédure. La collecte de données était d’une durée d’une heure au CPE ou à l’école fréquenté
par chaque participant à la recherche. Les tâches ont été administrées par deux assistantes de
recherche ayant reçu une formation au préalable afin qu’elles soient familières avec les
procédures de passation. Les rencontres étaient filmées afin de faciliter la codification des
résultats. Des 121 enfants recrutés, 34 ont été revus deux semaines suivant la première
administration afin de déterminer la fidélité test-retest de l’outil.
Analyse des données
L’analyse des données a été réalisée avec le logiciel SPSS version 22. Pour évaluer la validité de
construit, une analyse en composantes principales a été réalisée avec une rotation varimax (Kim
& Mueller, 1978; Tabachnick & Fidell, 2001). Cette analyse factorielle a permis d’examiner les
patrons de corrélations entre les six types d’inférences mesurées par l’outil de mesure Un abri
pour Pinson. À cette occasion, l’indice Kaiser-Meyer-Olkin (KMO) a permis d’évaluer la qualité
5 Conçu à partir du livre pour enfant Petit cochon s’est perdu (Amery & Cartwright, 2001) en
s’inspirant des travaux réalisés par van Kleeck et al. (2006).
56
représentationnelle de l’échantillon, c’est-à-dire à quel point la matrice de corrélation est
cohérente. Idéalement, l’indice KMO doit se situer entre 0,6 et 0,9 (Kaiser, 1974). Pour
déterminer la validité concurrente, des mesures de coefficients de corrélation de Pearson ont été
réalisées entre le nouvel outil et les deux autres tâches (Petit cochon s’est perdu, Desmarais et al.,
(2013a) et l’ÉVIP, Dunn et al., (1993)). La fidélité inter-juge a été mesurée avec la statistique
pondérée AC1 de Gwet (Wongpakaran, Wongpakaran, Wedding, & Gwet, 2013), un estimateur
développé pour corriger les paradoxes de la statistique Kappa de Cohen. La pondération permet
aussi de prendre en compte la taille de l’écart entre les cotes attribuées par les juges. En effet, une
différence de 1 entre deux catégories (p.ex. A et B), équivalait à un coefficient de 0.89, une
différence de 2 (p.ex. A et C) à un coefficient de 0.56, alors qu’une différence de 3 (p.ex. entre A
et D) à un coefficient de 0. Finalement, une analyse de coefficient intra-classe (ICC) de type 2,1 a
été employée afin de mesurer la fidélité test-retest à 2 semaines d’intervalle (modèle mixte, ICC
relatif). En effet, cette analyse s’avère préférable au coefficient de corrélation de Pearson (r), car
elle permet de comparer la variabilité des cotations d'un même sujet à la variation totale du test et
de tous les sujets (Steffen & Seney, 2008).
Résultats
Comme pour la section Méthode, les résultats sont présentés en deux sections distinctes : 1)
validité de contenu (matériel employé et fonctionnalité sur iPad) et 2) qualités psychométriques
(validité de construit, validité concurrente, fidélité inter-juge et test-retest).
1) Validité de contenu
Le matériel employé dans Un abri pour Pinson. Les experts ont jugé très pertinentes la structure
narrative du récit (pour 73% des éléments à juger), les questions posées (72%) et dans une
moindre mesure, les illustrations (57%). Tous les éléments jugés « assez pertinents » (structure
narrative du récit = 27%, questions posées = 27% et illustrations = 36%) et « peu pertinents »
(questions posées = 1% et illustrations = 7%) ont été discutés lors du premier focus groupe afin
de dégager un consensus quant aux modifications à apporter à l’outil. À titre d’exemple, la
formulation de certaines questions a été simplifiée et adaptée pour se rapprocher davantage du
langage utilisé avec de jeunes enfants en contexte franco-québécois. Également, certains éléments
des illustrations jugés non pertinents et pouvant nuire à l’interprétation visuelle ont été retirés.
57
La fonctionnalité de l’outil sur iPad. Le deuxième focus groupe a permis d’obtenir un consensus
quant aux modifications à effectuer à l’application informatique. Il est à noter que puisque les
modifications suggérées ont été abordées dans le contexte d’une discussion qui a mené à un
consensus, aucune analyse quantitative n’a été effectuée quant à l’échelle de pertinence. Suite au
focus groupe, certains paramètres et fonctionnalités ont été modifiés pour en simplifier
l’utilisation. Par ailleurs, le temps de passation, d’environ 15 minutes, était jugé adéquat. Les
résultats obtenus lors des deux focus groupes ont donc permis de modifier l’outil d’évaluation
afin de vérifier que le contenu de l’outil était adéquat et que son administration sur iPad
convenait à des enfants âgés de 3 à 6 ans.
2) Qualités psychométriques
Validité de construit. La matrice de corrélations soumise à l’analyse en composantes principales
affiche une excellente structure représentationnelle (indice KMO = ,865; p<,001; Kaiser, 1974).
Quatre facteurs ont été retenus parce que leur valeur propre (eigenvalue) était supérieure à un. Le
Tableau 10 présente les résultats de l’analyse en composantes principales à 4 facteurs après
rotation varimax (Tabachnick & Fidell, 2001), et ce, pour les six types d’inférences. Seules les
saturations supérieures ou égales à .4 en valeur absolue sont soulignées. Les types d’inférence
« Résolution de problème » (facteur 1) (eigenvalue = 1,5, saturation=,855), But (facteur 2,
eigenvalue = 1,4, saturation=,930), Événement déclencheur (facteur 3, eigenvalue = 1,3,
saturation=,917) et Solution (facteur 4, eigenvalue=1,2, saturation=,867) sont des sources de
variance presque indépendantes. Elles apportent donc des informations différentes. Les types
d’inférence « Réponse interne » et « Prédiction » sont associés à plusieurs facteurs dont
l’information est déjà contenue dans les 4 types d’inférence précédents. Selon l’analyse en
composantes principales, ces deux types sont plus hétérogènes et conséquemment plus difficiles à
préciser conceptuellement. Ces deux types d’inférences pourraient également partager des
sources de variance avec les autres types de questions posées. En effet, la « Réponse Interne »
partage vraisemblablement des sources de variance avec « Résolution de problème », « But » et
« Solution » alors que la « Prédiction » pourrait partager des sources de variance avec
« Événement déclencheur » et « Résolution de problème ».
58
Tableau 10. Résultats de l’analyse factorielle après rotation varimax de l’outil Un abri pour
Pinson.
Facteurs
1 2 3 4
Événement déclencheur .149 .125 .917 .248
Réponse interne .487 .554 .144 .450
But .162 .930 .161 .148
Résolution du problème .855 .160 .155 .325
Prédiction .607 .348 .571 .087
Solution .291 .201 .273 .867
Validité concurrente. Les résultats obtenus par les enfants aux trois tâches sont présentés dans le
Tableau 11. Les résultats montrent une forte corrélation de Pearson (r(119) = .77 ; p<.001) entre
le nouvel outil Un abri pour Pinson et la tâche expérimentale Petit cochon s’est perdu.
Également, ils partagent 58% de variance commune. Il existe une corrélation moindre, quoique
modérée, entre Un abri pour Pinson et l’ÉVIP-A, soit de r(119) = .43, p<.001 et où la part de
variance commune est de 18%.
Tableau 11. Résultats des trois groupes de participants à l’outil Un abri pour Pinson, à la tâche
Petit cochon s’est perdu et à l’ÉVIP.
Groupe Un abri pour
Pinson M (É-T)
Petit cochon s’est
perdu M (É-T)
EVIP
score brut M (É-T)
3 ans fréquentant un milieu de
garde (N=30) 46.1% (18.4) 44.6% (15.8) 41.0 (13.0)
4 et 5 ans fréquentant un milieu
de garde (N=37) 68.9% (11.1) 69.4% (15.0) 69.5 (16.7)
5 et 6 ans fréquentant la
maternelle (N=54) 78.6% (7.1) 78.9% (8.9) 89.8 (13.8)
59
Fidélité inter-juge de la grille de codification. La fidélité inter-juge de l’utilisation de la grille de
codification a été mesurée à deux reprises. La première mesure a porté sur la classification des
scores obtenus par les enfants à la tâche. Un document compilant toutes les possibilités de
réponses aux questions fournies par les 121 enfants a été soumis à trois juges experts, soit deux
orthophonistes ayant participé aux focus groupes et un chercheur expert dans le domaine. Les
trois juges ont été invités à compléter une grille de codification des réponses en accordant la
catégorie de qualité de réponses (A, B, C ou D) qu’ils jugeaient la plus appropriée pour chaque
réponse obtenue, et ce, pour chaque question. Les décisions quant aux catégories de réponse à
sélectionner ont été prises selon le nombre de juges en accord pour chaque question de l’histoire.
Pour cette analyse, les statistiques AC1 de Gwet (pondération quadratique) pour chaque question
de l’outil se situent entre 0.53, p<.001 (question 3) et 0.96 p<.001 (question 19), pour une
moyenne de 0.83. Cette première analyse de fidélité inter-juge a permis de concevoir la version
finale de la grille de codification.
Une seconde analyse inter-juge visait la vérification de cette version finale de la grille.
Pour ce faire, deux examinateurs à l’aveugle ont codifié les réponses des enfants à l’aide de la
version finale de la grille de codification, et ce, sur 20% des données (n = 25 enfants sélectionnés
au hasard). Les valeurs des statistiques AC1 de Gwet avec la version finale de la grille de
codification de la qualité des réponses des enfants sont plus élevées et se situent entre 0.92
p<.001 et 1.00, pour une moyenne de 0.99. Les statistiques AC1 de Gwet et les intervalles de
confiance à 95% pour chacune des 19 questions de l’outil lors des deux analyses inter-juge sont
présentées dans le Tableau 12.
60
Tableau 12. Statistiques AC1 de Gwet et intervalle de confiance à 95% pour chaque question lors
des deux analyses inter-juges.
No.
questions
Trois experts lors de la conception de la
grille de codification
Deux examinateurs à l’aveugle qui
utilisent la version finale de la grille
AC1 de Gwet IC à 95% AC1 de Gwet IC à 95%
1 0.84* 0.78 – 0.91 0.99* 0.98 – 1.00
2 0.82* 0.59 – 1.00 0.98* 0.93 – 1.00
3 0.53* 0.21 – 1.00 1.00* 1.00 – 1.00
4 0.82* 0.67 – 0.98 0.99* 0.94 – 1.00
5 0.86* 0.67 – 1.00 1.00* 1.00 – 1.00
6 0.91* 0.82 – 1.00 0.96* 0.91 – 1.00
7 0.95* 0.90 – 1.00 0,94* 0.76 – 1.00
8 0.81* 0.50- 1.00 1.00* 1.00 – 1.00
9 0.80* 0.45 – 1.00 0.99* 0.96 – 1.00
10 0.93* 0.84 – 1.00 1.00* 1.00 – 1.00
11 0.88* 0.72 – 1.00 0.98* 0.94 – 1.00
12 0.86* 0.75 – 0.98 0.99* 0.97 – 1.00
13 0.87* 0.64 – 1.00 1.00* 1.00 – 1.00
14 0.85* 0.58 – 1.00 0.99* 0.97 – 1.00
15 0.70* 0.11 – 1.00 0.98* 0.95 – 1.00
16 0.84* 0.67 – 1.00 0,96* 0.88 – 1.00
17 0.86* 0.65 – 1.00 0.98* 0.94 – 1.00
18 0.86* 0.74 – 0.97 0.97* 0.91 – 1.00
19 0.96* 0.86 – 1.00 0.95* 0.87 – 1.00
*p.001
Fidélité test-retest. La stabilité temporelle de l’outil à deux semaines d’intervalle a été réalisée
sur 34 enfants sélectionnés au hasard, soit 28% de l’échantillon, basée sur les procédures
employées par le Wechsler Preschool and Primary Scale of Intelligence – Third Edition (WPPSI-
III) (Wechsler, 2002). Les résultats obtenus par les enfants lors des deux temps de mesure sont
présentés dans le Tableau 13. Le coefficient de corrélation intra-classe (ICC) de type 2,1 entre les
61
deux temps de mesure est très élevé, valeur relative de l’ICC (33) = 0.948, p<.001 (Shrout, 1998).
Également, les résultats au deuxième temps de mesure sont plus élevés qu’au premier temps de
mesure. Le test T pairé révèle effectivement que les scores au temps 2 (deux semaines post)
s’avèrent significativement plus élevée (M =76.47) que ceux au temps 1 (M = 67.08), t(33) = -
6.461, p< .001. Les enfants ont obtenu en moyenne 9,4% de plus au temps 2 que lors de la
première administration.
Tableau 13. Résultats obtenus par les 34 enfants sélectionnés au hasard au test-retest à 2 semaines
d’intervalle.
Groupe
Un abri pour Pinson M (E-T) Différence de
moyenne (E-T) Temps 1 Temps 2
3 ans fréquentant un milieu de
garde (n=8)
41.5% (22.8) 53.2% (21.6) +11.7% (11.8)
4 et 5 ans fréquentant un milieu
de garde (n=10)
73.2% (14.1) 80.6% (11.1) +7.4% (7.5)
5 et 6 ans fréquentant la
maternelle (n=16)
76.0% (7.6) 85.5% (6.4) +9.5% (7.3)
T (n=34) 67.1% (20.0) 76.47% (18.08) +9.4% (8.5)
Discussion
Un abri pour Pinson constitue un outil d’évaluation sur iPad de la compréhension de six types
d’inférences causales ciblant les éléments de la grammaire de récit. Il a été conçu de manière à
tenir compte d’éléments méthodologiques précédemment mis en lumière dans les écrits
scientifiques (Bianco & Coda, 2002; Filiatrault-Veilleux et al., 2015; Gibbons et al., 1986;
Kendeou et al., 2008; van den Broek et al., 2005; van Kleeck, 2008; Wenner, 2004). Cette étude
visait à en déterminer la validité de contenu, de même que les qualités psychométriques. Elle a
ainsi permis de démontrer que l’outil est valide et fidèle pour mesurer cette habileté chez les
enfants âgés de 3 à 6 ans.
62
Validité de contenu l’outil. L’outil a été validé à l’aide d’orthophonistes et d’un chercheur expert
dans le domaine. Cette étape de validation de contenu a permis d’optimiser l’outil en cours de
création, tant sur le plan du matériel employé (structure narrative du récit, questions posées et
illustrations qui accompagnent l’histoire) que de sa fonctionnalité et son applicabilité sur iPad.
Compte tenu des résultats obtenus lors des focus groupes, des modifications importantes (ex. :
contenu de l’histoire et des illustrations, formulation des questions et paramètres de l’application
informatique) ont été réalisées afin de le rendre le plus adéquat et fonctionnel possible pour une
utilisation auprès de jeunes enfants.
Qualités psychométriques. En ce qui concerne la validité de construit, nos résultats indiquent que
l'ensemble des variables mesurées, soit les différents types d’inférences, forme un ensemble
cohérent permettant de mesurer adéquatement la compréhension inférentielle chez les enfants de
3 à 6 ans dans un contexte de récit (Tabachnick & Fidell, 2001). Bien que l’analyse factorielle ait
permis de distinguer conceptuellement plus facilement quatre types d’inférences, soit les
tentatives de résolution de problème, le but, l’événement déclencheur et la solution, elle met en
évidence que la réponse interne et la prédiction partagent des sources de variance avec d’autres
types d’inférences. Comme les six inférences mesurées sont de type « causales », elles requièrent
par conséquent la compréhension d’un lien de causalité entre des événements (Bianco & Coda,
2002; Dupin de Saint-André, 2008; Graesser et al., 1994; Johnson & Johnson, 1986; Lefebvre et
al., 2012; Richards & Anderson, 2003). Ceci s’avère particulièrement vrai pour ces deux types
d’inférences plus difficiles à distinguer conceptuellement. À titre d’exemple, dans la réponse
interne, afin de bien saisir l’émotion ressentie par le personnage, la compréhension du but de
l’histoire, soit la cause de l’émotion, est essentielle. La compréhension des tentatives de
résolution du problème et de la solution de l’histoire sont également liées à l’émotion du
personnage (par ex. : Pinson est content, parce qu’il a une nouvelle maison et que ses parents sont
revenus). De même, l’inférence de type « prédiction » est une inférence causale liée à
l’événement déclencheur et aux tentatives de résolution du problème. L’enfant ayant besoin de
capacité prédictive pour y arriver. Ainsi, dans Un abri pour Pinson, ces deux types d’inférences
apparaissent davantage dépendants et impliquent des processus déjà contenus dans les inférences
précédentes.
63
La validité concurrente démontre quant à elle que l’outil permet d’obtenir des scores
fortement corrélés à ceux d’une tâche expérimentale, Petit cochon s’est perdu (Desmarais et al.,
2013a), mesurant un construit similaire. Il est à noter que puisque Petit cochon s’est perdu ne
comprenait que 10 questions inférentielles, était administré en format papier et qu’il mesurait des
inférences portant à la fois sur le récit et les connaissances du monde, la corrélation entre les deux
ne pouvait être parfaite. D’autre part, la valeur de la corrélation obtenue entre le nouvel outil et le
vocabulaire réceptif était plus faible, s’expliquant entre autres par le fait que d’autres habiletés
langagières et cognitives sont impliquées dans la compréhension d’inférences, en plus du niveau
de vocabulaire réceptif (Florit, Roch, & Levorato, 2011; Kendeou et al., 2008).
À partir d’une démarche réalisée en collaboration avec des experts dans le domaine, il a
été possible de concevoir un outil qui permet de capter la performance des enfants au construit
complexe qu’est la compréhension inférentielle en contexte de récit. En effet, tout en tenant
compte du fait que les réponses s’appuient sur un processus de réflexion et non pas sur une
décision oui/non, l’outil propose une échelle de codification valide s’appuyant sur un continuum
de qualité des réponses en fonction d’une cible. Un abri pour Pinson démontre aussi de bonnes
qualités psychométriques quant à la fidélité. Grâce à une démarche minutieuse, la fidelité inter-
juge de la grille de codification créée a atteint un niveau très élevé. Il faut toutefois rappeler que
cette dernière a été évalué dans un deuxième temps auprès de deux juges seulement, plutôt que
trois, ce qui augmente nécessairement la valeur obtenue. Cependant, de façon générale, les
mesures obtenues tendent à démontrer que la compréhension inférentielle chez les jeunes enfants
peut être documentée de façon valide et fidèle, quand la démarche employée tient compte des
caractéristiques du construit.
Finalement, l’outil a démontré une bonne stabilité temporelle à 2 semaines d’intervalle.
Étant donné qu’il s’agissait de la même histoire, les enfants ont obtenu un résultat
significativement plus élevé au deuxième temps de mesure. Ceci n’est pas surprenant compte
tenu de l’effet connu et documenté dans les écrits scientifiques de la lecture répétée d’une même
histoire (Morrow, 1988; Verhallen et al., 2006).
Forces et limites de la présente étude. Bien que la compréhension inférentielle soit un construit
complexe à définir et, par le fait même, difficile à mesurer, l’outil d’évaluation Un abri pour
64
Pinson permet de répondre à ce besoin. En effet, il a permis de dégager des résultats valides et
fidèles avec des enfants de 3 à 6 ans dans un contexte narratif. Qui plus est, cette étude mènera
sous peu à la diffusion de cet outil d’évaluation, un apport considérable à la pratique clinique
franco-québécoise auprès de la jeune clientèle pédiatrique (Bouchard et al., 2009). Un abri pour
Pinson permettra aux orthophonistes et autres professionnels ayant une expertise en
développement langagier de l’enfant de mesurer une habileté langagière démontrée essentielle à
la compréhension en lecture (Cain & Oakhill, 2007; Ferreiro & Taberoski, 1982; Joffe et al.,
2007; Kontos & Wells, 1986; van Kleeck, 2008) et par le fait même, à la réussite éducative du
jeune enfant. Effectivement, il faut souligner l’unique et relativement indépendante contribution
de la compréhension inférentielle sur le développement de la compréhension en lecture (Bianco
et al., 2010; Kendeou et al., 2009; NICHD, Early Child Care research Network, 2005). Les
habiletés en compréhension orale chez les jeunes enfants, dont la compréhension inférentielle,
jouent un rôle de premier plan dans l’apprentissage et la réussite en compréhension en lecture à
l’âge scolaire (Cain, Oakhill, & Elbro, 2003; Kintsch & Kintsch, 2005; Trabasso & Wiley, 2005;
van den Broek et al., 2005; Zwaan & Singer, 2003). Il importe donc d’outiller les orthophonistes
dans l’évaluation et la prévention des enfants, et ce, dès leur plus jeune âge.
Cependant, une limite importante de ce projet de recherche concerne la représentativité de
l’échantillon. En effet, comme l’échantillonnage n’est pas représentatif de l’ensemble des enfants
québécois de 3 à 6 ans, les données développementales obtenues ne permettent de proposer des
normes développementales. Une utilisation de type critérié visant à décrire le développement des
enfants est donc conseillée.
Perspectives futures : l’outil d’évaluation, utile en clinique comme en recherche. L’outil
d’évaluation de la compréhension des inférences Un abri pour Pinson a été conçu afin d’être utile
non seulement en recherche, mais également dans les milieux cliniques. En recherche, ce nouvel
outil pourrait contribuer à uniformiser les méthodes employées pour pallier la grande diversité
des outils actuellement utilisés (Filiatrault-Veilleux et al., 2015; Florit et al., 2011; Hudson &
Slackman, 1990). En contexte clinique, plus précisément auprès des petits, les orthophonistes
pourront désormais avoir accès à un outil valide et fidèle permettant d’évaluer et de documenter
le niveau de développement des enfants. Également, il pourrait être pertinent de poursuivre la
65
validation de cet outil auprès d’un plus grand échantillon et d’autres populations, tels qu’auprès
d’enfants de familles à plus faible revenu ainsi qu’auprès d’autres types de clientèles d’enfants
(ex. : déficience langagière ou trouble du spectre de l’autisme).
Conclusion
L’outil Un abri pour Pinson a été conçu selon une procédure rigoureuse et structurée et ses
qualités psychométriques confirment qu’il est fidèle et valide. Malgré le fait que l’évaluation de
la compréhension inférentielle demeurera un défi pour les orthophonistes, ce nouvel contribue à
la pratique clinique en proposant une démarche standardisée permettant de décrire la
compréhension inférentielle du jeune enfant. Cet outil pourra aussi être utilisé en recherche dans
le but de poursuivre le développement des connaissances sur la compréhension inférentielle
pendant la petite enfance.
66
Note des auteurs
Les demandes au sujet de cet article devraient être adressées à Chantal Desmarais, Ph.D.,
orthophoniste et professeure agrégée à l’Université Laval, 1050, avenue de la Médecine, Québec,
Qc, G1V 0A6. [email protected]
Remerciements
Les auteurs désirent remercier les programmeurs en informatique de la compagnie Skyglow ayant
conçu l’application informatique ainsi que la graphiste Édith Julien qui a réalisé les illustrations
de l’outil. Ils souhaitent également exprimer leur reconnaissance aux orthophonistes et au
chercheur ayant participé à la validation de contenu de l’outil de même qu’aux enfants et aux
parents ayant participé à la collecte des données. Il est à noter que cette recherche a été financée
par le Conseil de recherches en sciences humaines du Canada.
67
Annexe
Résumé de l’histoire : Un abri Pour Pinson
Le personnage principal de l’histoire est un oiseau qui se nomme Pinson. Au début de
l'histoire, ses parents quittent le nid à la recherche de nourriture et Pinson reste seul. Soudain, un
orage arrive et son nid tombe et se brise. Pinson est mouillé et a peur. Il part donc à la recherche
d’un abri pour se protéger. Il rencontre alors trois animaux (une moufette, un porc-épic et un
castor) qui tentent de l'aider, mais qui n’y arrivent pas. À la fin, un petit garçon construit une
nouvelle maison pour Pinson et ses parents.
68
Références
Adams, C., Clarke, E., & Haynes, R. (2009). Inference and sentence comprehension in children
with specific or pragmatic language impairments. International Journal of Language &
Communication Disorders, 44(3), 301-318. doi: 10.1080/13682820802051788
AERA, APA, & NCME. (1999). Standards for educational and psychological testing.
Washington, DC: American Educational Research Assocation.
Amery, H., & Cartwright, S. (2001). Petit cochon s'est perdu. Usborne. Les contes de la ferme.
Bianco, M., Bressoux, P., Doyen, A.-L., Lambert, E., Lima, L., Pellenq, C., & Zorman, M.
(2010). Early Training in Oral Comprehension and Phonological Skills: Results of a
Three-Year Longitudinal Study. Scientific Studies of Reading, 14(3), 211-246.
doi:10.1080/10888430903117518
Bianco, M., & Coda, M. (2002). La compréhension en quelques points... In M. Bianco, M. Coda,
& D. Gourgue (Eds.), La compréhension (pp. 93-97). Grenoble: Éditions de la Cigale.
Bishop, D. V. M. (1997). Uncommon understanding : development and disorders of language
comprehension in children. Hove, UK: Psychology Press.
Blanc, N. (2010). La compréhension des contes entre 5 et 7 ans: Quelle représentation des
informations émotionnelles? [The comprehension of the tales between 5 and 7 year-olds:
Which representation of emotional information?]. Canadian Journal of Experimental
Psychology/Revue canadienne de psychologie expérimentale, 64(4), 256-265. doi:
10.1037/a0021283
Blank, M., Rose, S. A., & Berlin, L. J. (2003). Preschool Language Assessment Instrument –
Second Edition. Austin, TX: Pro-Ed.
Botting, N., & Adams, C. (2005). Semantic and inferencing abilities in children with
communication disorders. International Journal of Language & Communication
Disorders, 40(1), 49-66.
Bouchard, M., Fitzpatrick, E., & Olds, J. (2009). Analyse psychométrique d’outils d’évaluation
utilises auprès d’enfants francophones. Revue canadienne d’orthophonie et d’audiologie,
33(3), 129-139.
Bowyer-Crane, C., & Snowling, M. J. (2005). Assessing children's inference generation: what do
tests of reading comprehension measure? The British Journal of Educational Psychology,
75(Pt 2), 189-201.
Brandone, A. C., Michnick Golinkoff, R., & Hirsh-Pasek, K. (2008). Feasibility of Computer-
Administered Language Assessment. Perspectives on School-based Issues, 9, 57-65. doi:
doi:10.1044/sbi9.2.57
Cain, K., & Oakhill, J. (2007). Children's comprehension problems in oral and written language
: a cognitive perspective. New York: Guilford Press.
Cain, K., Oakhill, J. V., & Elbro, C. (2003). The ability to learn new word meanings from context
by school-age children with and without language comprehension difficulties. Journal of
Child Language, 30(3), 681-694.
Currie, N. K., & Cain, K. (2015). Children's inference generation: The role of vocabulary and
working memory. Journal of Experimental Child Psychology, 137, 57-75.
doi:10.1016/j.jecp.2015.03.005
Das Gupta, P., & Bryant, P. E. (1989). Young children's causal inferences. Child Development,
60(5), 1138-1146.
69
Deconti, K. A., & Dickerson, D. J. (1994). Preschool children's understanding of the situational
determinants of others' emotions. Cognition & Emotion, 8(5), 453-472. doi:
10.1080/02699939408408952
Desmarais, C., Archambault, M.-C., Filiatrault-Veilleux, P., & Tarte, G. (2012). La
compréhension d’inférences : comparaison des habiletés d’enfants de quatre et de
cinq ans en lecture partagée. Revue des sciences de l'éducation, 38(3), p. 555-578. doi:
10.7202/1022712ar
Desmarais, C., Nadeau, L., Trudeau, N., Filiatrault-Veilleux, P., & Maxes-Fournier, C. (2013a).
Effet d’une intervention visant à améliorer la compréhension inférentielle des enfants de
4 à 6 ans ayant une dysphasie : une collaboration orthophonistes-chercheurs. Glossa,
Glossa 113 Spécial XIIIèmes Rencontres d’orthophonie, 45-62.
Desmarais, C., Nadeau, L., Trudeau, N., Filiatrault-Veilleux, P., & Maxes-Fournier, C. (2013b).
Intervention for improving comprehension in 4-6 year old children with specific language
impairment: practicing inferencing is a good thing. Clinical Linguistics & Phonetics,
27(6-7), 540-552. doi: 10.3109/02699206.2013.791880
Dunn, L., Thériault-Whalen, C. M., & Dunn, M. (1993). Échelle de vocabulaire en images
Peabody. Adaptation française du Peabody picture vocabulary test-revised. Toronto, ON:
Pearson Psychcorp.
Dupin De Saint-André, M. (2008). Pistes pour travailler les inférences. Québec français, 150, 66-
67.
Ferreiro, E., & Taberoski, A. (1982). Literacy before schooling. Portsmouth, NH: Heinemann
Educational Books Inc.
Filiatrault-Veilleux, P., Bouchard, C., Trudeau, N. and Desmarais, C. (2015). Inferential
comprehension of 3-6 year-olds within the context of story grammar: a scoping review.
International Journal of Language & Communication Disorders, 50(6), 737-49. doi:
10.1111/1460-6984.12175
Filiatrault-Veilleux, P., Tarte, G., & Desmarais, C. (2015). Inferential Comprehension in a
Dialogic Reading Task: Comparing 5-year-old Children with Specific Language
Impairment to Typically Developing Peers. In G. M. Martinez (Ed.), Recent Advances in
Language and Communication (pp. pp. 119-133). New York: Nova Science Publishers,
Inc.
Florit, E., Roch, M., Altoè, G., & Levorato, M. C. (2009). Listening comprehension in
preschoolers: the role of memory. Br J Dev Psychol, 27(Pt 4), 935-951.
Florit, E., Roch, M., & Levorato, M. C. (2011). Listening text comprehension of explicit and
implicit information in preschoolers: The role of verbal and inferential skills. Discourse
Processes, 48(2), 119-138. doi: 10.1080/0163853x.2010.494244
Ford, J. A., & Milosky, L. M. (2003). Inferring emotional reactions in social situations:
differences in children with language impairment. Journal of Speech, Language &
Hearing Research, 46(1), 21-30.
Ford, J. A., & Milosky, L. M. (2008). Inference generation during discourse and its relation to
social competence: an online investigation of abilities of children with and without
language impairment. Journal of Speech, Language & Hearing Research, 51(2), 367-380.
Gibbons, J., Anderson, D. R., Smith, R., Field, D. E., & Fisher, C. (1986). Young Children's
Recall and Reconstruction of Audio and Audiovisual Narratives. Child Development,
57(4), 1014-1023.
70
Gnepp, J., & Gould, M. E. (1985). The Development of Personalized Inferences: Understanding
Other People's Emotional Reactions in Light of Their Prior Experiences. Child
Development, 56(6), 1455-1464.
Graesser, A. C., Singer, M., & Trabasso, T. (1994). Constructing inferences during narrative text
comprehension. Psychological Review, 101(3), 371-395.
Hudson, J. A., & Slackman, E. A. (1990). Children's use of scripts in inferential text processing.
Discourse Processes, 13(4), 375-385. doi: 10.1080/01638539009544766
Isbell, R., Sobol, J., Lindauer, L., & Lowrance, A. (2004). The Effects of Storytelling and Story
Reading on the Oral Language Complexity and Story Comprehension of Young Children.
Early Childhood Education Journal, 32(3), 157-163. doi:
10.1023/B:ECEJ.0000048967.94189.a3
Joffe, V. L., Cain, K., & Maric, N. (2007). Comprehension problems in children with specific
language impairment: does mental imagery training help? International Journal of
Language & Communication Disorders, 42(6), 648-664. doi:
10.1080/13682820601084402
Johnson, D. D., & von Hoff Johnson, B. (1986). Highlighting Vocabulary in Inferential
Comprehension Instruction. Journal of Reading, 29(7), 622-625.
Justice, L. M., & Kaderavek, J. N. (2004). Embedded-explicit emergent literacy intervention I:
Background and description of approach. Language, speech, and hearing services in
schools, 35(3), 201-211.
Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39, 401-415.
Kendeou, P., Bohn-Gettler, C., White, M. J., & van den Broek, P. (2008). Children's inference
generation across different media. Journal of Research in Reading, 31(3), 259-272. doi:
10.1111/j.1467-9817.2008.00370.x
Kendeou, P., van den Broek, P., White, M. J., & Lynch, J. S. (2009). Predicting Reading
Comprehension in Early Elementary School: The Independent Contributions of Oral
Language and Decoding Skills. Journal of Educational Psychology, 101(4), 765-778.
Kim, J.-O., & Mueller, C. W. (1978). Introduction to Factor Analysis. Thousand Oaks, CA:
SAGE Publications, Inc.
Kintsch, W., & Kintsch, E. (2005). Comprehension. In S. G. Paris & S. A. Stahl (Eds.), Current
issues on reading comprehension and assessment (pp. 71-92). Mahwah, NJ: Erlbaum.
Kontos, S., & Wells, W. (1986). What preschool children know about reading and how they learn
it. Young Children, 42(1), 58–66.
Laveault, D., & Grégoire, J. (2014). Introduction aux théories des tests en psychologie et en
sciences de l'éducation. Bruxelles: De Boeck.
Lefebvre, P., Bruneau, J., & Desmarais, C. (2012). Analyse conceptuelle de la compréhension
inférentielle en petite enfance à partir d’une recension des modèles théoriques. Revue des
sciences de l'éducation, 38(3), p. 533-553. doi:10.7202/1022711ar
Makdissi, H., & Boisclair, A. (2006). Interactive reading: A context for expanding the expression
of causal relations in preschoolers. Written Language and Literacy, 9(2), 177-211. doi:
10.1075/wll.9.2.02mak
Mandler, J. M., & Johnson, N. S. (1977). Remembrance of Things Parsed: Story Structure and
Recall. Cognitive Psychology, 9(1), 111-151.
Morrow, L. M. (1988). Young Children's Responses to One-to-One Story Readings in School
Settings. Reading Research Quarterly, 23(1), 89-107.
NICHD, Early Child Care research Network. (2005). Pathways to reading: the role of oral
language in the transition to reading. Developmental Psychology, 41(2), 428-442.
71
Richards, J. C., & Anderson, N. A. (2003). How Do You Know? A strategy to help emergent
readers make inferences. The Reading Teacher, 57(3), 290-293.
Ryder, N., Leinonen, E., & Schulz, J. (2008). Cognitive Approach to Assessing Pragmatic
Language Comprehension in Children with Specific Language Impairment. International
Journal of Language & Communication Disorders, 43(4), 427-447.
Shrout, P. E. (1998). Measurement reliability and agreement in psychiatry. Stat Methods Med
Res, 7(3), 301-317.
Skarakis-Doyle, E., & Dempsey, L. (2008). The detection and monitoring of comprehension
errors by preschool children with and without language impairment. Journal of Speech
Language and Hearing Research, 51(5), 1227-1243. doi: 10.1044/1092-4388(2008/07-
0136)
Spackman, M. P., Fujiki, M., & Brinton, B. (2006). Understanding emotions in context: the
effects of language impairment on children's ability to infer emotional reactions.
International Journal of Language & Communication Disorders, 41(2), 173-188. doi:
10.1080/13682820500224091
Steffen, T., & Seney, M. (2008). Test-retest reliability and minimal detectable change on balance
and ambulation tests, the 36-item short-form health survey, and the unified Parkinson
disease rating scale in people with parkinsonism. Physical Therapy, 88(6), 733-746. doi:
10.2522/ptj.20070214
Stein, N. L., & Glenn, C. G. (1979). An analysis of story comprehension in elementary school
children. (Vol. Discourse processing: Multidisciplinary perspectives). Norwood, NJ:
Ablex: R. Freedle (Ed.).
Tabachnick, B. G., & Fidell, L. S. (2001). Using Multivariate Statistics (4th ed.). Boston: Allyn
and Bacon.
Trabasso, T., & Nickels, M. (1992). The development of goal plans of action in the narration of a
picture story. Discourse Processes, 15(3), 249-275. doi: 10.1080/01638539209544812
Trabasso, T., & Wiley, J. (2005). Goal plans of action and inferences during comprehension of
narratives. Discourse Processes, 39, 129-164.
van den Broek, P., Kendeou, P., Kremer, K., Lynch, J. S., Butler, J., White, M. J., & Lorch, E. P.
(2005). Assessment of comprehension abilities in young children. In S. Stahl & S. Paris
(eds.), Children’s Reading Comprehension and Assessment, (pp.107-130). Mahwah, NJ:
Erlbaum.
van den Broek, P., Tzeng, Y., Risden, K., Trabasso, T., & Basche, P. (2001). Inferential
questioning: Effects on comprehension of narrative texts as a function of grade and
timing. Journal of Educational Psychology, 93(3), 521-529. doi: 10.1037/0022-
0663.93.3.521
van Kleeck, A. (2008). Providing preschool foundations for later reading comprehension : the
importance of and ideas for targeting inferencing in storybook-sharing interventions.
Psychology in the schools, 45(7), 627-643.
van Kleeck, A., Vander Woude, J., & Hammett, L. (2006). Fostering literal and inferential
language skills in Head Start preschoolers with language impairment using scripted book-
sharing discussions. American Journal of Speech-Language Pathology, 15(1), 85-95. doi:
10.1044/1058-0360(2006/009)
Verhallen, M. J. A. J., Bus, A. G., & de Jong, M. T. (2006). The Promise of Multimedia Stories
for Kindergarten Children at Risk. Journal of Educational Psychology, 98(2), 410-419.
Wechsler, D. (2002). Wechsler Preschool and Primary Scale of Intelligence™ - Third Edition
(WPPSI™ - III). San Antonio, TX: The Psychological Corporation.
72
Wenner, J. A. (2004). Preschoolers' comprehension of goal structure in narratives. Memory,
12(2), 193-202. doi: 10.1080/09658210244000478
Whitehurst, G. J., & Lonigan, C. J. (1998). Child development and emergent literacy. Child
Development, 69(3), 848-872.
Wongpakaran, N., Wongpakaran, T., Wedding, D., & Gwet, K. L. (2013). A comparison of
Cohen's Kappa and Gwet's AC1 when calculating inter-rater reliability coefficients: a
study conducted with personality disorder samples. BMC Medical Research Methodology
, 13, 61. doi: 10.1186/1471-2288-13-61
Zwaan, R.A. & Singer, M. (2003). Text comprehension. In A.C. Graesser, M.A. Gernsbacher &
S.R. Goldman (Eds.), Handbook of discourse processes (pp. 83-121). Mahwah, NJ:
Erlbaum.
73
Chapitre 4. Développement de la compréhension des
inférences causales chez les enfants âgés de 3 à 6 ans
Le présent chapitre consiste en une étude empirique portant sur le développement de la
compréhension des inférences causales en contexte de récit chez les enfants de 3 à 6 ans, qui a été
réalisée à l’aide de l’outil sur iPad, Un abri pour Pinson. Cet article est intitulé “Comprehension
of inferences in a narrative in 3 to 6 year old children” et est inséré tel que re-soumis le 30
novembre 2015 pour publication au Journal of Speech Language and Hearing Research suite aux
demandes de corrections des réviseurs.
Filiatrault-Veilleux, P., Desmarais, C., Bouchard, C. and Trudeau, N. (Re-soumis le 30-11-2015
au Journal of Speech Language and Hearing Research). Comprehension of inferences in a
narrative in 3 to 6 year old children.
Résumé
Objectif : Cette étude vise à décrire le développement des habiletés inférentielles des enfants
âgés de 3 à 6 ans en contexte de lecture partagée sur un outil iPad.
Méthode : Les participants sont 121 enfants au développement typique âgés de 3 à 6 ans divisés
en trois groupes selon les tranches d’âge (3-4 ans, 4-5 ans et 5-6 ans). Des analyses ANOVA sont
réalisées afin de vérifier si les groupes diffèrent quant au score total de compréhension
inférentielle, aux types d’inférences causales comprises et à la qualité des réponses produites.
Résultats : L’habileté à comprendre des inférences apparaît tôt, dès l’âge de 3-4 ans, et se
développe significativement pendant la petite enfance. Les résultats indiquent que, dès l’âge de 4-
5 ans, les enfants font des inférences en lien avec les réponses internes, le problème de l’histoire
et les prédictions. Vers 5 et 6 ans, les enfants deviennent plus habiles à inférer le but du récit, les
tentatives de résolution du problème et la solution. Finalement, en grandissant, les enfants se
rapprochent des cibles attendues en termes de qualité de réponses produites.
74
Conclusion : Cette étude contribue à une meilleure caractérisation d’un pilier important à la
compréhension en lecture et ouvre sur des pistes de réflexion quant à la stimulation de cette
habileté au préscolaire.
75
Abstract
Purpose: This study aims to describe the development of inferential abilities of children aged 3
to 6 years in a narrative using a dialogic reading task on iPad.
Method: Participants were 121 typically developing children, divided into three groups
according to age range (3-4 years old; 4-5 years old; 5-6 years old). Total score of inferential
comprehension, subscores by causal inference type targeting elements of the story grammar, and
quality of response were examined across groups.
Results: Inferential comprehension emerged early, from 3-4 years old, with considerable inter-
individual variability. Inferential comprehension scores increased significantly in relation to age,
leading to developmental steps with regards to the type of causal inferences. The ability to infer
the problem of the story, the internal response of a character, and predictions were easier starting
at age 4. Then, the 5-6 years old were better able to infer the goal, the attempt to solve the
problem, and the resolution. Finally, between the ages of 3 and 6, children improved in terms of
the quality of response they provided.
Conclusion: This study addresses important gaps in our knowledge of inferential comprehension
in young children and has implications for planning of early education in this realm.
Key words: inferential comprehension, causal inferences, dialogic reading, early education.
76
Introduction
Inferential abilities play a critical role in reading comprehension and, therefore, in the
educational success of children (Cain, Oakhill, & Elbro, 2003; Gineste & Le Ny, 2002; Kendeou,
van den Broek, White, & Lynch, 2009; Lynch & van den Broek, 2007; Paul, 2000; van Kleeck,
2008). This has been the topic of numerous studies relating to school age children (Bishop &
Adams, 1990; Catts, 2009; Oakhill & Cain, 2000; Serpell, Baker, & Sonnenschein, 2005).
Indeed, the link between good inferential skills and reading comprehension has often been
highlighted (Cain & Oakhill, 1999; Cain et al., 2003; Oakhill & Cain, 2000; Trabasso & Wiley,
2005; van den Broek, Tzeng, Risden, Trabasso, & Basche, 2001).
Besides this interest, recent studies have demonstrated that inferential abilities emerge
within the preschool period, before children enter school (Blanc, 2010; Filiatrault-Veilleux,
Bouchard, Trudeau, & Desmarais, 2015; Florit, Roch, & Levorato, 2014; Kendeou, Bohn-Gettler,
White, & van den Broek, 2008; Makdissi & Boisclair, 2006; Reed, Hurks, Kirschner, & Jolles,
2015; Tompkins, Guo, & Justice, 2013; van Kleeck, Vander Woude, & Hammett, 2006). From
the age of 3 or 4, children begin to be able to construct a coherent mental representation of a story
heard by using complex cognitive processes such as inference production (Filiatrault-Veilleux et
al., 2015; Florit et al., 2014; Kendeou et al., 2008; Potocki, Ecalle, & Magnan, 2012; Tompkins
et al., 2013; van den Broek, Kendeou, Kremer, Lynch, Butler, White & Lorch, 2005). Moreover,
some authors suggest that fostering inferential abilities would be beneficial in the preschool
period (Desmarais, Nadeau, Trudeau, Filiatrault-Veilleux, & Maxes-Fournier, 2013; Reed et al.,
2015; van Kleeck, 2008; van Kleeck et al., 2006) due to the likely impact on later reading
comprehension (Cain & Oakhill, 1999; Cain et al., 2003; Kintsch & Kintsch, 2005; Trabasso &
Wiley, 2005; van den Broek et al., 2001).
While there is growing interest for this topic, it is still not well known when and how
inferential comprehension emerges and develops in young children because few studies have
been conducted on this subject. Hence, a better description of the developmental steps in
inferential comprehension that occur in the preschool period is needed. Such information would
provide evidence to the field of early intervention and would foster new initiatives to promote the
development of inferential comprehension in young children. We sought to address this
77
important gap in the research by investigating the inferential comprehension of typically
developing children between 3-6 years in a narrative using a dialogic reading task.
Inferential Comprehension within a Dialogic Reading Task
Inferential comprehension is defined as the ability to understand a message in a context
where some elements are not explicitly known (Bishop, 1997; Gineste & Le Ny, 2002; van
Kleeck, 2008). To obtain a valid measure of this ability with children who have not yet learned to
read, a task that does not rely on reading must be used (van den Broek et al., 2005). In recent
studies, dialogic reading stands out as the task of choice (Ford & Milosky, 2008; Makdissi &
Boisclair, 2006; Price, van Kleeck, & Huberty, 2009; Tompkins et al., 2013; van Kleeck et al.,
2006). For both assessment and intervention, a storybook is read aloud to children and inferential
questions about elements that are not explicitly presented in the story are embedded in the
interaction between the adult and the child (Blanc, 2010; Justice & Kaderavek, 2004; Price et al.,
2009; van Kleeck, 2008). To answer correctly, children must make connections between what is
heard in the story, the context in which it is occurring, and their world knowledge (Desmarais et
al., 2013; Paul, 2000; van Kleeck, 2008). Beyond their role of potentially fostering
comprehension, the questions help maintain the child’s participation and interest in the story
(Ford & Milosky, 2008; Makdissi & Boisclair, 2006; van Kleeck et al., 2006; Wenner, 2004).
Dialogic reading with children optimally relies upon a format of a predictable story
structure minimally comprising the following story grammar elements: a setting, a triggering
event, a problem that generates an internal response, a goal that motivates the character's actions
during the narrative, an attempt to solve the problem and its consequence, and finally, a
resolution to the story (Mandler & Johnson, 1977; Stein & Glenn, 1979). A variety of inference
questions can be asked within a story. For the preschool population, the most pertinent questions
target causal inferences that are directly related to elements of story grammar (Bowyer-Crane &
Snowling, 2005; Makdissi & Boisclair, 2006; van Kleeck, 2008). This is important because
causal inferences are crucial to establishing causal connections between events in the story
(Bianco & Coda, 2002; Blanc, 2009, 2010; Filiatrault-Veilleux et al., 2015; Graesser, Singer, &
Trabasso, 1994; Lefebvre, Bruneau, & Desmarais, 2012; Reed et al., 2015; Richards & Anderson,
2003; Tompkins et al., 2013). As they listen to narratives, children learn to create coherent and
78
meaningful representations of a story that include concepts about causal relationships between
story elements (Lever & Senechal, 2011; Reed et al., 2015). This is why most studies of young
children have specifically focussed on causal inferences (Lynch & van den Broek, 2007;
Makdissi & Boisclair, 2006; Reed et al., 2015; Tompkins et al., 2013; Wenner, 2004).
Causal Inferences Targeting Elements of Story Grammar
To date, some studies have attempted to propose developmental steps in the acquisition of
inferential comprehension, but results are heterogeneous which is largely due to methodological
differences between studies including the definition of inferences, the tasks used and the
diversity in the age groups that are examined, etc. (Blanc, 2009, 2010; Filiatrault-Veilleux et al.,
2015). Nevertheless, van den Broek and colleagues (2005) proposed broad developmental steps
in the comprehension of inferences without associating them with precise age ranges because of
the many factors involved in inference comprehension including the influence of the child’s
knowledge and experience as well as the variability in the complexity of inferences. In this
proposal, the first step is the ability to establish inferential links between concrete physical events
that occur close together (e.g., a simple prediction about a doll broken if a child pulls its arm).
The second type of inference concerns the ability to identify relationships between more distant
events. Thirdly, children learn to identify causal relationships involving characters’ goals,
emotions, and desires (e.g. consequently, the little girl is sad and wants her mom to repair the
doll). As they get older, children can then more easily recognize the hierarchical and thematic
relations between clusters of events. Finally, children will understand the translation of the story
theme into a moral or lesson (e.g. it’s important to play carefully with toys).
As mentioned before, researchers have paid particular attention to causal inferences
because of their centrality to the comprehension of narratives. At this time, the two inference
types that have been most examined in the literature are the internal response and the
goal/purpose of the protagonist in a story (Adams, Clarke, & Haynes, 2009; Florit, Roch, &
Levorato, 2011; Ford & Milosky, 2008; Gnepp & Gould, 1985; Makdissi & Boisclair, 2006;
Spackman, Fujiki, & Brinton, 2006; Wenner, 2004). For example, if a child listens to a story
about a little boy who lost his favorite toy, this child could infer that the boy is sad (the internal
response) and that he would want to find it (the goal), even if these elements are not explicitly
79
stated. Some studies show that the ability to infer the internal response of a character appears
early in the development, between 3 and 4 years of age (Adams et al., 2009; Blanc, 2010; Ford &
Milosky, 2008; Gnepp & Gould, 1985; Lynch & van den Broek, 2007). In addition, from the age
of 4, children may also demonstrate some ability to infer the problem (Makdissi & Boisclair,
2006) and the goal of the character in a story (Wenner, 2004). It should be highlighted that the
goals are important because they provide the motivation for the actions that are carried out
throughout a narrative (Lynch & van den Broek, 2007). From the age of 5, children begin to
spontaneously produce characters’ goals in narrative recall (Trabasso & Nickels, 1992). Children
between the age of 5 and 6 are able to make inferences concerning the consequences of events in
short pictorial sequences (Schmidt & Paris, 1978), infer the attempt to solve a problem (e.g. the
little boy will look in the playroom for his toy) (Kendeou et al., 2008), and demonstrate the
ability to predict the next event of a story (Adams et al., 2009; Crais & Chapman, 1987). In sum,
these studies suggest that children as young as 6 years of age can make all these inference types
while listening to a story. In light of these results, there may be predictable developmental steps
in the acquisition of inferential comprehension with regard to the causal inference types related to
story grammar elements (Blanc, 2010; Filiatrault-Veilleux et al., 2015; Makdissi & Boisclair,
2006; van den Broek et al., 2005). There is a gap in the literature in that regard because none one
of the studies to date have examined all these inferences types in one task and no study has
included children across the age range from 3 to 6.
Quality of Response
In a dialogic reading task, children’s responses to inferential questions illustrate whether
or not they understand the links between events presented in the story. Recent studies have been
interested in the quality of responses that children provide to inferential questions because it most
likely reflects a refinement in the ability to infer (Adams et al., 2009; Desmarais et al., 2013;
Lynch & van den Broek, 2007). For example, to the question “what will the boy do next?” in
reference to the lost toy, a child who answers “The boy will look everywhere to find the toy”
instead of “the boy will go to sleep” could demonstrate a higher quality level of comprehension.
Thus, some authors suggest using a classification system to categorize the children’s responses in
order to take into account the quality of responses provided (Adams et al., 2009; Barnes, Dennis,
& Haefele-Kalvaitis, 1996; Desmarais et al., 2013; Spackman et al., 2006). The categorization of
80
children’s responses allows a nuanced analysis of their performance, thus determining the quality
level of a response and providing information for effective strategies that can be used in clinical
practice. In a recent intervention study with SLI children, coding the quality of responses
revealed that most children progressively reached the expected target (Desmarais et al., 2013). In
typically developing children, it would be of interest to investigate whether the quality of
response develops with age. Such information would be of great value to clinicians in selecting
and sequencing intervention targets and strategies.
This study aims to describe inferential abilities in the context of a narrative of children
aged 3 to 6. A task, created on the iPad, was used to collect that data that was then compared
across three age groups: (1) 3-4 year olds; (2) 4-5 year olds and (3) 5-6 year olds. The first
objective was to compare the total score obtained across group as well as the subscores on each
causal inference type targeting elements of story grammar. The second objective was to compare
the groups in terms of the quality of responses produced by children.
Method
Participants
One hundred and twenty-one (59 girls and 62 boys) French-speaking typically developing
preschoolers ranging from 3 to 6 years of age were recruited in collaboration with daycare centers
and schools in Quebec City, Canada. Some children were also recruited through an email
invitation to students and employees of Université Laval in Quebec City, Canada. All parents
signed informed consent on behalf of their children. The participants were divided into three
groups according to the age range (3-4 year olds; 4-5 year-olds; 5-6 year olds) and the preschool
program they were enrolled in (daycare centers or kindergarten). The gender distribution was
comparable across groups (M = 51.2 % of boys, 2 = .042, p = 0.959). Based on information
collected from questionnaire completed by the parents, children were excluded from the study if a
language delay or any other cognitive or developmental problem had been identified. To verify
the typical language development status, the French version of the Peabody Picture Vocabulary
Test (Dunn, Thériault-Whalen, & Dunn, 1993) was administered. On the sociodemographic
questionnaire, parents reported that the children spoke French as their native language, all
81
mothers had completed high school, and the majority of the families were above the low income
cut-offs (LICOs) as calculated by the Institut de la Statistique du Québec. Participant
characteristics are presented in Table 14.
Table 14. Participants’ characteristics.
Group N
(Girls-Boys)
Mean age
Months (S-D)
PPVT Percentile
Rank (S-D)
Income > LICOs
(%)
3-4 years old from
daycare center
30
(14-16)
42,2
(3,4)
70,43
(25,23)
100
4-5 years old from
daycare center
37
(18-19)
56,2
(6,2)
87,49
(14,2)
91,9
5-6 years old from
kindergarten
54
(27-27)
73,3
(3,7)
90,17
(16,88)
94,4
Material
For the purpose of this study, an application (app) for the iPad entitled “Un abri pour
Pinson” (“A Shelter for Pinson”) was designed (Filiatrault-Veilleux, Desmarais, Bouchard,
Trudeau, & Leblond, in Press). This tool consists of a 20 page story that follows a predictable
narrative structure, i.e. initiative event/problem, internal response, goal, prediction, attempt to
solve the problem and resolution. It is a story of a bird, named Pinson. It starts when Pinson’s
parents go away to find food and he stays alone in his nest. A storm begins and his nest falls. He
is wet and scared and goes to look for a new place to stay. He meets three characters (a skunk, a
porcupine and a beaver) who try to help him but nothing works. In the end, a boy builds a new
house for Pinson and his parents come back. Nineteen questions are asked on-line, within the
story, to assess six causal inference types that target the story grammar elements. The inference
types, story context, examples of questions and scoring of the task on iPad are shown in Table 15
(inspired from Botting & Adams, 2005; Makdissi & Boisclair, 2006; van Kleeck, 2008).
Administration and scoring were finalized after a thorough validation process reported in
a previous paper (Filiatrault-Veilleux et al., in Press) and summarized here. The responses
obtained to questions are scored in four categories following a quality continuum from expected
82
to inadequate (A-Expected, B-Incomplete, C-Low contingency, D-Inadequate or off topic). This
scale was inspired by the scoring system of the Preschool Language Assessment Instrument-
Second Edition (PLAI-2) (Blank, Rose, & Berlin, 2003) and was used in a previous experimental
task of inferential comprehension in story context (Desmarais, Archambault, Filiatrault-Veilleux,
& Tarte, 2012; Desmarais et al., 2013; Filiatrault-Veilleux, Tarte, & Desmarais, 2015). The total
score is calculated out of 78 points and then converted to a percentage score. The definitions,
examples of responses, and scores attributed for each category are presented in Appendix A. In
order to create the scoring system, the first step was carried out by three experts in the field who
read the verbatim transcription of the children’s responses and coded the quality of the responses
according to the scale from A (expected) to D (inadequate or off topic). These external experts
also provided recommendations for scoring. For each response to the 19 questions provided by
the 121 participants, assignment to a category was done according to the number of judges who
had selected that category. In the next step of the validation of the scoring system, two
independent judges coded 20% of the responses with the scoring scale obtained in step one. This
yielded a high inter-rater reliability (Gwet’s AC1 value = 0.99, p < .01).
The tool’s adequacy with regards to psychometric properties was confirmed and is
reported in a previous paper (Filiatrault-Veilleux et al., in press). Of note, the Kaiser-Meyer-
Olkin (KMO) value resulting from the factorial analysis of the six inference types of the tool
indicated good sampling adequacy (KMO value = 0.87, p < .01). The concurrent validity with an
experimental tool of inferential comprehension was high (r(119) = 0.77, p < .01), while the
convergent validity demonstrated a moderate link with receptive vocabulary (r(119) = 0.43, p <
.01) (as assessed with French version of the Peabody Picture Vocabulary Test (Dunn et al.,
1993)) which is consistent with previous studies where the link between receptive vocabulary and
inferential comprehension has been shown (Currie & Cain, 2015; Florit et al., 2014; Kendeou et
al., 2008). The test -retest reliability two weeks post on 20% of participants was excellent (ICC
(33) = 0.95, p < .01).
83
Table 15. Inference types, number of questions asked, story context, examples from the task and scores
Inference types N Story context Examples of questions6 Scores
Problem 2 The storm breaks the nest Pinson. What is happening? / 6
Internal response 4 Afraid because his parents leave the nest and the storm
arrives. / Sad because his nest is broken. / Discouraged
because he cannot find a shelter. / Happy because his
parents are back and has a new home.
How does Pinson feel?
Why?
Emotions / 12
Justification / 12
Goal of the
character
2 He is searching for a shelter. What is Pinson looking for? / 6
Attempts to solve
the problem
7 He meets three animals that try to help but cannot do it (a
skunk, porcupine and beaver).
Do you think Pinson will be
ok with this animal? Why?
Prediction / 16
Evaluation / 12
Prediction 2 The tree branch will break due to the storm. / Tom will
build him a wood bird house
What do you think will
happen next?
/ 6
Resolution 2 Tom built him a solid wood bird house. Who helped him the most?
Why?
/ 8
6 Filiatrault-Veilleux, P., Desmarais, C., Bouchard, C., Trudeau, N., & Leblond, J. (in press). Conception et qualités psychométriques d’un outil d’évaluation de la
compréhension d'inférences en contexte de récit chez les enfants âgés de 3 à 6 ans. Revue Canadienne d’Orthophonie et d’Audiologie.
84
Procedure
Children were seen individually by an experimenter in a quiet room at their daycare or
school. They were instructed to listen closely to the story so that they could answer the questions
embedded into the story in the app. On each page, the child looked at the illustration while
listening to the narration in audio output from the app and, subsequently, to a question also in
audio output from the app. The first two pages of the story served as training for the task with
five literal questions. On pages 3 to 20, one inferential question per page was programmed into
the app except for one page comprising two questions. The data collection was videotaped for
subsequent analysis. Each participant’s responses were coded in four categories (A-Expected, B-
Incomplete, C-Low contingency, D-Inadequate or off topic) based on the scoring system
described above. The administration of the task lasted approximately 15 minutes.
Data analysis
Data analysis was performed using SPSS Version 22. Analyses of variance (ANOVA)
and multivariate analyses of variance (MANOVA) were used to verify if there were differences
between the three groups in terms of total inferential score, subscores by inference type, and
number of expected responses obtained. A t-test was also used to verify if there was a gender
effect. Based on the significant differences that emerged, post hoc analyses using the Bonferroni
correction were performed due to the multiple comparisons tested.
Results
Total Score of Inferential Comprehension
The total scores obtained by each group on the task are presented in Figure 4. The
ANOVA revealed a significant difference between groups F (2,118) = 71.389, p .001. Post hoc
analysis showed that 3-4 year olds’ results (M = 46.1%, SD = 18.4, range = 6.4-76.9) were
significantly weaker than those of 4-5 year olds (M = 68.9%, SD = 11.1, range = 39.7-92.3), who
in turn obtained results also significantly weaker than those of kindergarteners of 5-6 years of
age (M = 78.6%, SD = 7.1, range = 62.8-89.7). The majority of the 3-4 year old children were
able to answer some inferential questions and to get through the task. However, as we can see in
85
Figure 4, there was greater variance in scores for the youngest age group compared to the two
older age groups.
Figure 4. Scatter plot by groups of the means and standard deviation for the total score of
inferential comprehension.
Subscores by Inference Types
Table 16 provides the descriptive statistics by groups for each subscore, corresponding to
the six types of causal inferences included in the task (i.e. internal response, problem, goal,
attempt to solve the problem, predictions and resolution). Figures 5 and 6 illustrate the
performance of children, with means and standard deviation of each group for each inference
types. The MANOVA showed a significant effect of age group on inference type: Problem (F
(2,118) = 19.397, p .001); Internal response (F (2,118) = 27.497, p .001); Goal (F (2,118) =
0
10
20
30
40
50
60
70
80
90
100
Per
cen
tage
(%)
Mean significantly different between groups (p .01)
Standard deviation by each group
3-4 years old 4-5 years old 5-6 years old
86
23.342, p .001); Attempt to solve the problem (F (2,118) = 50.077, p .001); Prediction (F
(2,118) = 33.008, p .001) and Resolution (F (2,118) = 34.628, p .001). For the inferences
targeting the Problem, the Internal response and the Prediction (Figure 5), there was a significant
difference between the 3-4 year old children and the two older groups. However, no difference
was found between the two older groups (Problem: t(2) = -0.647, p = 1.00; Internal response: t(2)
= -1.367, p = .523; Prediction: t(2) = - 2.143, p = .103). For the Goal, the Attempt to solve the
problem, and the Resolution inference types (Figure 6), the 3-4 year olds were significantly
weaker than the 4-5 year olds who were also significantly weaker than the kindergarteners of 5-6
years old. Again, the variance in scores for the participants in each group and for each inference
type decreased with age.
Finally, independent t-test analyses revealed no gender effect for the total inferential score
(t(119) = -0.514, p=0.608), age (3-4: t(28) = -0.450, p =0.656 / 4- 5: t(35) = 0.061, p =0.952 / 5-6:
t(52) = -0.315 5, p =0,754) or inference type (Internal response : t(119) = -0.25, p=0,81; Goal :
t(119) = 0.05, p =0,96; Problem: t(119) = -1.99, p=0,05; Attempts to solve the problem : t(119) =
-0.32, p=0,75; Prediction : t(119)= -0.10, p =0,92; Resolution: t(119) = -0.89, p =0,38).
Table 16. Subscores in percentage for the six inference types of each group.
Inference types 3-4 years old
(n=30)
4-5 years old
(n=37)
5-6 years old
(n=54)
Problem 53.3%* (18.8) 72.5% (18.5) 74.7% (11.1)
Internal response 48.9%* (27.1) 73.8% (17.1) 79.1% (11.9)
Prediction 40.6%* (20.4) 67.1% (19.4) 75.9% (18.5)
Goal of the character 52.8%* (28.4) 71.6%* (23.5) 85.2% (12.4)
Attempts to the problem 41.5%* (21.3) 62.7%* (14.1) 76.6% (12.1)
Resolution 47.1%* (26.8) 72.6%* (20.8) 84.3% (13.1)
*Significantly lower scores than the older group, p < .01
87
Figure 5. Scatter plot by groups of the mean and standard deviation of the inferences targeting the
problem, the internal response and the prediction.
0
10
20
30
40
50
60
70
80
90
100
Per
cen
tage
(%)
3 year-olds from daycares4 ans 5 year-olds from daycares5 and 6 year-olds from kindergarten
Problem Internal response Prediction
Mean significantly higher than the younger group (p .01)
No significant difference with the younger group
Standard deviation by each group
3-4 years old
4-5 years old
5-6 years old
88
Figure 6. Scatter plot by groups of the mean and standard deviation of the inferences targeting the
goal, the attempt to solve the problem, and resolution.
Quality of Responses Obtained
To address the second objective, we compared the three groups on the quality of
responses obtained. A cumulative frequency histogram shows the mean number of responses per
category (A-Expected, B-Incomplete, C-Low contingency, D-Inadequate or off topic) obtained
by each group. As can be seen in the histogram, the quality of responses increased with age
(Figure 7). Indeed, an ANOVA confirmed the significant difference between groups for the
expected response category, the most contingent response to the question, F (2,118) = 53.186, p
.001. The 3-4 year olds produced significantly less expected responses than the 4-5 year olds,
who in turn, produced significantly less expected responses than the 5-6 year olds. As for the
0
10
20
30
40
50
60
70
80
90
100
Per
cen
tage
(%)
3 year-olds from daycares
4 and 5 year-olds from daycares
5 and 6 year-olds from kindergarten
Goal Attempt to solve the problem Resolution
Mean significantly higher than the younger group (p .01)
No significant difference with the younger group
Standard deviation by each group
3-4 years old
4-5 years old
5-6 years old
89
production of the acceptable response, which may be imprecise or incomplete, the ANOVA
showed no difference between groups, F (2,118) = 2.957, p = .056. With regards to low-
contingency and inadequate responses categories, the two ANOVAs were significant (Low-
contingency: F (2,118) = 18.879, p .001; Inadequate: F (2,118) = 43.306, p .001). The
frequency of low-contingency responses, in which the information is not relevant to the question,
decreased significantly with age by comparing the three groups. Finally, the group of 3-4 year
olds produced more inadequate responses than the two older groups. In sum, children improve
with age in terms of the quality of response they provide. Appendix B shows the changes from 3
to 6 years of age in terms of the quality of responses provided to an inference question about the
"Attempts to solve the problem".
*Mean significantly higher than the older group (p .01)
+Mean significantly higher than the younger group (p .01)
Figure 7. Cumulative frequency histogram of mean number of responses per category.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
3-y.-o. from daycares 4- and 5-y.-o. from daycares 5- and 6-y.-o. from
Kindergartens
D-Inadequate or off topic C-Low contingency B-Incomplete A-Expected
+
*
*
*
+
3-4 years old 4-5 years old 5-6 years old
90
Discussion
The aim of the current study was to describe the inferential abilities of children between
the ages of 3 and 6 in the context of a story. Children’s performance, using a dialogic reading app
entitled “A Shelter for Pinson”, was measured across three groups, i.e. 3-4 years old, 4-5 years
old and 5-6 years old. The first objective was to compare the total quantitative score obtained
across groups as well as the subscores by causal inference type targeting elements of story
grammar. The second objective was to assess whether the groups varied in terms of the quality of
responses produced by children.
The results for the first objective indicate that inferential comprehension in the context of
a story emerges early, from 3 years of age, and increases gradually afterwards. Moreover,
increments in the development of inferential comprehension by inference type were related to the
children’s age, supporting the notion of a developmental trajectory for that ability. From 4 to 5
years of age, children more easily inferred the problem of the story and the internal responses of
the character. They were also providing good predictions for the next steps in the story. For their
part, children at age 5 to 6 were better at making inferences about the goal of the main character,
the attempts to solve the problem, and the resolution of the story. The results of the second
objective reveal that the quality of responses produced by children also increased and evolved
significantly with age.
Taken together, these results contribute to a better understanding of young children’s
inferential comprehension. They provided answers to important research questions, which
received little attention to date. This understanding could subsequently lead to innovative
educational and interventional strategies in preschool contexts.
Development of Inferential Comprehension
With regards to the total score of inferential comprehension, our results support previous
studies by demonstrating that inferential comprehension emerges early in the development of
children, within the preschool period and certainly well before reaching school (Filiatrault-
Veilleux et al., 2015; Florit et al., 2014; Kendeou et al., 2008; Makdissi & Boisclair, 2006;
91
Potocki et al., 2012; Reed et al., 2015; Trabasso & Nickels, 1992; van den Broek et al., 2005; van
Kleeck, 2008). As anticipated, with an appropriate tool taking into account developmental
considerations (Filiatrault-Veilleux et al., 2015; van den Broek et al., 2005), even very young
children are able to answer questions suggesting that they are able to construct a mental
representation of a story heard by using complex cognitive processes such as inference generation
(Kendeou et al., 2008; Potocki et al., 2012; Trabasso & Nickels, 1992; van den Broek et al.,
2005; van Kleeck, 2008). Furthermore, the use of an iPad in this research is in line with recent
studies arguing that using non-traditional narrative media presentations, such as television or
interactive “e-books,” may reduce cognitive load, improve recall of narrative events, and enhance
story comprehension for children (Burris & Brown, 2014; Verhallen, Bus, & de Jong, 2006).
Another result of note is the greater variance in scores for the youngest age group when
compared to the two older age groups. This inter-individual variability may be linked to the
young age of children, a phenomenon well documented in the literature relating to other language
abilities (Fenson et al., 1994) and may also depend on many factors such as experience with
dialogic reading or socioeconomic background (Justice, Chow, Capellini, Flanigan, & Colton,
2003). Indeed, the link between exposure to language and literacy at home or in preschool
contexts and the development of oral language abilities and emergent literacy has been clearly
demonstrated (Frijters, Barron, & Brunello, 2000; Justice et al., 2003; Leseman & de Jong, 1998;
Purcell-Gates, 1996). Yet, these factors require further examination in order to more clearly
understand their impact, particularly with regards to the emergence of inferential comprehension
and its development. Therefore, expectations about the performance of very young children
should take inter-individual variability into consideration. With age, inferential scores become
more homogenous, a finding that is consistent with general cognitive development.
Our study also described the performance of children, with a considerable age range, in
terms of six causal inference types targeting the story grammar elements in a unique tool of
dialogic reading. Looking at our results chronologically, at 4-5 years old, children were able to
infer the problem of the story, the internal responses of characters and make predictions as
accurately as those who were 5-6 years old. For the internal response and the problem of the
story, these results were consistent with other studies. It has been previously highlighted that the
internal response inferences appear early in development, at approximately 3 to 4 years of age
92
(Ford & Milosky, 2003, 2008; Spackman et al., 2006), and may be one of the first causal
relations understood by young children within a story. It has further been suggested that the
visual characteristics of the character’s expressions present in the illustrations may help children
to infer the character’s emotions in relation with specific situations (Desmarais et al., 2013;
Makdissi & Boisclair, 2006).
The problem of the story also relies upon salient and concrete information in the story
that children have been able to infer early in development (Makdissi and Boisclair, 2006). With
regard to the ability to make predictions, literature to date has mostly addressed this type of
inference with older children ranging from 5 to 6 years of age (Adams et al., 2009; Crais &
Chapman, 1987). However, our results suggest that prediction questions could be understood
earlier in development. This is not surprising since predictions about future events as well as
internal response inferences are known to be more often stimulated by adults and educators in
dialogic reading activities (Girolametto, Weitzman, Lefebvre, & Greenberg, 2007). The results of
our study are also consistent with the developmental steps proposed by van den Broek and
colleagues in 2005. Young children begin by making inferences that link physical relations and
concrete events (van den Broek et al., 2005) as measured in our task by questions about the
problem faced by the character (e.g. the storm breaks Pinson’s nest) and prediction questions
(e.g. the tree branch will break due to the storm).The internal responses based on the character’s
facial expression and on the story context also appear to be understood early in development. In
light of these results, knowing that typically developing 3 and 4 year olds can make these types of
inferences, they would be excellent first targets in educational or clinical contexts.
There are however inference types that are more challenging for young children. For
example, the goal, the attempt to solve the problem, and the resolution of the story were harder to
infer at 4-5 years of age. As was expected, the 5-6 years old performed significantly better than
the two younger groups with regards to those three inference types. Once again, these results
align with previous findings. Kendeou and colleagues (2008) obtained a clear developmental
pattern related to the ability to infer the goal of the character from children between the ages of 4
and 8. An interesting feature of our results was that the goal was harder to understand than the
problem of the story for the 4-5 year old group. In interpreting the children’s responses, it appears
that the goal of the character (Q.: What is Pinson looking for? A: He wants to find a shelter) may
93
have been more abstract for them than the problem the character was facing (Q: What is Pinson’s
problem? A: His nest is broken). As Kendeou and colleagues (2008) stated, it might be easier for
young children to focus their attention on observable, concrete actions rather than on internal
elements such as the character goal. This could also be related to the ability to consider another
person’s perspective, in this case the character of the story, that increases with age (Frye, Zelazo,
& Palfai, 1995).
Turning to the attempt to solve the problem, in this study, the children needed to rely on
specific world knowledge to give an adequate answer (i.e. knowledge about each animal Pinson
met: a skunk that smells bad, a porcupine that may release his, quills and a beaver that lives in a
dam with underwater entrances). The two younger groups might not have acquired that specific
knowledge yet or might experience difficulties in retrieving it (Cain & Oakhill, 1999). Another
possible explanation is that the plot episodes, consisting of the attempts to solve the problem,
may carry less crucial information required for global comprehension of the story, and thus, may
appear less relevant for the two younger groups (Morrow, 1985). Moreover, based on the
proposal by van den Broek et al. (2005), it is possible that the ability to infer the attempts to solve
the problem could involve links between clusters of events instead of between unique events and
thus being more challenging for younger children.
Finally, as for the resolution of the story, the results confirm previous findings indicating
that this element is understood later in development (Makdissi & Boisclair, 2006). In fact to
properly infer the resolution, hierarchical integration of other causal chain elements is required
(Makdissi & Boisclair, 2006; van den Broek et al., 2005). The ability to understand the causal
links between the character’s goal and consequences of the attempts to solve the problem could
gradually lead to the ability to infer the resolution. This developmental trajectory of inferential
comprehension with regards to causal inferences makes sense as children gradually produce more
connections between episodes constituting the narrative. These connections appear gradually
resulting in a causal chain that helps create a complete representation of a narrative (Kendeou et
al., 2008; Makdissi & Boisclair, 2006; Trabasso & Nickels, 1992; van Kleeck, 2008). Again, our
results are in line with the developmental steps proposed by van den Broek et al. (2005). As they
get older, children get better at making links between abstract and internal events, e.g. to infer the
goal of the character, and they increasingly connect groups of events as measured with inferences
94
about attempts to solve the problem and the resolution of the story. Tompkins et al. (2013)
obtained similar results with 4 and 5-year old children who were good at recognizing key
elements in the story. For instance, they inferred what the protagonist was thinking and seeing
and responded well to questions about the causal relationships among events in a story that had
been read to them. Targeting these crucial inferences in young children could potentially provide
them with the tools necessary to better comprehend text once they begin formal reading
instruction.
Changes in Term of Quality of Responses
The method consisting of classifying children’s answers in different levels of quality has
been recently used with a population of children with SLI in order to follow their response to
intervention (Desmarais et al., 2013). With typically developing children, our study observed a
progression in terms of the quality of responses provided as children get older. For 3-and 4 year
olds, the results show that children experienced difficulties producing the right response or the
expected target. In fact, 3-4 year old children produced more inadequate and low-contingency
responses than the two older groups. In Appendix B, examples of responses to the questions
“Will Pinson be ok in the skunk’s house? and why?” and “Is it a good shelter for Pinson?”
illustrate the development of inferential comprehension as a process that is refined qualitatively
over time. In shared-book reading activities in preschool programs, adults should thus foster the
development of inferences that becoming increasingly more abstract and more elaborate to follow
the progression in the thinking process of children. This is in line with the expectation that, over
time, children progressively learn to disengage from their own perspective (Frye et al., 1995;
Wellman, Cross, & Watson, 2001; Westby & Robinson, 2014) to provide answers to questions
about a story that are closer to the answers an adult might provide. Another hypothesis, relates to
the cognitive development of children, namely that cognitive flexibility could be less developed
in the younger group (Blaye & Chevalier, 2011; Chevalier & Blaye, 2006; Chevalier & Blaye,
2008). More specifically, cognitive flexibility is involved when the child must adaptively select a
new answer based on a new situation (Chevalier & Blaye, 2006). Thus, in the earlier stages of
cognitive flexibility development, children would have more difficulty adjusting their answers
with the story. Indeed, it was observed that young children, at 3 and 4 years of age, often gave the
95
same response to all the questions about attempts to solve the problem (for example in Appendix
B : "No, because it’s too small for him"). It would be interesting to further investigate this
hypothesis in relation with cognitive abilities of young children. In sum, our results suggest that
considering the progression in the quality of the response provides important information about
development.
Limitations of the study and future directions
Despite its strengths, this study has some limitations. The first limitation is the unequal
distribution of children within groups, due to the recruitment procedure used. The intention was
to divide the children in groups of 30 by age (3, 4, 5 and 6 years old). In order to assess
inferential abilities of children before learning how to read, all 6 year old children were attending
kindergarten. As a result, the 5 year-old kindergarteners in our sample performed similarly to 6
year olds. As for 5 year olds from daycare, their scores were equivalently to the 4 year old group
from daycare. For that reason, the division of participants in three groups in function of age range
and preschool program (daycare centers and kindergarten) was seen as more advantageous for
making developmental observations. Another limit concerns the moderate or high socioeconomic
background of the children participating in this study. It is well known that children coming from
middle class families tend to benefit from more exposure to dialogic reading and language
stimulation (Frijters et al., 2000; Justice et al., 2003; Leseman & de Jong, 1998; Purcell-Gates,
1996), which might have positively affected the children’s performances. Future research may
thus examine how children from lower socioeconomic backgrounds perform on such tasks.
Also, to bolster the results and confirm the developmental pattern of inferential
comprehension, it would be of great interest to collect data with a longitudinal study design, with
follow-up on inferential abilities throughout the preschool years. Finally, due to inferential
comprehension being an ability frequently affected in children with a diagnosis of SLI, autism
spectrum disorder, or other language difficulties or disorders (Adams et al., 2009; Bishop, 1997;
Botting & Adams, 2005; Desmarais et al., 2013; Dodwell & Bavin, 2008; Ryder, Leinonen, &
Schulz, 2008; Skarakis-Doyle & Dempsey, 2008), comparison studies with these populations
96
during those early years of development would be helpful in providing a better and more
complete description of this ability.
Conclusion
By including a larger age range of children, this study confirms and expands previous
findings. Namely, the results show that there is an early emergence of inferential comprehension,
as young as 3 years of age and that inferential abilities then increase gradually until 6.
Consequently, the period that spans 3 to 6 years of age appears to be critical for the emergence
and development of this ability. Moreover, interactive media such the use of the iPad to measure
inferential comprehension appeared promising. Finally, the current study has implications to both
educators and speech-language pathologists working with children in special education who
could include stimulation of the different inferences types in classroom and therapy activities.
Furthermore, describing inferential ability in terms of quality of responses allows tracking of the
improvement of children’s ability in an in-depth fashion, a relevant consideration for
intervention. In summary, this study contributes to knowledge of the early developmental steps of
inferential comprehension, a crucial ability to later reading comprehension success.
97
Acknowledgment
This research was supported by a grant to Chantal Desmarais from the Social Sciences
and Humanities Research Council of Canada and received ethical approval from the Institut de
réadaptation en défience physique de Québec (IRDPQ). We are grateful to the children, parents,
daycares, and schools who participated in the study as well as the research assistants who helped
with the data collection and coding. We also acknowledge the statistical expertise of Jean
Leblond in the design and analysis for this study.
98
Appendix A. Four categories of quality of responses, examples to the question “What does
Pinson want?” and scores.
Quality of response continuum
Low High
Category D – Inadequate
or off topic
C - Low
contingency
B - Incomplete A - Expected
Definition
Unrelated to the
question or no
response
Contains
information that is
not relevant to the
question
Potential
justification; may be
imprecise or
incomplete
Most frequent
and contingent
response
Examples He wants to go
in the water. He wants to eat.
He wants to find his
parents.
He wants to find
a shelter.
Score 0 1 2 3
99
Appendix B. Examples of responses obtained by each age group to the "Attempts to solve the
problem", questions 7 (in prediction) and 8 (in evaluation).
3-4 years old 4-5 years old 5-6 years old
Is Pinson will be ok in the
skunk’s house
No I don’t think so No
Why? Because it’s too
small for him
Because it’s too
hot
Because it will
smell bad
Is it a good shelter for Pinson? No No No
Why? Because it’s too
small for him
Because it’s too
hot … and it will
smell bad too.
Because the skunk
smells bad.
100
References
Adams, C., Clarke, E., & Haynes, R. (2009). Inference and sentence comprehension in children
with specific or pragmatic language impairments. International Journal of Language &
Communication Disorders, 44(3), 301-318.
Barnes, M. A., Dennis, M., & Haefele-Kalvaitis, J. (1996). The effects of knowledge availability
and knowledge accessibility on coherence and elaborative inferencing in children from six
to fifteen years of age. Journal of Experimental Child Psychology, 61(3), 216-241.
Bianco, M., & Coda, M. (2002). La compréhension en quelques points... In M. Bianco, M. Coda,
& D. Gourgue (Eds.), La compréhension (pp. 93-97). Grenoble: Éditions de la Cigale.
Bishop, D. V. M. (1997). Uncommon understanding : development and disorders of language
comprehension in children. Hove, UK: Psychology Press.
Bishop, D. V. M., & Adams, C. (1990). A prospective study of the relationship between specific
language impairment, phonological disorders and reading retardation. Journal of Child
Psychology and Psychiatry, 31, 1027-1050.
Blanc, N. (2009). La compréhension de contes présentés oralement en classes de CP et CE1 :
quelle utilisation des dimensions situationnelles ? L'Année Psychologique, 109(04), 607-
628.
Blanc, N. (2010). La compréhension des contes entre 5 et 7 ans: Quelle représentation des
informations émotionnelles? [The comprehension of the tales between 5 and 7 year-olds:
Which representation of emotional information?]. Canadian Journal of Experimental
Psychology/Revue canadienne de psychologie expérimentale, 64(4), 256-265.
Blank, M., Rose, S. A., & Berlin, L. J. (2003). Preschool Language Assessment Instrument –
Second Edition. Austin, TX: Pro-Ed.
Blaye, A., & Chevalier, N. (2011). The role of goal representation in preschoolers' flexibility and
inhibition. Journal of Experimental Child Psychology, 108(3), 469-483.
Botting, N., & Adams, C. (2005). Semantic and inferencing abilities in children with
communication disorders. International Journal of Language & Communication
Disorders, 40(1), 49-66.
Bowyer-Crane, C., & Snowling, M. J. (2005). Assessing children's inference generation: what do
tests of reading comprehension measure? The British Journal of Educational Psychology,
75(2), 189-201.
Burris, S. E., & Brown, D. D. (2014). When all children comprehend: increasing the external
validity of narrative comprehension development research. Frontiers in Psycholgy, 5,
168.
Cain, K., & Oakhill, J. V. (1999). Inference making ability and its relation to comprehension
failure in young children. Reading and Writing, 11(5-6), 489-503.
Cain, K., Oakhill, J. V., & Elbro, C. (2003). The ability to learn new word meanings from context
by school-age children with and without language comprehension difficulties. Journal of
Child Language, 30(3), 681-694.
Catts, H. W. (2009). The narrow view of reading promotes a broad view of comprehension.
Language, Speech and Hearing Services in Schools, 40, 178-183.
Chevalier, N., & Blaye, A. (2006). Le développement de la flexibilité cognitive chez l'enfant
préscolaire: Enjeux théoriques. [The development of cognitive flexibility in preschoolers:
Theoretical issues.] (Version 4) France: Armand Colin.
Chevalier, N., & Blaye, A. (2008). Cognitive flexibility in preschoolers: the role of representation
activation and maintenance. Developmental Science, 11(3), 339-353.
101
Crais, E. R., & Chapman, R. S. (1987). Story Recall and Inferencing Skills in
Language/Learning-Disabled and Nondisabled Children. Journal of Speech and Hearing
Disorders, 52(1), 50-55.
Currie, N. K., & Cain, K. (2015). Children’s inference generation: The role of vocabulary and
working memory. Journal of Experimental Child Psychology, 137, 57-75.
Desmarais, C., Archambault, M.-C., Filiatrault-Veilleux, P., & Tarte, G. (2012). La
compréhension d’inférences : comparaison des habiletés d’enfants de quatre et de
cinq ans en lecture partagée. Revue des sciences de l'éducation, 38(3), p. 555-578.
doi:10.7202/1022712
Desmarais, C., Nadeau, L., Trudeau, N., Filiatrault-Veilleux, P., & Maxes-Fournier, C. (2013).
Intervention for improving comprehension in 4-6 year old children with specific language
impairment: practicing inferencing is a good thing. Clinical Linguistics & Phonetics,
27(6-7), 540-552.
Dodwell, K., & Bavin, E. L. (2008). Children with specific language impairment : an
investigation of their narratives and memory. International Journal of Language &
Communication Disorders, 43(2), 201-218.
Dunn, L., Thériault-Whalen, C. M., & Dunn, M. (1993). Échelle de vocabulaire en images
Peabody. Adaptation française du Peabody picture vocabulary test-revised. Toronto, ON:
Pearson Psychcorp.
Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D. J., & Pethick, S. J. (1994). Variability
in early communicative development. Monographs of the Society for Research in Child
Development, 59(5), 1-173; discussion 174-185.
Filiatrault-Veilleux, P., Bouchard, C., Trudeau, N., & Desmarais, C. (2015). Inferential
comprehension of 3-6 year olds within the context of story grammar: a scoping review.
International Journal of Language & Communication Disorders, 50(6), 737-749.
Filiatrault-Veilleux, P., Desmarais, C., Bouchard, C., Trudeau, N., & Leblond, J. (in Press).
Conception et qualités psychométriques d’un outil d’évaluation de la compréhension
d'inférences en contexte de récit chez les enfants âgés de 3 à 6 ans. Revue Canadienne
d’Orthophonie et d’Audiologie.
Filiatrault-Veilleux, P., Tarte, G., & Desmarais, C. (2015). Inferential Comprehension in a
Dialogic Reading Task: Comparing 5-year-old Children with Specific Language
Impairment to Typically Developing Peers. In G. M. Martinez (Ed.), Recent Advances in
Language and Communication (pp. pp. 119-133). New York: Nova Science Publishers,
Inc.
Florit, E., Roch, M., & Levorato, M. C. (2011). Listening text comprehension of explicit and
implicit information in preschoolers: The role of verbal and inferential skills. Discourse
Processes, 48(2), 119-138.
Florit, E., Roch, M., & Levorato, M. C. (2014). Listening text comprehension in preschoolers: a
longitudinal study on the role of semantic components. Reading and Writing, 27(5), 793-
817.
Ford, J. A., & Milosky, L. M. (2003). Inferring emotional reactions in social situations:
differences in children with language impairment. Journal of Speech, Language &
Hearing Research, 46(1), 21-30.
Ford, J. A., & Milosky, L. M. (2008). Inference generation during discourse and its relation to
social competence: an online investigation of abilities of children with and without
language impairment. Journal of Speech, Language & Hearing Research, 51(2), 367-380.
102
Frijters, J. C., Barron, R. W., & Brunello, M. (2000). Direct and mediated influences of home
literacy and literacy interest on prereaders’ oral vocabulary and early written language
skill. Journal of Educational Psychology, 92, 466–477.
Frye, D., Zelazo, P. D., & Palfai, T. (1995). Theory of mind and rule-based reasoning. Cognitive
Development, 10(4), 483-527.
Gineste, M.-D., & Le Ny, J. F. (2002). Psychologie cognitive du langage : de la reconnaissance
à la compréhension. Paris: Dunod.
Girolametto, L., Weitzman, E., Lefebvre, P., & Greenberg, J. (2007). The Effects of In-Service
Education to Promote Emergent Literacy in Child Care Centers: A Feasibility Study.
Language, Speech, and Hearing Services in Schools, 38(1), 72-83.
Gnepp, J., & Gould, M. E. (1985). The Development of Personalized Inferences: Understanding
Other People's Emotional Reactions in Light of Their Prior Experiences. Child
Development, 56(6), 1455-1464.
Graesser, A. C., Singer, M., & Trabasso, T. (1994). Constructing inferences during narrative text
comprehension. Psychological Review, 101(3), 371-395.
Justice, L. M., Chow, S. M., Capellini, C., Flanigan, K., & Colton, S. (2003). Emergent literacy
intervention for vulnerable preschoolers: relative effects of two approaches. American
Journal of Speech-Language Pathology, 12(3), 320-332.
Justice, L. M., & Kaderavek, J. N. (2004). Embedded-explicit emergent literacy intervention I:
Background and description of approach. Language, speech, and hearing services in
schools, 35(3), 201-211.
Kendeou, P., Bohn-Gettler, C., White, M. J., & van den Broek, P. (2008). Children's inference
generation across different media. Journal of Research in Reading, 31(3), 259-272.
Kendeou, P., van den Broek, P., White, M. J., & Lynch, J. S. (2009). Predicting Reading
Comprehension in Early Elementary School: The Independent Contributions of Oral
Language and Decoding Skills. Journal of Educational Psychology, 101(4), 765-778.
Kintsch, W., & Kintsch, E. (2005). Comprehension. In S. G. P. S. A. S. (Eds.) (Ed.), Children's
reading comprehension and assessment (pp. 71-92). Mahwah, NJ: Erlbaum.
Lefebvre, P., Bruneau, J., & Desmarais, C. (2012). Analyse conceptuelle de la compréhension
inférentielle en petite enfance à partir d’une recension des modèles théoriques. Revue des
sciences de l'éducation, 38(3), p. 533-553.
Leseman, P. P., & de Jong, P. F. (1998). Home literacy: Opportunity, instruction, cooperation and
social-emotional quality predicting early reading achievement. Reading Research
Quarterly, 33, 294–318.
Lever, R., & Senechal, M. (2011). Discussing stories: on how a dialogic reading intervention
improves kindergartners' oral narrative construction. Journal of Experimental Child
Psychology, 108(1), 1-24.
Lynch, J. S., & van den Broek, P. (2007). Understanding the glue of narrative structure:
Children's on- and off-line inferences about characters' goals. Cognitive Development,
22(3), 323-340.
Makdissi, H., & Boisclair, A. (2006). Interactive reading: A context for expanding the expression
of causal relations in preschoolers. Written Language and Literacy, 9(2), 177-211.
Mandler, J. M., & Johnson, N. S. (1977). Remembrance of Things Parsed: Story Structure and
Recall. Cognitive Psychology, 9(1), 119; -151.
Morrow, L. M. (1985). Retelling Stories: A Strategy for Improving Young Children's
Comprehension, Concept of Story Structure, and Oral Language Complexity. The
Elementary School Journal, 85(5), 647-661.
103
Oakhill, J., & Cain, K. (2000). Children's Difficulties in Text Comprehension: Assessing Causal
Issues. Journal of Deaf Studies and Deaf Education, 5(1), 51-59.
Paul, R. (2000). Putting things in context - Literal and discourse approaches to comprehension
assessment. Seminars in Speech and Language, 21(3), 247-255.
Potocki, A., Ecalle, J., & Magnan, A. (2012). Narrative Comprehension Skills in 5-Year-Old
Children: Correlational Analysis and Comprehender Profiles. The Journal of Educational
Research, 106(1), 14-26.
Price, L. H., van Kleeck, A., & Huberty, C. J. (2009). Talk during Book Sharing between Parents
and Preschool Children: A Comparison between Storybook and Expository Book
Conditions. Reading Research Quarterly, 44(2), 171-194.
Purcell-Gates, V. (1996). Stories, coupons, and the TV Guide: Relationships between home
literacy experiences and emergent literacy knowledge. Reading Research Quarterly,, 31,
406–428.
Reed, H. C., Hurks, P. P. M., Kirschner, P. A., & Jolles, J. (2015). Preschoolers’ Causal
Reasoning During Shared Picture Book Storytelling: A Cross-Case Comparison
Descriptive Study. Journal of Research in Childhood Education, 29(3), 367-389.
Richards, J. C., & Anderson, N. A. (2003). How Do You Know? A strategy to help emergent
readers make inferences. The Reading Teacher, 57(3), 290-293.
Ryder, N., Leinonen, E., & Schulz, J. (2008). Cognitive Approach to Assessing Pragmatic
Language Comprehension in Children with Specific Language Impairment. International
Journal of Language & Communication Disorders, 43(4), 427-447.
Schmidt, C. R., & Paris, S. G. (1978). Operativity and reversibility in children's understanding of
pictorial sequences. Child Development, 49(4), 1219-1222.
Serpell, R., Baker, L., & Sonnenschein, S. (2005). Becoming literate in the city : The Baltimore
Early Childhood Project. Cambridge, UK: Cambridge University Press.
Skarakis-Doyle, E., & Dempsey, L. (2008). The detection and monitoring of comprehension
errors by preschool children with and without language impairment. Journal of Speech
Language and Hearing Research, 51(5), 1227-1243.
Spackman, M. P., Fujiki, M., & Brinton, B. (2006). Understanding emotions in context: the
effects of language impairment on children's ability to infer emotional reactions.
International Journal of Language & Communication Disorders, 41(2), 173-188.
Stein, N. L., & Glenn, C. G. (1979). An analysis of story comprehension in elementary school
children. (Vol. Discourse processing: Multidisciplinary perspectives). Norwood, NJ:
Ablex: R. Freedle (Ed.).
Tompkins, V., Guo, Y., & Justice, L. (2013). Inference generation, story comprehension, and
language skills in the preschool years. Reading and Writing, 26(3), 403-429.
Trabasso, T., & Nickels, M. (1992). The development of goal plans of action in the narration of a
picture story. Discourse Processes, 15(3), 249-275.
Trabasso, T., & Wiley, J. (2005). Goal plans of action and inferences during comprehension of
narratives. Discourse Processes, 39, 129-164.
van den Broek, P., Kendeou, P., Kremer, K., Lynch, J. S., Butler, J., White, M. J., & Lorch, E. P.
(2005). Assessment of comprehension abilities in young children. In S. Stahl & S. Paris
(eds.), Children’s Reading Comprehension and Assessment, (pp.107-130). Mahwah, NJ:
Erlbaum.
van den Broek, P., Tzeng, Y., Risden, K., Trabasso, T., & Basche, P. (2001). Inferential
questioning: Effects on comprehension of narrative texts as a function of grade and
timing. Journal of Educational Psychology, 93(3), 521-529.
104
van Kleeck, A. (2008). Providing Preschool Foundations for Later Reading Comprehension: The
Importance of and Ideas for Targeting Inferencing in Storybook-Sharing Interventions.
Psychology in the Schools, 45(7), 627-643.
van Kleeck, A., Vander Woude, J., & Hammett, L. (2006). Fostering literal and inferential
language skills in Head Start preschoolers with language impairment using scripted book-
sharing discussions. American Journal of Speech-Language Pathology, 15(1), 85-95.
Verhallen, M. J. A. J., Bus, A. G., & de Jong, M. T. (2006). The Promise of Multimedia Stories
for Kindergarten Children at Risk. Journal of Educational Psychology, 98(2), 410-419.
Wellman, H. M., Cross, D., & Watson, J. (2001). Meta-Analysis of Theory-of-Mind
Development: The Truth about False Belief. Child Development, 72(3), 655-684.
Wenner, J. A. (2004). Preschoolers' comprehension of goal structure in narratives. Memory,
12(2), 193-202.
Westby, C., & Robinson, L. (2014). A Developmental Perspective for Promoting Theory of
Mind. Topics in Language Disorders, 34(4).
105
Chapitre 5. Autres résultats et apports à la compréhension
inférentielle
À la lumière de l’ensemble des données analysées dans la présente thèse, certaines
observations ont été soulevées dont les résultats n’ont pas fait l’objet d’articles. Le présent
chapitre présente donc les résultats obtenus de ces observations fortuites, qui n’ont pas été
rapportées dans l’article 3, soit au chapitre 4 de la thèse. Ainsi, ce chapitre vise à fournir d’autres
apports ou pistes de réflexion quant au développement de la compréhension des inférences
causales chez les jeunes enfants âgés de 3 à 6 ans. Plus spécifiquement, les objectifs abordés dans
le présent chapitre sont de : 1) vérifier la présence d’un effet de genre quant aux résultats obtenus
par les enfants des différents groupes; et 2) vérifier l’effet de la structure répétée de l’histoire sur
la compréhension des tentatives de résolution du problème.
5.1 Effet de genre
La distinction entre les performances des garçons et celles les filles quant au
développement des habiletés langagières a souvent fasciné les chercheurs. En effet, il est
désormais bien documenté qu’en bas âge, les filles démontrent une acquisition langagière plus
précoce et plus rapide comparativement aux garçons (Bouchard, Trudeau, Sutton, Boudreault, &
Deneault, 2009; Fenson et al., 1994; Hyde & Linn, 1988). Selon ces auteurs, le développement
lexical et les habiletés morphosyntaxiques seraient particulièrement sensibles à cette distinction
entre les genres. Toutefois, la supériorité langagière des filles s’est atténuée jusqu’à disparaître à
partir de l’âge de 27 mois dans l’étude de Bouchard et al. (2009). Les garçons ont donc rattrapé
leurs paires filles quant aux sphères langagières évaluées pendant la période de la petite enfance
dans cette étude (Bouchard et al., 2009).
En ce qui concerne la compréhension des inférences, cette distinction des performances
selon le genre pourrait concerner un type d’inférence en particulier, soit la réponse interne des
personnages. En effet, il est reconnu que les filles pourraient produire davantage de mots
émotionnels et évaluer de manière plus marquée la valence émotionnelle des mots que les
garçons (Neshat-Doost, Taghavi, Moradi, Yule, & Dalgleish, 1998; Syssau & Monnier, 2009).
106
Une hypothèse explicative issue de ces études pourrait provenir de la nature même du contenu
des interactions entre parents et enfants. Les parents pourraient être portés à discuter davantage
des émotions avec les filles qu’avec les garçons (par ex. Dunn, Bretherton, & Munn, 1987;
Fivush, Brotman, Buckner, & Goodman, 2000).
Ainsi, dans cette lignée, Blanc (2010) s’est intéressée à cette différenciation garçon/fille
quant à la compréhension des inférences. Dans son étude, elle a obtenu que les inférences de type
causalité et réponse interne des personnages étaient mieux réussies par les filles que les garçons
âgés de 5 à 7 ans. Les performances quant aux autres types d’inférences, soit de types espace,
temps et personnages étaient cependant comparables. Ainsi, à notre connaissance, Blanc (2010)
est l’une des seules chercheuses à s’être intéressée à cette distinction de genre quant à la
compréhension inférentielle des jeunes enfants.
Afin de continuer d’alimenter les réflexions à ce sujet et de tenter de reproduire les
résultats précédemment obtenus, les performances des filles ont été comparées à celles des
garçons dans la présente thèse. De manière plus précise, ces performances ont été comparées à
l’aide d’un test t indépendant, tant pour le score total de compréhension inférentielle tous âges
confondus, par groupes d’âge (à 3-4 ans, 4-5 ans et 5-6 ans) et par types d’inférences.
Les statistiques descriptives relatives au score total de compréhension inférentielle
obtenues par les garçons et les filles à la tâche Un abri pour Pinson sont présentées dans la
Figure 8. Les analyses de test t indépendant effectuées n’ont révélé aucun effet de genre pour le
score total de compréhension inférentielle obtenu (t(119) = -0.52, p=0.61); pour les trois groupes
d’enfants (3-4 ans : t(28) = -0.45, p=0.66; 4-5 ans : t(35) = 0.06, p =0.95; 5-6 ans : t(52) = -0.32
5, p=0,75) et pour tous les types d’inférences évaluées (Réponse interne : t(119) = -0.25, p=0,81;
But : t(119) = 0.05, p =0,96; Tentative de résolution du problème : t(119) = -0.32, p=0,75;
Prédiction : t(119)= -0.10, p =0,92; et Solution : t(119) = -0.89, p =0,38). Il est à noter qu’un seul
résultat s’est approché du seuil de signification, c’est-à-dire celui obtenu pour la variable
« Problème de l’histoire » (t(119) = -1.99, p=0,05), où les filles ont obtenu des scores légèrement
plus élevés (M = 72,04, ET = 17,35) que les garçons (M = 65,59, ET = 18,07).
107
Figure 8. Comparaison des performances obtenues par les filles et les garçons quant au score
inférentiel total à la tâche Un abri pour Pinson.
Ainsi, dans le cadre de la présente étude, les garçons ont obtenu des résultats similaires
aux filles. À la lumière des travaux antérieurs menés par Blanc (2010) et autres études portant sur
les mots à caractère émotionnel (Dunn et al., 1987; Fivush et al., 2000; Neshat-Doost et al., 1998;
Syssau & Monnier, 2009), il aurait été attendu d’obtenir un résultat favorable quant aux réponses
internes des personnages. À notre connaissance, peu de chercheurs ayant étudié les habiletés de
compréhension inférentielle des enfants se sont intéressés à cette différence de genre. Cette
observation découle fort probablement qu’actuellement, peu de données laissent présager la
présence d’une telle distinction. Comme mentionnées précédemment, dans une étude portant sur
le développement langagier, les distinctions garçons/filles, si elles sont présentes, disparaissent
tout au plus vers l’âge de 27 mois (Bouchard et al., 2009). Ainsi, en contexte d’évaluation des
habiletés inférentielles, pour la tranche d’âge d’intérêt de 3 à 6 ans, les données ne nous
permettent pas de conclure à une différence de performance entre les garçons et les filles. Dans
cette thèse, les garçons ont été aussi performants que les filles à inférer les liens causaux entre les
éléments de l’histoire en contexte de lecture partagée.
0
10
20
30
40
50
60
70
80
90
100
3:5 ans des milieux de
garde
4:5 ans des milieux de
garde
6 ans de la maternelle
Po
urc
enta
ge
(%)
Filles
Garçons
3-4 ans 4-5 ans 5-6 ans
108
5.2 Effet de de la structure répétitive de l’histoire sur la compréhension des tentatives de
résolution du problème.
Dans un récit, les différentes tentatives de résolution du problème prennent souvent une
place importante dans les histoires pour enfants. Bien qu’elles ne consistent pas en un élément
aussi central que le but, le caractère prévisible de leur échec est souvent bien connu par les
enfants. Elles leur permettent de générer des hypothèses plausibles tout au long de l’histoire afin
de résoudre le problème du personnage. Ainsi, intuitivement, les orthophonistes emploient
régulièrement des livres à caractère prévisible en thérapies orthophoniques. En effet, il est
raisonnable de penser qu’exposer les enfants à des tentatives de résolution du problème qui se
suivent et se ressemblent les aidera à anticiper plus facilement la suite des événements dans une
histoire.
Dans le cadre de la présente thèse, la curiosité des chercheuses les a poussées à vouloir
vérifier l’effet de la structure répétitive des tentatives de résolution du problème sur la
compréhension de l’histoire par les enfants. Cet aspect de la compréhension des inférences n’a, à
notre connaissance, fait l’objet d’aucune étude à ce jour. Pourtant, cette information serait
pertinente d’un point de vue clinique afin de valider cette pratique intuitive des orthophonistes.
Dans l’histoire Un abri pour Pinson, Pinson, le personnage principal, tente de résoudre
son problème à trois reprises lors de sa rencontre avec trois animaux, mais sans y arriver (voir les
Tableaux 17 et 18). Pour chaque tentative de résolution du problème, une question inférentielle
est posée en prédiction, avant que Pinson n’entre dans l’abri de l’animal (pages 9, 11 et 13 de
l’histoire). Une question est ensuite posée en évaluation, lorsque Pinson tente d’entrer dans l’abri
de l’animal (pages 10, 12 et 14 de l’histoire). S’il y a présence d’un effet de la structure répétitive
sur la compréhension des tentatives, il est attendu que le score de compréhension des enfants
s’améliore d’une tentative à l’autre, soit d’un animal à l’autre. Dans la présente thèse, cet effet de
la structure répétitive sur la compréhension a été évalué, tant pour les prédictions que pour les
évaluations des tentatives de résolution du problème.
Tout d’abord, relativement aux trois questions ciblant les prédictions des tentatives de
résolution du problème (Tableau 17), les résultats obtenus par les trois groupes d’enfants (3-4
ans, 4-5 ans et 5-6 ans) ont été comparés à l’aide d’analyses de variances ANOVA.
109
Tableau 17. Les trois questions inférentielles et réponses attendues portant sur les prédictions des
tentatives de résolution du problème.
Animal Question inférentielle Réponse attendue
1re tentative : la moufette
(page 9 de l’histoire)
Est-ce que Pinson va être bien dans
la maison de la moufette? Pourquoi?
Non, parce qu’elle va sentir
mauvais.
2e tentative : le porc-épic
(page 11 de l’histoire)
Est-ce que Pinson va être bien dans
la maison du porc-épic? Pourquoi?
Non, parce qu’il peut piquer
Pinson.
3e tentative : le castor
(page 13 de l’histoire)
Est-ce que Pinson va être bien dans
la maison du castor? Pourquoi?
Non, parce que Pinson ne
sait pas nager.
Les moyennes et écarts-types obtenus par les trois groupes d’enfants sont présentés dans
la Figure 9. En ce qui concerne les deux plus jeunes groupes d’enfants, aucune différence
significative n’a été obtenue entre les trois tentatives (3-4 ans : F (2,118) = 0.128, p = 0.880; 4-5
ans : F (2,118) = 2.227, p = 0.107). Toutefois, pour ce qui est du groupe d’enfants 5-6 ans, une
différence significative est présente (F (2,118) = 4.743, p = 0.01). L’analyse post hoc de
Bonferroni a démontré que les résultats obtenus pour la 2e tentative (le porc-épic) et la 3e
tentative (le castor) étaient significativement plus élevés que ceux obtenus pour la 1re tentative (la
moufette) (p < 0.05).
110
*p0,05
Figure 9. Moyennes et écarts-types obtenus par les trois groupes d’enfants pour les prédictions
des trois tentatives de résolution du problème.
Ces résultats indiquent donc que seuls les enfants âgés de 5-6 ans ont pu bénéficier de la
structure répétitive de l'histoire sur leur compréhension des inférences portant sur la prédiction
des tentatives de résolution du problème. Cependant, comme ces données s’appuient sur des
réponses à une seule question par tentative, il est important de soulever leur grande distribution
autour des moyennes. En effet, les écarts-types obtenus sont très élevés, diminuant du coup la
validité de ces résultats. Ainsi, il est suggéré de réaliser d’autres analyses auprès d’un plus grand
échantillon d’enfants.
Ensuite, la même procédure a été appliquée pour les trois questions ciblant les évaluations
des tentatives de résolution du problème (Tableau 18). Les résultats obtenus par les trois groupes
d’enfants (3-4 ans, 4-5 ans et 5-6 ans) ont été comparés à l’aide d’analyses de variance univariées
(ANOVAs).
0
10
20
30
40
50
60
70
80
90
100
3-4 ans 4-5 ans 5-6 ans
Pourc
enta
ge
(%)
Moufette
Porc-épic
Castor *
*
111
Tableau 18. Les trois questions inférentielles et réponses attendues portant sur les évaluations des
tentatives de résolution du problème.
Animal Question inférentielle Réponse attendue
1re tentative : la moufette
(page 10 de l’histoire)
Est-ce un bon abri pour Pinson?
Pourquoi?
Non, parce qu’elle sent
mauvais.
2e tentative : le porc-épic
(page 12 de l’histoire)
Est-ce un bon abri pour Pinson?
Pourquoi?
Non, parce qu’il pique
Pinson.
3e tentative : le castor
(page 14 de l’histoire)
Est-ce un bon abri pour Pinson?
Pourquoi?
Non, parce que Pinson ne
sait pas nager.
Les moyennes et écarts-types obtenus par les trois groupes d’enfants sont présentés dans
la Figure 10. Pour le plus jeune groupe d’enfants âgés de 3-4 ans, aucune différence significative
selon les trois tentatives n’a été obtenue (F (2,118) = 2.741, p = 0.070). Pour les deux autres
groupes d’enfants âgés de 4-5 ans et 5-6 ans, une différence significative était présente (4-5 ans :
F (2,118) = 6.845, p = 0.002; 5-6 ans : F (2,118) = 8.118, p < 0.001). Pour ces deux groupes,
l'analyse post hoc de Bonferroni a démontré que les résultats obtenus à la 2e tentative (porc-épic)
étaient significativement plus élevés que ceux obtenus à la 1re tentative (moufette). De plus, pour
le groupe d’enfants âgés de 5-6 ans, la 2e tentative (le porc-épic) a également été
significativement mieux réussi que la 3e tentative (le castor) (p < 0.01).
Ces résultats penchent également vers un effet bénéfique de la structure répétitive de
l'histoire sur la compréhension ces types d’inférences pour les deux groupes d’enfants plus âgés.
Également, ces résultats indiquent que le porc-épic a été la tentative la plus facile à inférer pour
les enfants âgés de 5-6 ans comparativement aux deux autres animaux. Encore une fois, les
écarts-types obtenus sont très élevés et la conduction d’analyses auprès d’un plus grand
échantillon d’enfants est suggérée.
112
*p0,05
Figure 10. Moyennes et écarts-types obtenus par les trois groupes d’enfants pour les évaluations
des trois tentatives de résolution du problème.
Ces résultats exploratoires ouvrent la voie vers de nouvelles réflexions quant à l’effet des
tentatives de résolution prévisibles sur la compréhension d’un récit par l’enfant. Cet effet s’est
avéré particulièrement présent chez les enfants âgés de 5 et 6 ans, et ce, tant en prédiction qu’en
évaluation. Pour les enfants âgés de 4-5 ans, cet effet a été démontré significatif seulement lors
des évaluations, celles-ci étant plus faciles à inférer que les prédictions. Ainsi, ces résultats
démontrent qu’à partir de cet âge, les enfants commencent à pouvoir ajuster la qualité de leurs
réponses produites en cours de récit. Il pourrait donc exister un effet bénéfique de la structure
répétitive du récit sur la compréhension des tentatives de résolution du problème. Il apparaît ainsi
pertinent d’inclure ce type d’histoire à caractère prévisible lors des interventions auprès des
0
10
20
30
40
50
60
70
80
90
100
3-4 ans 4-5 ans 5-6 ans
Pourc
enta
ge
(%)
Moufette
Porc-épic
Castor *
* *
113
enfants, surtout pour ceux âgés de 5 et 6 ans. Cependant, comme mentionné précédemment, ces
résultats sont exploratoires et davantage d’études sont nécessaires afin de les confirmer.
Or, pour les plus jeunes enfants âgés de 3-4 ans, cette structure répétitive ne semble pas
aussi bénéfique. Quelques hypothèses peuvent être soulevées. D’abord, il a été précédemment
démontré dans cette thèse que les tentatives de résolution du problème étaient difficiles à cet âge,
puisqu’elles seraient maîtrisées que vers l’âge de 5 ou 6 ans. Ainsi, ces types d’inférences sont
peut-être tout simplement trop difficiles à cet âge, ou encore, les enfants ne possèdent peut-être
pas encore les connaissances générales nécessaires leur permettant de réaliser ces inférences.
Une autre hypothèse, en lien davantage avec les habiletés cognitives des enfants en
développement, pourrait être le plus faible niveau de flexibilité cognitive à cet âge (Blaye &
Chevalier, 2011; Chevalier & Blaye, 2006; Chevalier & Blaye, 2008). Plus précisément, la
flexibilité est impliquée lorsque l’enfant doit sélectionner de manière adaptative une nouvelle
réponse en fonction d’une variation de la situation (Chevalier & Blaye, 2006). Ainsi, cette plus
faible flexibilité cognitive pourrait empêcher les enfants de modifier leurs réponses au cours du
récit. En effet, lors de l’administration de la tâche, il a été observé que les jeunes enfants de 3 et 4
ans répondaient plus souvent la même réponse, et ce, à toutes les tentatives de résolution
auxquelles Pinson était confronté (par ex. : « non, parce qu’il fait trop chaud » pour les trois
animaux). Ces pistes de réflexion seraient par conséquent très intéressantes à investiguer
davantage en lien avec les habiletés cognitives des jeunes enfants.
Finalement, les données actuellement obtenues dans Un abri pour Pinson laissent
présager que la tentative du porc-épic était plus facile à inférer pour les enfants de tous âges que
celles des deux autres animaux (moufette et castor). Plusieurs facteurs pourraient être liés à ces
résultats. Notons la présence d’indices visuels sur les illustrations qui auraient pu influencer la
compréhension des inférences, les connaissances générales des enfants en lien avec les trois
animaux, l’ordre de présentation des tentatives et finalement, la sévérité dans la codification de la
qualité des réponses accordées aux enfants. Lors de futures études, un paramètre important à
contrôler pour vérifier l’effet de la structure répétitive sur la compréhension serait l’ordre de
présentation des animaux. En effet, si cet effet est présent, la performance des enfants devraient
s’améliorer, et ce, peu importe l’ordre dans lequel les animaux sont présentés. Encore une fois, la
conduction d’études supplémentaires en contrôlant ces différents facteurs serait appropriée.
114
Références
Blanc, N. (2010). La compréhension des contes entre 5 et 7 ans: Quelle représentation des
informations émotionnelles? [The comprehension of the tales between 5 and 7 year-olds:
Which representation of emotional information?]. Canadian Journal of Experimental
Psychology/Revue canadienne de psychologie expérimentale, 64(4), 256-265. doi:
10.1037/a0021283
Blaye, A., & Chevalier, N. (2011). The role of goal representation in preschoolers' flexibility and
inhibition. Journal of Experimental Child Psychology, 108(3), 469-483. doi:
10.1016/j.jecp.2010.09.006
Bouchard, C., Trudeau, N., Sutton, A. N. N., Boudreault, M.-C., & Deneault, J. (2009). Gender
differences in language development in French Canadian children between 8 and 30
months of age. Applied Psycholinguistics, 30(04), 685-707
Chevalier, N., & Blaye, A. (2006). Le développement de la flexibilité cognitive chez l'enfant
préscolaire: Enjeux théoriques. [The development of cognitive flexibility in preschoolers:
Theoretical issues.] (Version 4) [doi:10.4074/S0003503306004040]. France: Armand
Colin.
Chevalier, N., & Blaye, A. (2008). Cognitive flexibility in preschoolers: the role of representation
activation and maintenance. Developmental Science, 11(3), 339-353. doi: 10.1111/j.1467-
7687.2008.00679.x
Dunn, J., Bretherton, I., & Munn, P. (1987). Conversations about feeling states between mothers
and their young children. Developmental Psychology, 23(1), 132-139. doi: 10.1037/0012-
1649.23.1.132
Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D. J., & Pethick, S. J. (1994). Variability
in early communicative development. Monographs of the Society for Research in Child
Development, 59(5), 1-173; discussion 174-185.
Fivush, R., Brotman, M., Buckner, J., & Goodman, S. (2000). Gender Differences in Parent–
Child Emotion Narratives. Sex Roles, 42(3-4), 233-253. doi: 10.1023/A:1007091207068
Hyde, J. S., & Linn, M. C. (1988). Gender differences in verbal ability: A meta-analysis.
Psychological Bulletin, 104(1), 53-69.
Neshat-Doost, H. T., Taghavi, M. R., Moradi, A. R., Yule, W., & Dalgleish, T. (1998). Memory
for emotional trait adjectives in clinically depressed youth. Journal of Abnormal
Psychology, 107(4), 642-650.
Syssau, A., & Monnier, C. (2009). Children's emotional norms for 600 French words. Behavior
Research Methods, 41(1), 213-219. doi: 10.3758/BRM.41.1.213
115
Chapitre 6. Discussion générale
Ce dernier chapitre comporte d’abord un retour sur les principaux résultats obtenus dans le
cadre de cette thèse. Il vise à résumer l’ensemble des données recueillies et analysées afin de
dégager les faits saillants sur le développement de la compréhension des inférences causales chez
les enfants âgés de 3 à 6 ans. Par la suite, les forces et les limites des différentes études seront
présentées. Les retombées de cette thèse, tant en recherche qu’en clinique ainsi que les
perspectives futures seront ensuite discutées. Finalement, une courte conclusion clôturera la
présente thèse.
6.1 Synthèse des résultats
Cette thèse doctorale visait à décrire le développement de la compréhension des inférences
causales en contexte de récit chez les jeunes enfants âgés de 3 à 6 ans. Pour ce faire, trois
objectifs généraux étaient poursuivis via trois grandes étapes réalisées de manière séquentielle.
Le premier objectif était de dégager des écrits scientifiques les éléments soutenant une
proposition théorique de la façon dont se met en place la compréhension des inférences causales à
l’âge préscolaire (article 1, chapitre 2). Le deuxième objectif de cette thèse consistait à concevoir
un outil d’évaluation à partir des jalons développementaux mis de l’avant et à en évaluer les
qualités psychométriques (article 2, chapitre 3). Enfin, le troisième objectif était de valider de
façon empirique et préciser, le cas échéant, cette proposition théorique auprès d’enfants âgés de 3
à 6 ans (article 3, chapitre 4). Les résultats de cette démarche ont permis de mettre en évidence
une proposition scientifiquement appuyée d’une séquence de développement de la
compréhension des inférences causales chez le jeune enfant.
En premier lieu, afin de répondre au premier objectif général de cette thèse, une recension
des écrits de type « scoping review » a été réalisée, compte tenu de la nature du construit
théorique ciblé. Ce type novateur de recension des écrits permet effectivement de dégager des
thèmes et d’identifier les lacunes dans les écrits scientifiques publiés sur un sujet de recherche
examiné, selon des perspectives et domaines diversifiés (Arksey & O'Malley, 2005; Drager,
Reichle, & Pinkoski, 2010; Rumrill, Fitzgerald, & Merchant, 2010). Comme la compréhension
inférentielle est un construit complexe qui a fait l’objet d’études provenant de différents champs
116
d’expertise, tels que l’orthophonie, la psychologie et l’éducation, ce type de recension s’est avéré
un choix judicieux. Ainsi, à partir de la synthèse des écrits publiés sur le sujet, cette étude a
dégagé deux éléments clés.
D’abord, elle a fourni une première ébauche de séquence développementale de la
compréhension de six types d’inférences causales ciblant les éléments de la grammaire de récit
chez les enfants de 3 à 6 ans. Le résumé de cette séquence de développement est présenté dans le
Tableau 4 de la thèse. L’ensemble des études évaluées a confirmé que la compréhension des
inférences causales en contexte de récit est une habileté qui est bien présente à l’âge préscolaire,
bien avant l’entrée à l’école. Dès l’âge de 3 ans ou 4 ans, les enfants seraient capables de réaliser
plus facilement des inférences portant sur les réponses internes, le problème et le but du
personnage dans une histoire. Puis, ils développeraient graduellement l’habileté à inférer les
tentatives de résolution du problème, à faire des prédictions et à inférer la solution du récit vers
l’âge de 5 ou 6 ans.
Cependant, comme aucune étude n’avait étudié l’ensemble de ces six types d’inférences sur
l’étendue d’âge visée, la conduction d’études supplémentaires était recommandée afin de
confirmer et/ou préciser cette séquence développementale. Également, la très grande
hétérogénéité des cadres théoriques utilisés a été soulevée, menant à une interprétation prudente
des résultats obtenus. Ce constat venait appuyer une fois de plus que davantage de recherches
étaient nécessaires afin de préciser ce construit théorique.
Le deuxième élément clé a été de mettre en lumière les méthodes novatrices utilisées par
les différentes études pour mesurer cette habileté chez les jeunes enfants. Le résumé de ces
méthodes utilisées est présenté dans le Tableau 5 de la thèse. Les modalités suivantes se sont
avérées des choix adaptés à la réalité des petits qui n’ont pas encore appris à lire : la lecture
partagée qui combine un support visuel et auditif, les questions posées pendant la lecture pour
enrichir la qualité des relations causales exprimées et/ou comprises par l’enfant et la codification
des réponses des enfants en différentes catégories afin de mieux en discriminer les performances.
Néanmoins, l’absence d’uniformité dans les méthodes utilisées pour évaluer la compréhension
inférentielle des enfants a été soulevée. En effet, aucune étude n’a pu fournir ou identifier un outil
d’évaluation standardisé, pouvant être utilisé en clinique comme en recherche. Cet outil serait
pourtant essentiel afin de poursuivre les recherches dans le domaine. Ainsi, la présente thèse a
117
voulu répondre à ce besoin de connaissances, en concevant un nouvel outil d’évaluation de la
compréhension d’inférences qui tient compte des résultats obtenus dans la recension des écrits
(article 1).
Le deuxième objectif général de la thèse consistait à concevoir un outil d’évaluation de la
compréhension des six inférences causales impliquées en contexte de récit à partir des jalons
théoriques mis de l’avant dans la recension des écrits. Il visait également à en déterminer la
validité de contenu de même qu’à en mesurer les qualités psychométriques. L’outil Un abri pour
Pinson a été conçu sur iPad, sous forme de livre interactif, afin d’assurer une uniformité dans la
procédure de passation. L’outil permet d’évaluer les six types d’inférences causales, en plus de
tenir compte des éléments méthodologiques importants soulevés précédemment (la lecture
partagée, les questions posées pendant la lecture et la codification des réponses des enfants sur un
continuum de qualité).
La première étape de validation via une consultation de six experts cliniques a servi à
peaufiner l’outil d’évaluation, tant sur le plan du contenu que de sa fonctionnalité sur iPad, afin
qu’il soit optimal avant son utilisation auprès de jeunes enfants. Par la suite, l’outil a été
administré à 121 enfants âgés de 3 à 6 ans afin d’en déterminer la validité de construit et la
validité concurrente de même que sa fidélité interjuge et test-retest. Les résultats ont démontré
qu’Un abri pour Pinson présentait de bonnes qualités psychométriques. En effet, il a démontré
une excellente structure représentationnelle de l’échantillon avec un indice KMO élevé de 0.87
(p<0.01).
L’analyse factorielle en composantes principales a permis de dégager quatre types
d’inférences provenant de sources de variance presque indépendantes, soit la tentative de
résolution du problème, le but, l’événement déclencheur et la solution. La réponse interne et la
prédiction se sont avérées quant à elles dépendantes et associées à d’autres types d’inférences.
Également, pour ce qui est de la validité concurrente, une corrélation élevée de 0.77 (p<0.01) a
été obtenue avec un outil expérimental de la compréhension des inférences, alors qu’il a obtenu
une corrélation plus faible, quoique modérée de 0.43 (p<0.01) avec le vocabulaire réceptif. Ces
résultats étaient conformes aux attentes par rapport aux écrits dans le domaine. Sur le plan de la
fidélité, les valeurs des coefficients obtenus ont été élevées, soit de 0.83 et 0.99 (p<0.01) aux
analyses interjuges et de 0.95 (p<0.01) au test-retest. Ce nouvel outil d’évaluation s’avérait donc
118
un excellent outil pouvant préciser le développement de la compréhension inférentielle chez les
jeunes enfants en contexte de récit.
Enfin, le troisième objectif général de la thèse consistait à décrire le développement de la
compréhension des six types inférences causales chez 121 enfants âgés de 3 à 6 ans à l’aide de
l’outil d’évaluation créé sur iPad. Plus spécifiquement, cette étude visait à comparer les
performances des enfants quant au score total de compréhension inférentielle ainsi qu’aux six
différents sous-scores obtenus, soit les six types d’inférences causales ciblant les éléments de la
grammaire de récit. Elle avait également pour but de comparer les performances des enfants en
fonction de la qualité des réponses produites. Il a été confirmé que la compréhension inférentielle
apparaît tôt dans le développement de l’enfant, soit dès 3 ans, les scores affichant toutefois une
plus grande variabilité interindividuelle. Ensuite, elle s’est améliorée significativement avec
l’âge. Les différents types d’inférences, tels qu’avancés dans l’article 1, se mettent en place
graduellement selon une séquence développementale. La classification des réponses des enfants a
également permis de distinguer les groupes d’âge et de décrire l’évolution de la compréhension
inférentielle de manière qualitative.
Finalement, d’autres observations quant à l’effet de genre et l’effet de la structure répétitive
des tentatives de résolution du problème sur la compréhension ont été examinées. Pour ce qui est
des différences entre genre, elles ne se sont pas avérées concluantes pour la tranche d’âge
d’intérêt. Quant aux tentatives de résolution du problème, les enfants de 5 et 6 ans ont davantage
pu bénéficier de la structure répétitive de l’histoire sur leur compréhension. Pour leur part, les
enfants de 3 et 4 ans n’ont pas démontré d’amélioration d’une tentative à l’autre, ce qui pourrait
être en lien avec le niveau de compréhension, les connaissances générales ou la plus faible
flexibilité cognitive de ces derniers. L’ensemble de ces données a permis de dégager le tableau
synthèse suivant qui vise à mieux décrire comment la compréhension des inférences causales se
met en place chez les enfants de 3 à 6 ans, en contexte de récit (Tableau 19). Le Tableau 19 met
en relation les éléments développementaux clés avancés dans la recension des écrits (article 1) et
les nuances apportées par l’étude empirique menée dans la présente thèse (article 3) grâce à
l’outil Un abri pour Pinson (article 2).
119
Tableau 19. Synthèse du développement de la compréhension des inférences causales en contexte de récit chez les enfants de 3 à 6 ans
Âge Éléments saillants de la séquence développementale
Résultats du « scoping review » Ajouts et nuances apportés par l’étude empirique
Habiletés
inférentielles
3-4 ans
Apparaissent très précocement et peuvent être
mesurées en contexte de lecture partagée.
Plus grande variabilité interindividuelle à cet âge, caractéristiques
personnelles et environnementales à investiguer.
4-5 ans Amélioration avec l’augmentation en âge. Amélioration significative et diminution de la variabilité
interindividuelle.
5-6 ans Amélioration avec l’augmentation en âge, et
continuent de s’améliorer au-delà de cet âge.
Amélioration significative et diminution importante de la
variabilité interindividuelle.
À 5-6 ans, les enfants n’ont pas obtenu des scores parfaits.
Types
d’inférences
3-4 ans
Compréhension de la relation entre une situation
et une émotion.
Absence de types d’inférences mieux comprises.
4-5 ans Types d’inférences : réponse interne, problème
et but de l’histoire.
Types d’inférences : réponse interne, problème et prédiction.
Réponse interne : 1re relation causale comprise.
Problème : action concrète et observable.
But : sensible à la structure de l'histoire et aux
nombres de relations causales du récit.
Réponse interne : comprise précocement.
Problème : action concrète et observable. Comprise avant le but.
Prédiction : apparaît précocement (avant 5 ou 6 ans), surtout celles
de type cause à effet.
5-6 ans Types d’inférences : tentative de résolution du
problème, prédiction et solution.
Types d’inférences : tentative de résolution du problème, but et
solution.
Tentative de résolution du problème : moins
saillante pour les enfants.
Prédiction : évaluée plus tardivement.
Solution : comprise tardivement, intégration des
différents éléments de la chaine causale exigée.
Tentative de résolution du problème : nécessite des/accès à des
connaissances plus spécifiques. La structure répétitive de l’histoire
est bénéfique à la compréhension.
But : exige une prise en considération de la perspective de l’autre.
Solution : comprise tardivement.
Qualité des
réponses
aux questions
inférentielles
3-4 ans Absence de données développementales.
Proportion élevée de réponses « inadéquates » et « ambigües ».
4-5 ans Absence de données développementales. Proportion élevée de réponses « ambigües » et augmentation
graduelle des réponses « attendues ».
5-6 ans Absence de données développementales. Proportion élevée de réponses « attendues ».
120
6.2 Forces et limites
Plusieurs forces et limites peuvent être dégagées des trois études réalisées dans le cadre de la
présente thèse, les principales ayant déjà été abordées dans les chapitres 2 à 5. Comme mentionné
précédemment, la grande force de ce projet doctoral réside en son intérêt envers une habileté peu
étudiée à ce jour chez les jeunes enfants malgré son lien indéniable avec le succès en lecture plus
tard (Cain, Oakhill, & Elbro, 2003; Kintsch & Kintsch, 2005; Trabasso & Wiley, 2005; van den
Broek et al., 2005; Zwaan & Singer, 2003) et, par ricochet, avec la réussite scolaire (Conseil
canadien sur l’apprentissage, 2007; Conseil des statistiques canadiennes de l’éducation, 2009).
Relativement au premier article de la thèse, l’utilisation d’un type de recension des écrits
novateur s’est avérée être une grande force dans cette démarche doctorale. En effet, cette
recension des écrits de type « scoping review » a servi à dégager des constats théoriques, même
s’il s’agissait d’un construit complexe qui a avait été examiné selon différents angles et
perspectives. Ce type de recension des écrits mérite certainement une place plus importante dans
les futurs écrits scientifiques.
La force principale du deuxième article consiste en la démarche rigoureuse employée pour la
conception de l’outil Un abri pour Pinson. L’élaboration et la validation de cet outil se sont
basées sur les recommandations des « Standards for Educational and Psychological Testing » des
associations de l’American Educational Research Association, l’American Psychological
Association et le National Council of Measurement in Education (AERA, APA, & NCME,
1999). Cette démarche a ainsi permis la conception d’un outil robuste tant sur le plan scientifique
que clinique. Elle a donc certainement contribué à assurer les qualités psychométriques de
l’outil. Enfin, il ne faut pas négliger l’aspect novateur de l’utilisation des nouvelles
technologiques dans ce projet de recherche, qui bénéficie d’un intérêt grandissant en éducation.
Les tâches d’évaluation sur tablette informatique sont prometteuses en favorisant l’intérêt,
l’attention et le plaisir de l’enfant lors de leur administration (Brandone et al., 2008; Friend &
Keplinger, 2008). Ce projet doctoral pourrait inspirer d’autres études à emboîter le pas dans cette
avenue novatrice.
Finalement, la force du troisième article réside en l’étendue d’âge ciblée (3 à 6 ans), en
lien avec une description détaillée de la compréhension des inférences causales. En effet, elle a
été examinée sous plusieurs aspects, incluant des résultats quantitatifs, grâce au score total et
121
sous-scores par types d’inférences, en y intégrant des résultats qualitatifs grâce à la qualité des
réponses fournies. Il ne faut également pas oublier l’effet de genre et l’effet de la structure
répétée sur la compréhension de l’histoire qui ont été de surcroît abordés dans le chapitre 5.
L’ensemble de ces données contribue à mieux situer la compréhension des inférences en contexte
de langage oral, de son émergence à son rôle comme passerelle vers l’apprentissage de la lecture
à l’âge scolaire.
En contrepartie, quelques limites générales de la présente thèse méritent d’être soulevées.
En premier lieu, la nature complexe du construit théorique évalué limite la possibilité d’établir
des conclusions robustes quant à la compréhension inférentielle. Il apparaît ainsi très important
de poursuivre la réalisation d’études dans le domaine. Une meilleure conceptualisation et
opérationnalisation de ce construit permettra sans aucun doute aux chercheurs et aux cliniciens
œuvrant auprès des enfants d’âge préscolaire de mieux cerner la compréhension inférentielle, et
par conséquent, d’intervenir plus efficacement sur son développement.
Pour ce qui est de l’étude réalisée auprès des 121 enfants, l’absence de mesure de
variables confondantes potentiellement contributoires à la compréhension inférentielle des
enfants est une limite n’ayant pas été mentionnée auparavant. À titre d’exemple, bien que le
vocabulaire réceptif ait été mesuré avec l’ÉVIP (Dunn, Thériault-Whalen, & Dunn, 1993), le
vocabulaire spécifiquement employé ainsi que la compréhension littérale des énoncés présentés
dans Un abri pour Pinson n’ont pas été préalablement mesurés chez les enfants. Par conséquent,
il n’est pas exclu que les difficultés de compréhension des enfants puissent être expliquées, du
moins en partie, par leurs habiletés langagières de plus bas niveau. Également, il faut ajouter que
la mesure du niveau de compréhension inférentielle des enfants a été évaluée par les réponses
expressives produites. Ainsi, les habiletés langagières expressives des enfants ont sans doute pu
influencer à certains égards le niveau de qualité des réponses produites par ces derniers. Enfin,
certains facteurs environnementaux, tels que l’exposition à la lecture partagée dans le milieu
éducatif ou à la maison, n’ont pas été pris en considération dans ces études.
Finalement, une dernière limitation qui a nécessité de longues réflexions tout au long de
cette démarche doctorale concerne la formation des groupes d’enfants. Initialement, le projet de
recherche visait à comparer la performance de quatre groupes de 30 enfants, divisés par âge
chronologique : 3, 4, 5 et 6 ans. Comme le but de l’étude était de documenter cette habileté avant
122
l’apprentissage de la lecture, tous les enfants de 6 ans ont été recrutés en maternelle. Toutefois,
un problème s’est présenté avec le groupe d’enfants âgés de 5 ans, dont certains fréquentaient
également la maternelle. En effet, lors de l’analyse des données, il est apparu n’y avoir qu’un seul
groupe d’enfants, âgés de 5 et 6 ans, qui fréquentent la maternelle. La situation était similaire
pour les enfants de 5 ans fréquentant le milieu de garde. Ils partageaient davantage de similitudes
avec les enfants âgés de 4 ans, eux-aussi issus des milieux de garde. Ainsi, pour ces
considérations développementales, l’équipe de recherche a décidé d’opter pour une division des
enfants en trois groupes, en fonction de l’âge des enfants tout en considérant le milieu éducatif
fréquenté (3-4 ans des milieux de garde; 4-5 ans des milieux de garde; 5-6 ans de la maternelle).
Ce changement de répartition des participants au sein des groupes explique pourquoi les groupes
d’enfants ne sont pas égaux en nombre. Heureusement, la puissance statistique n’a pas été
affectée par ce changement.
6.3 Retombées
La présente thèse contribue à l’avancement des connaissances en recherche et en clinique,
en fournissant des premiers barèmes développementaux de la compréhension des inférences
causales. Les articles 1 et 3 de la présente thèse permettent de dégager une proposition théorique
de sa période de développement appuyée scientifiquement. Elle porte sur une grande étendue
d’âge d’enfants (3 à 6 ans) en ce qui a trait aux 6 types d’inférences causales impliquées dans la
compréhension d’un récit. Une autre contribution significative de ce projet doctoral est la
proposition d’un outil d’évaluation de lecture partagée de la compréhension des inférences
causales en contexte de récit (article 2), Un abri pour Pinson. Les aspects de ces retombées qui
en découlent en recherche et en clinique sont discutés.
Tout d’abord, en recherche, les études avaient davantage documenté la compréhension des
inférences de la période scolaire jusqu’à l’âge adulte (Ackerman, Silver, & Glickman, 1990;
Bishop & Adams, 1992; Letts & Leinonen, 2001; van den Broek, Tzeng, Risden, Trabasso, &
Basche, 2001). Puisqu’elle occupe le premier rang des compétences à développer pour favoriser
la réussite en lecture à l’âge scolaire (Cain, Oakhill, & Elbro, 2003; Kintsch & Kintsch, 2005;
Trabasso & Wiley, 2005; van den Broek et al., 2005; Zwaan & Singer, 2003), il était d’une
importance capitale de fournir des données de recherche plus précises sur la période de son
123
émergence et de son développement, qui avait jusqu’alors été peu étudiée. Ce projet doctoral
contribue donc à détailler le développement de ce prérequis crucial à l’apprentissage de la lecture.
De plus, la grande hétérogénéité des méthodes et outils utilisés pour évaluer cette habileté
chez les enfants avait été soulevée dans les écrits scientifiques (Blanc, 2009, 2010; Filiatrault-
Veilleux et al., 2015; van den Broek et al., 2005). Ainsi, en fournissant un outil d’évaluation
valide et fidèle, la présente thèse pourra contribuer à uniformiser les méthodes employées en
recherche pour évaluer cette habileté auprès des enfants de 3 à 6 ans. Un abri pour Pinson pourra
effectivement être utilisé en recherche pour documenter le niveau d’habiletés inférentielles des
enfants, en plus de servir de mesure de progrès des interventions orthophoniques effectuées en
recherche (par ex., en pré et post intervention).
En contexte clinique, les répercussions sont également largement positives. La pratique
fondée sur les données probantes est au cœur des préoccupations actuelles dans le domaine
orthophonique. Jusqu’à aujourd’hui, il était ardu pour les orthophonistes de stimuler ou de
promouvoir le développement de cette habileté fréquemment atteinte chez les enfants parce
qu’aucun repère développemental n’était disponible. Ainsi, ces premiers barèmes
développementaux viseront à fournir aux orthophonistes des balises sur lesquelles appuyer leurs
interventions.
Également, une importante préoccupation en clinique actuellement concerne la rareté des
outils d’évaluation orthophoniques disponibles en contexte franco-québécois (Bouchard,
Fitzpatrick, & Olds, 2009). Qui plus est, aucun outil d’évaluation ne permettait actuellement de
documenter les habiletés inférentielles des enfants d’âge préscolaire. Les retombées cliniques
sont donc majeures pour l’évaluation orthophonique. Le dépistage des enfants à risque ou
éprouvant des difficultés de compréhension inférentielle se verra ainsi facilité pour les
orthophonistes.
Finalement, les retombées en intervention sont importantes et toucheront non seulement le
domaine orthophonique, mais également l’éducation au préscolaire. En effet, le dépistage et
l'intervention précoces sont essentiels dans un but de prévenir les difficultés de lecture des jeunes
enfants. En agissant plus efficacement et précocement sur le développement de la compréhension
des inférences, il est possible de croire en une amélioration des chances de réussite scolaire des
124
enfants. L’apport de ces connaissances sur cet important prérequis à la réussite en lecture pourrait
permettre l’amélioration des services orthophoniques et pédagogiques offerts aux enfants.
6.4 Perspectives futures
Au terme du présent projet doctoral et en tenant compte des études antérieures sur le sujet,
des interrogations demeurent et pourraient faire l’objet de recherches futures. Tout d’abord, sur le
plan développemental, comment les enfants évoluent-ils dans le temps par rapport à eux-mêmes
quant à leurs habiletés de compréhension inférentielle ? Les données recueillies dans la présente
thèse répondent à certaines interrogations à l’aide d’un devis expérimental transversal. Or, selon
une perspective développementale, beaucoup reste à investiguer quant à l’évolution des habiletés
langagières des jeunes enfants dans le temps, à l’aide de devis de recherche longitudinal.
Ensuite, considérant ce qui a été précédemment soulevé concernant les difficultés de
compréhension inférentielle chez les enfants, il est légitime de se questionner sur les trajectoires
développementales des enfants présentant des difficultés langagières. En effet, la première étape
consistait à décrire comment cette habileté se met en place chez les enfants tout-venant.
Toutefois, qu’en est-il des enfants aux prises avec un retard ou un trouble de langage, ou un
trouble du spectre autistique ? Dans ce même ordre d’idée, qu’en est-il des enfants provenant de
milieux socioéconomiques plus faibles, provenant de milieux vulnérables ou des enfants négligés
? Une poursuite de la recherche en utilisant des groupes comparatifs d’enfants s’avère en ce sens
fondamentale. Qui plus est, ceci permettrait également de documenter un autre type de validité de
l’outil Un abri pour Pinson, soit la validité discriminante. En fonction des données obtenues, Un
abri pour Pinson pourrait ultimement s’ajouter aux outils pouvant contribuer au dépistage des
enfants en difficultés langagières.
De surcroît, il reste beaucoup à faire pour mieux décrire les liens qu’entretient la
compréhension des inférences avec les autres habiletés cognitives et langagières. Quels sont les
liens existant entre les habiletés de discours narratif, les compétences en théorique de l’esprit, la
mémoire de travail et les fonctions exécutives tel que la flexibilité cognitive ? La réalisation
d’études qui prendraient en considération ces autres variables potentiellement contributoires ou
liées à la compréhension des inférences serait essentielle.
125
Finalement, rappelons qu’Un abri pour Pinson a été conçu et validé en contexte franco-
québécois. Quelles en seraient les qualités psychométriques après traduction dans une autre
langue et administration à d’autres populations ? Également, les résultats d’une version papier
seraient-ils comparables à la version informatique ? Serait-il possible de concevoir d’autres
versions d’outils d’évaluation et/ou du matériel d’intervention orthophonique à partir de
l’application informatique créée ? Il s’agit en effet d’une innovation qui pourrait faire l’objet de
recherches futures.
Ainsi donc, force est de constater que ce projet doctoral ouvre sur une multitude de nouvelles
pistes à explorer afin d’approfondir les connaissances quant au développement des habiletés
inférentielles des enfants pendant cette période cruciale qu’est la petite enfance.
6.5 Conclusion
Ce projet doctoral visait à mieux documenter le développement des habiletés de
compréhension des inférences causales en contexte de récit chez les jeunes enfants âgés de 3 à 6
ans. Il y est arrivé par le biais d’une démarche scientifique rigoureuse comportant trois grandes
étapes, soit la réalisation d’une recension des écrits, suivie de la conception de l’outil
d’évaluation Un abri pour Pinson et conclue par une étude expérimentale auprès de 121 enfants
au développement typique âgés de 3 à 6 ans. L’habileté à comprendre des inférences est une
habileté complexe qui émerge très tôt et qui se construit graduellement, de l’âge de 3 à 6 ans.
Cette période s’avère critique pour le développement de cette habileté qui joue un rôle clé dans la
réussite ultérieure en lecture. À partir des données développementales recueillies dans la présente
thèse, il importe donc de mettre en place des stratégies efficaces pour en promouvoir le
développement. Les activités de stimulation permettront de jouer un important rôle en prévention
auprès des jeunes enfants et ainsi, ultimement pouvoir influencer la suite de leur parcours
académique. La réussite scolaire des enfants est au cœur de nos préoccupations dans notre
société. Il est donc primordial de continuer de réaliser des recherches dans ce domaine afin de
constamment viser l’amélioration des services orthophoniques et pédagogiques offerts aux
enfants.
126
Bibliographie
Ackerman, B. P., Silver, D., & Glickman, I. (1990). Concept availability in the causal inferences
of children and adults. Child Development, 61(1), 230-246.
Adams, C., Clarke, E., & Haynes, R. (2009). Inference and sentence comprehension in children
with specific or pragmatic language impairments. International Journal of Language &
Communication Disorders, 44(3), 301-318.
Adams, C., Cooke, R., Crutchley, A., Hesketh, A. and Reeves, D. 2001. Assessment of
Comprehension and Expression 6–11 (ACE 6–11), Windsor, NFER-Nelson.
AERA, APA, & NCME. (1999). Standards for educational and psychological testing.
Washington, DC: American Educational Research Assocation.
Amery, H., & Cartwright, S. (2001). Petit cochon s'est perdu. Usborne. Les contes de la ferme.
Arksey, H., & O'Malley, L. (2005). Scoping studies: towards a methodological framework.
International Journal of Social Research Methodology, 8(1), 19-32. doi:
10.1080/1364557032000119616
Barnes, M. A., Dennis, M., & Haefele-Kalvaitis, J. (1996). The effects of knowledge availability
and knowledge accessibility on coherence and elaborative inferencing in children from six
to fifteen years of age. Journal of Experimental Child Psychology, 61(3), 216-241. doi:
10.1006/jecp.1996.0015
Bianco, M., Bressoux, P., Doyen, A.L., Lambert, E., Lima, L., Pellenq, C. and Zorman, M.
(2010). Early training of oral comprehension and phonological skills at preschool: the
results of a 3 years longitudinal study. Scientific Studies of Reading, 14(3), 211-246.
doi:10888430903117518.
Bianco, M., & Coda, M. (2002). La compréhension en quelques points... In M. Bianco, M. Coda,
& D. Gourgue (Eds.), La compréhension (pp. 93-97). Grenoble: Éditions de la Cigale.
Bianco, M., Pellenq, C., Lambert, E., Bressoux, P., Lima, L. and Doyen, A.-L. (2012), Impact of
early code-skill and oral-comprehension training on reading achievement in first grade.
Journal of Research in Reading, 35: 427–455. doi: 10.1111/j.1467-9817.2010.01479.x
Bishop, D. V., & Adams, C. (1992). Comprehension problems in children with specific language
impairment: literal and inferential meaning. Journal of Speech And Hearing Research,
35(1), 119-129.
Bishop, D. V. M. (1997). Uncommon understanding : development and disorders of language
comprehension in children. Hove, UK: Psychology Press.
Bishop, D. V. M., & Adams, C. (1990). A prospective study of the relationship between specific
language impairment, phonological disorders and reading retardation. Journal of Child
Psychology and Psychiatry, 31, 1027-1050.
Blanc, N. (2009). La compréhension de contes présentés oralement en classes de CP et CE1 :
quelle utilisation des dimensions situationnelles ? L'Année Psychologique, 109(04), 607-
628. doi:10.4074/S0003503309004023
Blanc, N. (2010). La compréhension des contes entre 5 et 7 ans: Quelle représentation des
informations émotionnelles? [The comprehension of the tales between 5 and 7 year-olds:
Which representation of emotional information?]. Canadian Journal of Experimental
Psychology/Revue canadienne de psychologie expérimentale, 64(4), 256-265. doi:
10.1037/a0021283
Blank, M., Rose, S. A., & Berlin, L. J. (2003). Preschool Language Assessment Instrument –
Second Edition. Austin, TX: Pro-Ed.
127
Blaye, A., & Chevalier, N. (2011). The role of goal representation in preschoolers' flexibility and
inhibition. Journal of Experimental Child Psychology, 108(3), 469-483. doi:
10.1016/j.jecp.2010.09.006
Botting, N., & Adams, C. (2005). Semantic and inferencing abilities in children with
communication disorders. International Journal of Language & Communication
Disorders, 40(1), 49-66.
Bouchard, M., Fitzpatrick, E., & Olds, J. (2009). Analyse psychométrique d’outils d’évaluation
utilises auprès d’enfants francophones. Revue canadienne d’orthophonie et d’audiologie,
33(3), 129-139.
Bouchard, C., Trudeau, N., Sutton, A. N. N., Boudreault, M.-C., & Deneault, J. (2009). Gender
differences in language development in French Canadian children between 8 and 30
months of age. Applied Psycholinguistics, 30(04), 685-707
Bowyer-Crane, C., & Snowling, M. J. (2005). Assessing children's inference generation: what do
tests of reading comprehension measure? The British Journal of Educational Psychology,
75(Pt 2), 189-201.
Brandone, A. C., Michnick Golinkoff, R., & Hirsh-Pasek, K. (2008). Feasibility of Computer-
Administered Language Assessment. Perspectives on School-based Issues, 9, 57-65. doi:
doi:10.1044/sbi9.2.57
Burris, S. E., & Brown, D. D. (2014). When all children comprehend: increasing the external
validity of narrative comprehension development research. Frontiers in Psycholgy, 5,
168.
Cain, K., & Oakhill, J. (2007). Children's comprehension problems in oral and written language
: a cognitive perspective. New York: Guilford Press.
Cain, K., & Oakhill, J. V. (1999). Inference making ability and its relation to comprehension
failure in young children. Reading and Writing, 11(5-6), 489-503. doi:
10.1023/a:1008084120205
Cain, K., Oakhill, J. V., & Elbro, C. (2003). The ability to learn new word meanings from context
by school-age children with and without language comprehension difficulties. Journal of
Child Language, 30(3), 681-694.
Catts, H. W. (2009). The narrow view of reading promotes a broad view of comprehension.
Language, Speech and Hearing Services in Schools, 40, 178-183.
Chevalier, N., & Blaye, A. (2006). Le développement de la flexibilité cognitive chez l'enfant
préscolaire: Enjeux théoriques. [The development of cognitive flexibility in preschoolers:
Theoretical issues.] (Version 4) [doi:10.4074/S0003503306004040]. France: Armand
Colin.
Chevalier, N., & Blaye, A. (2008). Cognitive flexibility in preschoolers: the role of representation
activation and maintenance. Developmental Science, 11(3), 339-353. doi: 10.1111/j.1467-
7687.2008.00679.x
Conseil des statistiques canadiennes de l’éducation (2009). Facteurs clés de réussite en littératie
parmi les populations d'âge scolaire. Toronto, Ontario : Conseil des statistiques
canadiennes de l'éducation.
Conseil Canadien Sur L’apprentissage. (2007). État de l’apprentissage au Canada : pas le temps
de s’illusionner. Ottawa, Ontario : Conseil Canadien Sur L’apprentissage.
Crais, E. R., & Chapman, R. S. (1987). Story Recall and Inferencing Skills in
Language/Learning-Disabled and Nondisabled Children. Journal of Speech and Hearing
Disorders, 52(1), 50-55.
128
Currie, N. K., & Cain, K. (2015). Children's inference generation: The role of vocabulary and
working memory. Journal of Experimental Child Psychology, 137, 57-75.
doi:10.1016/j.jecp.2015.03.005
Das Gupta, P., & Bryant, P. E. (1989). Young children's causal inferences. Child Development,
60(5), 1138-1146.
Deconti, K. A., & Dickerson, D. J. (1994). Preschool children's understanding of the situational
determinants of others' emotions. Cognition & Emotion, 8(5), 453-472. doi:
10.1080/02699939408408952
Desmarais, C., Archambault, M.-C., Filiatrault-Veilleux, P., & Tarte, G. (2012). La
compréhension d’inférences : comparaison des habiletés d’enfants de quatre et de
cinq ans en lecture partagée. Revue des sciences de l'éducation, 38(3), p. 555-578. doi:
10.7202/1022712ar
Desmarais, C., Nadeau, L., Trudeau, N., Filiatrault-Veilleux, P., & Maxes-Fournier, C. (2013a).
Effet d’une intervention visant à améliorer la compréhension inférentielle des enfants de
4 à 6 ans ayant une dysphasie : une collaboration orthophonistes-chercheurs. Glossa,
Glossa 113 Spécial XIIIèmes Rencontres d’orthophonie, 45-62.
Desmarais, C., Nadeau, L., Trudeau, N., Filiatrault-Veilleux, P., & Maxes-Fournier, C. (2013b).
Intervention for improving comprehension in 4-6 year old children with specific language
impairment: practicing inferencing is a good thing. Clinical Linguistics & Phonetics,
27(6-7), 540-552. doi: 10.3109/02699206.2013.791880
Dodwell, K., & Bavin, E. L. (2008). Children with specific language impairment : an
investigation of their narratives and memory. International Journal of Language &
Communication Disorders, 43(2), 201-218.
Drager, K. D., Reichle, J., & Pinkoski, C. (2010). Synthesized speech output and children: a
scoping review. American Journal of Speech-Language Pathology, 19(3), 259-273. doi:
10.1044/1058-0360(2010/09-0024)
Dunn, J., Bretherton, I., & Munn, P. (1987). Conversations about feeling states between mothers
and their young children. Developmental Psychology, 23(1), 132-139. doi: 10.1037/0012-
1649.23.1.132
Dunn, L., Thériault-Whalen, C. M., & Dunn, M. (1993). Échelle de vocabulaire en images
Peabody. Adaptation française du Peabody picture vocabulary test-revised. Toronto, ON:
Pearson Psychcorp.
Dupin De Saint-André, M. (2008). Pistes pour travailler les inférences. Québec français, 150, 66-
67.
Duval, S. et Bouchard, C. (2013). Soutenir la préparation à l’école et à la vie des enfants issus de
milieux défavorisés ou en difficulté. Québec, Canada : Ministère de la Famille et des
Ainés, 169 pages. Disponible en
ligne : https://www.mfa.gouv.qc.ca/fr/publication/Documents/preparation-ecole.pdf
Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D. J., & Pethick, S. J. (1994). Variability
in early communicative development. Monographs of the Society for Research in Child
Development, 59(5), 1-173; discussion 174-185.
Ferreiro, E., & Taberoski, A. (1982). Literacy before schooling. Portsmouth, NH: Heinemann
Educational Books Inc.
Filiatrault-Veilleux, P., Bouchard, C., Trudeau, N. and Desmarais, C. (2015). Inferential
comprehension of 3-6 year-olds within the context of story grammar: a scoping review.
International Journal of Language & Communication Disorders, 50(6), 737-49. doi:
10.1111/1460-6984.12175
129
Filiatrault-Veilleux, P., Desmarais, C., Bouchard, C., Trudeau, N. et Leblond, J. (Sous presse).
Conception et qualités psychométriques d’un outil d’évaluation de la compréhension
d’inférences en contexte de récit chez des enfants âgés de 3 à 6 ans. Revue Canadienne
d’Orthophonie et d’Audiologie.
Filiatrault-Veilleux, P., Tarte, G., & Desmarais, C. (2015). Inferential Comprehension in a
Dialogic Reading Task: Comparing 5-year-old Children with Specific Language
Impairment to Typically Developing Peers. In G. M. Martinez (Ed.), Recent Advances in
Language and Communication (pp. pp. 119-133). New York: Nova Science Publishers,
Inc.
Fivush, R., Brotman, M., Buckner, J., & Goodman, S. (2000). Gender Differences in Parent–
Child Emotion Narratives. Sex Roles, 42(3-4), 233-253. doi: 10.1023/A:1007091207068
Florit, E., Roch, M., Altoè, G., & Levorato, M. C. (2009). Listening comprehension in
preschoolers: the role of memory. British Journal of Developmental Psychology, 27(Pt 4),
935-951.
Florit, E., Roch, M., & Levorato, M. C. (2011). Listening text comprehension of explicit and
implicit information in preschoolers: The role of verbal and inferential skills. Discourse
Processes, 48(2), 119-138. doi: 10.1080/0163853x.2010.494244
Florit, E., Roch, M., & Levorato, M. C. (2014). Listening text comprehension in preschoolers: a
longitudinal study on the role of semantic components. Reading and Writing, 27(5), 793-
817.
Ford, J. A., & Milosky, L. M. (2003). Inferring emotional reactions in social situations:
differences in children with language impairment. Journal of Speech, Language &
Hearing Research, 46(1), 21-30.
Ford, J. A., & Milosky, L. M. (2008). Inference generation during discourse and its relation to
social competence: an online investigation of abilities of children with and without
language impairment. Journal of Speech, Language & Hearing Research, 51(2), 367-380.
Friend, M., & Keplinger, M. (2008). Reliability and validity of the computerized comprehension
task (CCT): data from American English and Mexican Spanish infants. Journal of Child
Language, 35(1), 77-98.
Frijters, J. C., Barron, R. W., & Brunello, M. (2000). Direct and mediated influences of home
literacy and literacy interest on prereaders’ oral vocabulary and early written language
skill. Journal of Educational Psychology, 92, 466–477.
Frye, D., Zelazo, P. D., & Palfai, T. (1995). Theory of mind and rule-based reasoning. Cognitive
Development, 10(4), 483-527. doi: http://dx.doi.org/10.1016/0885-2014(95)90024-1
Gibbons, J., Anderson, D. R., Smith, R., Field, D. E., & Fisher, C. (1986). Young Children's
Recall and Reconstruction of Audio and Audiovisual Narratives. Child Development,
57(4), 1014-1023.
Gineste, M.-D., & Le Ny, J. F. (2002). Psychologie cognitive du langage : de la reconnaissance
à la compréhension. Paris: Dunod.
Girolametto, L., Weitzman, E., Lefebvre, P., & Greenberg, J. (2007). The Effects of In-Service
Education to Promote Emergent Literacy in Child Care Centers: A Feasibility Study.
Language, Speech, and Hearing Services in Schools, 38(1), 72-83. doi: 10.1044/0161-
1461(2007/007)
Gnepp, J., & Gould, M. E. (1985). The Development of Personalized Inferences: Understanding
Other People's Emotional Reactions in Light of Their Prior Experiences. Child
Development, 56(6), 1455-1464.
130
Graesser, A. C., Singer, M. and Trabasso, T. 1994. Constructing inferences during narrative text
comprehension. Psychological Review, 101, 371-95.
Gupta, P. D., & Bryant, P. E. (1989). Young Children's Causal Inferences. Child Development,
60(5), 1138-1146.
Hudson, J. A., & Slackman, E. A. (1990). Children's use of scripts in inferential text processing.
Discourse Processes, 13(4), 375-385. doi: 10.1080/01638539009544766
Hyde, J. S., & Linn, M. C. (1988). Gender differences in verbal ability: A meta-analysis.
Psychological Bulletin, 104(1), 53-69.
Isbell, R., Sobol, J., Lindauer, L., & Lowrance, A. (2004). The Effects of Storytelling and Story
Reading on the Oral Language Complexity and Story Comprehension of Young Children.
Early Childhood Education Journal, 32(3), 157-163. doi:
10.1023/B:ECEJ.0000048967.94189.a3
Joffe, V. L., Cain, K., & Maric, N. (2007). Comprehension problems in children with specific
language impairment: does mental imagery training help? International Journal of
Language & Communication Disorders, 42(6), 648-664. doi:
10.1080/13682820601084402
Johnson, D. D., & von Hoff Johnson, B. (1986). Highlighting Vocabulary in Inferential
Comprehension Instruction. Journal of Reading, 29(7), 622-625.
Justice, L. M., Chow, S. M., Capellini, C., Flanigan, K., & Colton, S. (2003). Emergent literacy
intervention for vulnerable preschoolers: relative effects of two approaches. American
Journal of Speech-Language Pathology, 12(3), 320-332.
Justice, L. M., & Kaderavek, J. N. (2004). Embedded-explicit emergent literacy intervention I:
Background and description of approach. Language, speech, and hearing services in
schools, 35(3), 201-211.
Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39, 401-415.
Kendeou, P., Bohn-Gettler, C., White, M. J., & van den Broek, P. (2008). Children's inference
generation across different media. Journal of Research in Reading, 31(3), 259-272. doi:
10.1111/j.1467-9817.2008.00370.x
Kendeou, P., van den Broek, P., White, M. J., & Lynch, J. S. (2009). Predicting Reading
Comprehension in Early Elementary School: The Independent Contributions of Oral
Language and Decoding Skills. Journal of Educational Psychology, 101(4), 765-778.
Kintsch, W., & Kintsch, E. (2005). Comprehension. In S. G. Paris & S. A. Stahl (Eds.), Current
issues on reading comprehension and assessment (pp. 71-92). Mahwah, NJ: Erlbaum.
Kim, J.-O., & Mueller, C. W. (1978). Introduction to Factor Analysis. Thousand Oaks, CA:
SAGE Publications, Inc.
Kontos, S., & Wells, W. (1986). What preschool children know about reading and how they learn
it. Young Children, 42(1), 58–66.
Laveault, D., & Grégoire, J. (2014). Introduction aux théories des tests en psychologie et en
sciences de l'éducation. Bruxelles: De Boeck.
Lefebvre, P., Bruneau, J., & Desmarais, C. (2012). Analyse conceptuelle de la compréhension
inférentielle en petite enfance à partir d’une recension des modèles théoriques. Revue des
sciences de l'éducation, 38(3), p. 533-553. doi:10.7202/1022711ar
Leseman, P. P., & de Jong, P. F. (1998). Home literacy: Opportunity, instruction, cooperation and
social-emotional quality predicting early reading achievement. Reading Research
Quarterly, 33, 294–318.
131
Letts, C., & Leinonen, E. (2001). Comprehension of inferential meaning in language-impaired
and language normal children. International Journal of Language & Communication
Disorders, 36(3), 307-328. doi: 10.1080/13682820110045829
Levorato, M. C. and Roch, M. 2007. Valutare la comprensione del testo orale: Il TOR 3–8
[Evaluation of listening comprehension: The TOR 3–8] Florence, Italy: Organizzazioni
Speciali.
Lever, R., & Senechal, M. (2011). Discussing stories: on how a dialogic reading intervention
improves kindergartners' oral narrative construction. Journal of Experimental Child
Psychology, 108(1), 1-24.
Lynch, J. S., & van den Broek, P. (2007). Understanding the glue of narrative structure:
Children's on- and off-line inferences about characters' goals. Cognitive Development,
22(3), 323-340. doi: 10.1016/j.cogdev.2007.02.002
Makdissi, H., & Boisclair, A. (2006). Interactive reading: A context for expanding the expression
of causal relations in preschoolers. Written Language and Literacy, 9(2), 177-211. doi:
10.1075/wll.9.2.02mak
Mandler, J. M., & Johnson, N. S. (1977). Remembrance of Things Parsed: Story Structure and
Recall. Cognitive Psychology, 9(1), 111-151. McKenzie, F. (2009). Un guide pour soutenir la réflexion et ... mieux faire grandir les tout-petits.
Montréal, Québec : Agence de la santé et des services sociaux de Montréal. Repéré à
http://www.csssdll.qc.ca/fileadmin/csss_dll/CSSS_DLL/Publications/Maturite_scolaire/Guide
_pratiques_0_5_ans.pdf. Ministère de l'Éducation, de l'Enseignement supérieur et de la Recherche (2015). Lecture à
l’école. Repéré à http://www.education.gouv.qc.ca/dossiers-thematiques/lecture-a-lecole/.
Morrow, L. M. (1985). Retelling Stories: A Strategy for Improving Young Children's
Comprehension, Concept of Story Structure, and Oral Language Complexity. The
Elementary School Journal, 85(5), 647-661.
Morrow, L. M. (1988). Young Children's Responses to One-to-One Story Readings in School
Settings. Reading Research Quarterly, 23(1), 89-107.
Murachver, T., Pipe, M.-E., Gordon, R., Owens, J. L. and Fivush, R. 1996. Do, Show, and Tell:
Children's Event Memories Acquired through Direct Experience, Observation, and
Stories. Child Development, 67, 3029-3044.
Neshat-Doost, H. T., Taghavi, M. R., Moradi, A. R., Yule, W., & Dalgleish, T. (1998). Memory
for emotional trait adjectives in clinically depressed youth. Journal of Abnormal
Psychology, 107(4), 642-650.
NICHD, Early Child Care research Network. (2005). Pathways to reading: the role of oral
language in the transition to reading. Developmental Psychology, 41(2), 428-442.
Oakhill, J., & Cain, K. (2000). Children's Difficulties in Text Comprehension: Assessing Causal
Issues. Journal of Deaf Studies and Deaf Education, 5(1), 51-59.
Paul, R. (2000). Putting things in context - Literal and discourse approaches to comprehension
assessment. Seminars in Speech and Language, 21(3), 247-255.
Potocki, A., Ecalle, J., & Magnan, A. (2012). Narrative Comprehension Skills in 5-Year-Old
Children: Correlational Analysis and Comprehender Profiles. The Journal of Educational
Research, 106(1), 14-26. doi: 10.1080/00220671.2012.667013
Price, L. H., van Kleeck, A., & Huberty, C. J. (2009). Talk during Book Sharing between Parents
and Preschool Children: A Comparison between Storybook and Expository Book
Conditions. Reading Research Quarterly, 44(2), 171-194.
132
Purcell-Gates, V. (1996). Stories, coupons, and the TV Guide: Relationships between home
literacy experiences and emergent literacy knowledge. Reading Research Quarterly, 31,
406–428.
Reed, H. C., Hurks, P. P. M., Kirschner, P. A., & Jolles, J. (2015). Preschoolers’ Causal
Reasoning During Shared Picture Book Storytelling: A Cross-Case Comparison
Descriptive Study. Journal of Research in Childhood Education, 29(3), 367-389.
Richards, J. C., & Anderson, N. A. (2003). How Do You Know? A strategy to help emergent
readers make inferences. The Reading Teacher, 57(3), 290-293.
Rumrill, P. D., Fitzgerald, S. M., & Merchant, W. R. (2010). Using scoping literature reviews as
a means of understanding and interpreting existing literature. Work, 35(3), 399-404. doi:
10.3233/WOR-2010-0998
Ryder, N., Leinonen, E., & Schulz, J. (2008). Cognitive Approach to Assessing Pragmatic
Language Comprehension in Children with Specific Language Impairment. International
Journal of Language & Communication Disorders, 43(4), 427-447.
Serpell, R., Baker, L., & Sonnenschein, S. (2005). Becoming literate in the city : The Baltimore
Early Childhood Project. Cambridge, UK: Cambridge University Press.
Shrout, P. E. (1998). Measurement reliability and agreement in psychiatry. Statistical Methods in
Medical Research, 7(3), 301-317.
Schmidt, C. R., & Paris, S. G. (1978). Operativity and reversibility in children's understanding of
pictorial sequences. Child Development, 49(4), 1219-1222.
Skarakis-Doyle, E., & Dempsey, L. (2008). The detection and monitoring of comprehension
errors by preschool children with and without language impairment. Journal of Speech
Language and Hearing Research, 51(5), 1227-1243. doi: 10.1044/1092-4388(2008/07-
0136)
Spackman, M. P., Fujiki, M., & Brinton, B. (2006). Understanding emotions in context: the
effects of language impairment on children's ability to infer emotional reactions.
International Journal of Language & Communication Disorders, 41(2), 173-188. doi:
10.1080/13682820500224091
Steffen, T., & Seney, M. (2008). Test-retest reliability and minimal detectable change on balance
and ambulation tests, the 36-item short-form health survey, and the unified Parkinson
disease rating scale in people with parkinsonism. Physical Therapy, 88(6), 733-746. doi:
10.2522/ptj.20070214
Stein, N. L., & Glenn, C. G. (1979). An analysis of story comprehension in elementary school
children. (Vol. Discourse processing: Multidisciplinary perspectives). Norwood, NJ:
Ablex: R. Freedle (Ed.).
Syssau, A., & Monnier, C. (2009). Children's emotional norms for 600 French words. Behavior
Research Methods, 41(1), 213-219. doi: 10.3758/BRM.41.1.213
Tabachnick, B. G., & Fidell, L. S. (2001). Using Multivariate Statistics (4th ed.). Boston: Allyn
and Bacon.
Tompkins, V., Guo, Y., & Justice, L. (2013). Inference generation, story comprehension, and
language skills in the preschool years. Reading and Writing, 26(3), 403-429.
Trabasso, T., & Nickels, M. (1992). The development of goal plans of action in the narration of a
picture story. Discourse Processes, 15(3), 249-275. doi: 10.1080/01638539209544812
Trabasso, T., Stein, N. L., Rodkin, P. C., Munger, C. P. and Baughn, C. R. (1992). Knowledge of
Goals and Plans in the On-line Narration of Events. Cognitive Development, 7, 133-170.
133
Trabasso, T., Suh, S., Payton, P. and Jain, R. 1995. Explanatory inferences and other strategies
during comprehension and their effect on recall. In R.F. Lorch (Ed.), Sources of
coherence in reading (pp. 219-239). Hillsdale, NJ : Lawrence Erlbaum Associates.
Trabasso, T., & Wiley, J. (2005). Goal plans of action and inferences during comprehension of
narratives. Discourse Processes, 39, 129-164.
van den Broek, P., Kendeou, P., Kremer, K., Lynch, J. S., Butler, J., White, M. J., & Lorch, E. P.
(2005). Assessment of comprehension abilities in young children. In S. Stahl & S. Paris
(eds.), Children’s Reading Comprehension and Assessment, (pp.107-130). Mahwah, NJ:
Erlbaum.
van den Broek, P., Tzeng, Y., Risden, K., Trabasso, T., & Basche, P. (2001). Inferential
questioning: Effects on comprehension of narrative texts as a function of grade and
timing. Journal of Educational Psychology, 93(3), 521-529. doi: 10.1037/0022-
0663.93.3.521
van Kleeck, A. (2008). Providing Preschool Foundations for Later Reading Comprehension: The
Importance of and Ideas for Targeting Inferencing in Storybook-Sharing Interventions.
Psychology in the Schools, 45(7), 627-643.
van Kleeck, A., Vander Woude, J., & Hammett, L. (2006). Fostering literal and inferential
language skills in Head Start preschoolers with language impairment using scripted book-
sharing discussions. American Journal of Speech-Language Pathology, 15(1), 85-95. doi:
10.1044/1058-0360(2006/009)
Verhallen, M. J. A. J., Bus, A. G., & de Jong, M. T. (2006). The Promise of Multimedia Stories
for Kindergarten Children at Risk. Journal of Educational Psychology, 98(2), 410-419.
Walt Disney (1993). Les 3 Petits Cochons et le Grand Méchant Loup. Disney Hachette Edition :
États-Unis.
Wechsler, D. (2002). Wechsler Preschool and Primary Scale of Intelligence™ - Third Edition
(WPPSI™ - III). San Antonio, TX: The Psychological Corporation.
Wellman, H. M., Cross, D., & Watson, J. (2001). Meta-Analysis of Theory-of-Mind
Development: The Truth about False Belief. Child Development, 72(3), 655-684.
Wenner, J. A. (2004). Preschoolers' comprehension of goal structure in narratives. Memory,
12(2), 193-202. doi: 10.1080/09658210244000478
Westby, C., & Robinson, L. (2014). A Developmental Perspective for Promoting Theory of
Mind. Topics in Language Disorders, 34(4).
Whitehurst, G. J., & Lonigan, C. J. (1998). Child development and emergent literacy. Child
Development, 69(3), 848-872.
Wongpakaran, N., Wongpakaran, T., Wedding, D., & Gwet, K. L. (2013). A comparison of
Cohen's Kappa and Gwet's AC1 when calculating inter-rater reliability coefficients: a
study conducted with personality disorder samples. BMC Medical Research Methodology
, 13, 61. doi: 10.1186/1471-2288-13-61
Zwaan, R.A. & Singer, M. (2003). Text comprehension. In A.C. Graesser, M.A. Gernsbacher &
S.R. Goldman (Eds.), Handbook of discourse processes (pp. 83-121). Mahwah, NJ:
Erlbaum.
134
Annexes
135
136
137
138