children's school performance and their parents' causal attributions to ability and...

9
Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study Katja Natale , Kaisa Aunola, Jari-Erik Nurmi Department of Psychology, University of Jyväskylä, Finland article info abstract Available online 31 October 2008 The present study investigated the cross-lagged associations between parents' attributions of ability and effort concerning their children's success and failure, and children's academic performance in kindergarten and primary school. Two hundred seven children and their parents were followed over three years. The parents completed a questionnaire concerning their causal attributions for their children's performance three times. Children's performance in mathematics and reading was tested twice a year. The results showed that children's high academic performance predicted parents' attributions of their children's success to ability, whereas low performance predicted parental attributions to effort. Furthermore, parental attributions to ability were positively related to higher levels of later academic performance. © 2008 Elsevier Inc. All rights reserved. Keywords: Parents' causal attributions Academic performance School transition Ability Effort 1. Introduction Parents' beliefs about the causes of their children's academic outcomes are important as they may have an impact on parents' general conceptions regarding their children's academic skills. Such parental conceptions are then likely to impact their behavior toward their children, and consequently, the development of children's academic skills (Miller, 1995). Previous research has shown that parents' causal attributions are associated with children's performance at school: the higher the level of the children's performance, the more the parents attribute success to ability, and the less they attribute it to effort (Holloway & Hess, 1985; Dunton, McDevitt, & Hess, 1988; Yee & Eccles, 1988). However, only a few longitudinal studies have examined the cross-lagged relationships between parental causal attributions and children's performance, by controlling for the previous levels of the variable before predicting it by another. Consequently, little is known about the direction of inuence, that is, whether parents' causal attributions predict their children's performance (Hess, Holloway, Dickson, & Price, 1984) or whether children's performance contributes to parental causal attributions (Holloway & Hess, 1985). Most studies have also examined parental attributions for older children's performance (Cashmore & Goodnow, 1986; Georgiou, 1999), although it might be assumed that parents' causal attributions begin to evolve in the early school years (Miller, 1995). Consequently, the aim of the present study was to investigate the extent to which children's academic performance predicts their parents' causal attributions concerning the causes of academic success and failure, and the extent to which parents' attributions predict their children's subsequent performance. 1.1. Parental causal attributions The attribution theory of achievement motivation (Weiner, 1985, 1986, 1992) has been extended to encompass the ways in which parents explain and evaluate their children's academic performance. The causal attributions used by parents have typically Journal of Applied Developmental Psychology 30 (2009) 1422 This study is a part of the ongoing Jyväskylä Entrance into Primary School (JEPS) study, and was funded by a grant from the Finnish Academy (63099 and 778230). We would like to express our gratitude to all the children and their parents participating in this study, as well as their kindergarten and primary school teachers and local school authorities. Corresponding author. Department of Psychology (Agora), University of Jyväskylä, P.O. Box 35, 40014 Jyväskylä, Finland. E-mail address: [email protected].(K. Natale). 0193-3973/$ see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.appdev.2008.10.002 Contents lists available at ScienceDirect Journal of Applied Developmental Psychology

Upload: katja-natale

Post on 04-Sep-2016

219 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study

Journal of Applied Developmental Psychology 30 (2009) 14–22

Contents lists available at ScienceDirect

Journal of Applied Developmental Psychology

Children's school performance and their parents' causal attributions toability and effort: A longitudinal study☆

Katja Natale⁎, Kaisa Aunola, Jari-Erik NurmiDepartment of Psychology, University of Jyväskylä, Finland

a r t i c l e i n f o

☆ This study is a part of the ongoing Jyväskylä Entr778230). We would like to express our gratitude to allteachers and local school authorities.⁎ Corresponding author. Department of Psychology

E-mail address: [email protected] (K. Natale).

0193-3973/$ – see front matter © 2008 Elsevier Inc. Adoi:10.1016/j.appdev.2008.10.002

a b s t r a c t

Available online 31 October 2008

The present study investigated the cross-lagged associations between parents' attributions ofability and effort concerning their children's success and failure, and children's academicperformance in kindergarten and primary school. Two hundred seven children and theirparents were followed over three years. The parents completed a questionnaire concerningtheir causal attributions for their children's performance three times. Children's performance inmathematics and reading was tested twice a year. The results showed that children's highacademic performance predicted parents' attributions of their children's success to ability,whereas low performance predicted parental attributions to effort. Furthermore, parentalattributions to ability were positively related to higher levels of later academic performance.

© 2008 Elsevier Inc. All rights reserved.

Keywords:Parents' causal attributionsAcademic performanceSchool transitionAbilityEffort

1. Introduction

Parents' beliefs about the causes of their children's academic outcomes are important as they may have an impact on parents'general conceptions regarding their children's academic skills. Such parental conceptions are then likely to impact their behaviortoward their children, and consequently, the development of children's academic skills (Miller, 1995). Previous research has shownthat parents' causal attributions are associated with children's performance at school: the higher the level of the children'sperformance, the more the parents attribute success to ability, and the less they attribute it to effort (Holloway & Hess, 1985;Dunton, McDevitt, & Hess, 1988; Yee & Eccles, 1988). However, only a few longitudinal studies have examined the cross-laggedrelationships between parental causal attributions and children's performance, by controlling for the previous levels of the variablebefore predicting it by another. Consequently, little is known about the direction of influence, that is, whether parents' causalattributions predict their children's performance (Hess, Holloway, Dickson, & Price, 1984) or whether children's performancecontributes to parental causal attributions (Holloway & Hess, 1985). Most studies have also examined parental attributions forolder children's performance (Cashmore & Goodnow, 1986; Georgiou, 1999), although it might be assumed that parents' causalattributions begin to evolve in the early school years (Miller, 1995). Consequently, the aim of the present study was to investigatethe extent to which children's academic performance predicts their parents' causal attributions concerning the causes of academicsuccess and failure, and the extent to which parents' attributions predict their children's subsequent performance.

1.1. Parental causal attributions

The attribution theory of achievement motivation (Weiner, 1985, 1986, 1992) has been extended to encompass the ways inwhich parents explain and evaluate their children's academic performance. The causal attributions used by parents have typically

ance into Primary School (JEPS) study, and was funded by a grant from the Finnish Academy (63099 andthe children and their parents participating in this study, as well as their kindergarten and primary schoo

(Agora), University of Jyväskylä, P.O. Box 35, 40014 Jyväskylä, Finland.

ll rights reserved.

l

Page 2: Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study

15K. Natale et al. / Journal of Applied Developmental Psychology 30 (2009) 14–22

been divided into those that refer to a child's ability and effort, and those that refer to task difficulty and teachers' competence(Cashmore & Goodnow,1986; Yee & Eccles, 1988). Such causal attributions vary along three dimensions: locus (internal – external),stability (stable – unstable), and controllability (controllable – uncontrollable; Weiner, 1986). Ability and effort, for example, areboth internal causal attributions whereas task difficulty and teachers' competence are external attributions. Ability and effort differon the stability and controllability dimension, however. Ability is a stable characteristic and is beyond the control of the child. Incontrast, effort is unstable across settings, and is controllable by the child. Task difficulty and teachers' competence, in turn, areboth stable and uncontrollable properties. Empirical studies have shown that parents most often attribute their children's successand failure to either ability or effort rather than to task difficulty or teacher competence (Cashmore & Goodnow,1986; Cote & Azar,1997; Georgiou,1999; Holloway & Hess,1985; Kinlaw, Kurtz-Costes, & Goldman-Fraser, 2001; Räty, Vänskä, Kasanen & Kärkkäinen,2002; Rytkönen, Aunola, & Nurmi, 2005; Stevenson & Lee, 1990; Yee & Eccles, 1988). Hence, the present study focused onexamining the relation between these two causal attributions and academic performance.

Researchers have found associations between children's performance and parents' causal attributions. For example, Yee andEccles (1988) found that parents tended to attribute their children's current mathematics success to effort when that performancewas average or low, and attributed it to ability when their performance was high. For failure, effort was perceived as the mostimportant cause. There is also evidence from longitudinal studies that children's past performance predicts their parents' causalattributions. When children perform well at school, parents are likely to attribute their success to a stable cause, such as ability(Dunton et al., 1988; Holloway & Hess, 1985; Rytkönen et al., 2005). By contrast, failure is typically attributed to an unstable cause,such as effort (Holloway & Hess, 1985).

Previous research on the relationship between parental attributions and children's academic performance has several majorlimitations. First, although it has been shown that children's school performance predicts their parents' causal attributions, thoselongitudinal studies showing this relation have not controlled for prior parental attributions. Second, few studies have examinedthe possibility that the relation between parental causal attributions and child academic performance might be bidirectional, thatis, parental attributions might predict the child's performance (Hess et al., 1984). Parents' causal attributions may, for example,influence their expectations and aspirations concerning their children's performance, as well as the support, advice, and guidancethey give to their children (Murphey, 1992). Parents' attributions may also have an impact on their children's behavior at school. Ithas been shown, for example, that adult praise of intelligence increases children's performance-orientation, whereas praise foreffort promotes their mastery-orientation strategies (Kamins & Dweck, 1999; Mueller & Dweck, 1998). Such changes in children'sconceptions may in turn have an impact on their subsequent performance.

A further limitation of earlier research is that most studies have examined older school-aged children (Georgiou, 1999; Yee &Eccles, 1988). It might be assumed, however, that parents' causal attributions are established when children are young (Miller,1995). Moreover, the role of parents' causal attributions in their children's performance may change as children grow older, asmany changes take place in children's school-related activities and also in parent–child relations. A particularly important periodfor the formation of parents' causal attributions, and their impact on children's performance, is the time when their children beginto be faced with the challenge of learning basic academic skills, i.e., the transition from kindergarten to primary school. During thisperiod parents receive increasing amounts of information and feedback concerning their children's progress in learning.

Thus, the goal of the present study was to investigate the cross-lagged associations between parents' causal attributions, andtheir children's academic performance during their transition to primary school. Using longitudinal data in an academic setting,this study examined the extent to which children's school performance predicts their parents' causal attributions, the extent towhich parental causal attributions predict the child's performance, and whether these prospective associations show reciprocalpatterns, consisting of both kinds of cross-lagged associations (Pomerantz & Eaton, 2001).

Previous studies have also shown that parents' causal attributions vary with the child's sex. For example, mothers typicallythink that their sons succeed in mathematics because of their ability, but that their daughters' success is due to effort (e.g., Duntonet al., 1988; Eccles, Jacobs, & Harold, 1990; Lummis & Stevenson, 1990; Räty et al., 2002; Yee & Eccles, 1988). Not all studies,however, have found sex differences. For example, Cashmore and Goodnow (1986) and Rytkönen et al. (2005) found that bothparents attributed their daughters' and sons' success to ability. Because the previous research on sex differences has shownmixedfindings, the present study also investigated this topic.

In Finland, primary school with formal teaching begins one year later than in the United States. Children start their schoolcareer in the year of their 7th birthday. Before primary school, children normally participate in one year of kindergarten, which isequivalent to kindergarten in the U.S. In Finland, kindergarten is not compulsory, but almost all children attend. In Finnishkindergarten there is no formal teaching, but children are encouraged to play with letters and numbers, and concepts related toreading and mathematics. In kindergarten, parents are seldom provided with feedback on their children's progress in learning.However, when they move to primary school, parents typically meet children's teachers and discuss their children's progress withthem. Although grades are not given during the first two school years of primary school, teachers give written reports of children'sprogress in the major subjects (see Aunola, Leskinen, Lerkkanen, & Nurmi, 2004).

1.2. Research questions

The present study examined the following research questions:

(1) Does children's academic performance predict their parents' causal attributions of ability and effort concerning theirchildren's success and failure at school?

Page 3: Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study

16 K. Natale et al. / Journal of Applied Developmental Psychology 30 (2009) 14–22

(2) Do parents' attributions of ability and effort concerning their children's success and failure at school predict their children'ssubsequent academic performance?

(3) Are these lagged relationships different for the parents of boys and the parents of girls?

2. Method

2.1. Participants and procedure

This study was part of the Jyväskylä Entrance into Primary School (JEPS) Study (Nurmi & Aunola, 1999). The original sampleconsisted of all the children born in 1993 (N = 210) and resident in twomedium-sized communities (population 5019 and 8407) inCentral Finland and their parents. These two communities were selected because they included both suburban and rural areas. Thefact that all the children in the two communities were examined helped keep attrition rates low. Parental permission to gather datafrom their children was obtained in August 1999. Parents of 207 children (111 boys, 96 girls) who were 5–6 years old (M = 75month, SD = 3.30 month) at the baseline gave permission.

Information about the children's performance in reading and mathematics was gathered on six occasions: at the beginning(Time 1) and at the end (Time 3) of their kindergarten year, in October 1999 (N = 207) and April 2000 (N = 199); at the beginning(Time 4) and at the end (Time 6) of their first primary school year, in October 2000 (N = 196) and April 2001 (N = 196); and at thebeginning (Time 7) and at the end (Time 9) of their second primary school year, in October 2001 (N = 197) and March 2002 (N =196). The attrition of 11 childrenwas due to the fact that the families of these childrenmoved to other districts andwere not able tocontinue participation in the study later on.

All tests were carried out by trained investigators and took place in a suitable room in the kindergarten or primary schoolpremises. Childrenwhowere absent from the school on the day of testing (e.g., because of illness) were tested as soon as they wereback at school again. Children's math performance was assessed in an individual testing situation at Times 1 and 3, and in a groupsituation in the classroom at Times 4–9. Of the tests assessing children's reading performance only the Sentence Comprehensiontest was conducted in a group situation in the classroom at Times 4–9. Other reading performance tests were performed in anindividual testing situation at all measurement points. For the kindergarten assessment, the children came from 21 kindergartengroups with 3 to 21 (M = 9.8) children from any one kindergarten. In the first grade, the children came from 17 classes with 2 to 26(M = 12.0) children from any one class. In the second grade, the children came from 19 classes with 6 to 30 children (M = 17.2) fromany one class. The reason for the low numbers of participants in some of the classes was that in some schools children fromdifferent grades were combined to form a single class and in others not. Combining grades is common practice in Finland in themore remote small schools where there are few children in any one grade.

A causal attribution questionnaire was mailed to parents on three occasions (Times 2, 5, and 8) during kindergarten (December1999), thefirst primary school year (December 2000), and the secondprimary school year (December 2001).Mothers and fatherswereasked to answer the questionnaires independently and without conferring. Overall 189, 170 and 178 mothers, and 164, 147 and 160fathers answered the causal attribution questionnaires in the kindergarten, first and second years of primary school, respectively.

Families most often included two parents and their children (83%), either the mother or the father living with her/his new spouseand their children (10%); or a single mother with her child/children (7%). The number of children per family ranged from 1 to 11 (M =2.80, SD = 1.50). A total of 18% of themothers and 14% of the fathers had a degree from an institution of university standing, 68% of themothers and 75% of the fathers had a qualification from an institution of professional or vocational education, and 14% of themothersand 11% of the fathers had no occupational education. A total of 17% of themothers and 24% of the fathers wereworking in the higherwhite-collar professions; 53% of the mothers and 25% of the fathers were working in the lower white-collar professions; 17% of themothers and 34% of the fathers were working in blue-collar professions; 6% of the mothers and 12% of the fathers were in privatebusiness; and 7% of the mothers and 5% of the fathers were students, housewives/ husbands, or pensioners.

2.2. Child measures

Children's academic performance was measured in two basic academic skills, mathematics and reading. Children'smathematical performance was assessed with the Diagnostic Test for Basic Mathematical Concepts (Ikäheimo, 1996). Althoughthe structure of the test was identical at all six measurement points, more difficult items were added to the test as the childrenbecame more skilled in mathematics. The test consisted of five subtests (Aunola & Nurmi, 2004; Aunola et al., 2004):

(1) Knowledge of ordinal numbers: Knowledge of ordinal numbers was assessed with two tasks in which children were firstshown a picture of a sequence of boy figures and then asked to circle a particular one (“The boys are in a line. Circle the thirdboy from the beginning.”).

(2) Knowledge of cardinal numbers and basic mathematical concepts. Knowledge of cardinal numbers and basic mathematicalconcepts, such as ‘equal’, ‘more’, and ‘less’, weremeasured by taskswhichbecameprogressivelymore difficult (7 tasks at Time1and 12 tasks at the other measurement times). In each task, the childrenwere shown a picture representing a number of ballsand asked to draw a specific number of balls in the adjacent space (e.g. “Draw five balls less than there are in the model.”).

(3) Number identification. The children's ability to perceive the correspondence between a particular number and the number ofobjects in a picture was assessed using two kinds of tasks (4 tasks at Time 1 and 6 tasks at the other measurement times). Inthe first set of tasks, the children were shown a picture of a set of balls under which four different numbers were written.

Page 4: Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study

17K. Natale et al. / Journal of Applied Developmental Psychology 30 (2009) 14–22

They were asked to circle the number that corresponded to the number of balls pictured. In another set of tasks, the childrenwere shown a picture of a specific number and asked to draw as many balls as indicated by that number.

(4) Word problems. Childrenwere read simple verbal mathematical problems (e.g., “You have seven candies and you get 3 more.How many do you have now?”). After hearing each problem (2 tasks at Time 1, 6 tasks at Times 3 and 4, 10 tasks at Time 6,and 14 tasks at Times 7 and 9), the children were asked to write down the right solution on the answer sheet.

(5) Basic arithmetic. The children's skill in basic arithmetic was assessed by means of a set of visual addition (e.g., “9 + 3 = _”),subtraction (“11 – 2 = _”), multiplication (“8 × 7 = _”) and division tasks (“48 : 6 = _”), and combinations of these (“16 : 4 + 7 = _”).The childrenwere asked to do asmany of them as they could (4 tasks at Time 1, 8 tasks at Time 3,18 tasks at Time 4, 28 tasks atTime 6, and 42 tasks at Times 7 and 9).

One point was given for each correct answer. The total maximum score for the test was 19 in the first measurement (Time 1), 34in the second (Time 3), 44 in the third (Time 4), 58 in the fourth (Time 6), and 76 in the fifth (Time 7) and in the sixthmeasurement(Time 9). The test–retest reliability for the test ranged between .73 and .83. Correlations between the test scores and teacher-ratings of math performance ranged from .60 to .73 (Aunola & Nurmi, 2004).

Children's reading performance was assessed by three tests. The scores on each test were summed to form an overall score forreading. The three tests were:

(1) A letter identification task asked children to name 21 upper case letters. Scoring was based on the number of correctlyidentified letters.

(2) The ReadingWords and Sentences test (Normaalikoulu,1985) asked children to read aloud a set of words and sentences. Thetask beganwith oneword at a time, with 20writtenwords of increasing difficulty. The difficultywas increasedmainly by thefact that words became longer, for example, ‘ja’ (and), ‘isä’ (dad), ‘ikkuna’ (window), ‘tulitikku’ (match). The 20 words werefollowed by two sentences. Testing continued until four successive words/sentences were read incorrectly or were notattempted. One point was scored for each word/sentence read correctly, yielding a maximum score of 22.

(3) A Sentence Comprehension test (Lindeman, 1998) presented four sentences and one picture, and children were asked tochoose which of the 4 sentences best matched the meaning of the picture. The children read the sentences alone whileperforming the task. The test consisted of 20 items, which became progressively more difficult across the test. The test score(maximum 20) consisted of the number of right answers given within 2 min.

A sumscore for readingperformancewas calculatedbyaddingup the scores for the three tests. The test–retest reliability for readingperformancewas .93. Correlations between the test scores and teacher-ratings of reading performance ranged from .49 to .87 (Aunola& Nurmi, 1999). In terms of raw scores, no child received the maximum possible sum score. Because a sum score comprising severaltestswasused rather than separate test scores, nofloor effectwasevident at anymeasurement point. Anoverall academicperformancescore was formed by calculating the mean of the standardized mathematics and reading performance sum scores.

2.3. Parent measures

In the present study, parental causal attributionsweremeasured as general rather than subject-specific attributions. The reason forthis was that during the kindergarten year, when the study began, specific subject areas were not a focus of the children's education.

Parents' causal attributions concerning their children's success and failure at school weremeasured by a 4-item questionnaire thatwas based on the work of Ames and Archer (1987) and Parsons (1980). Two questions assessed parents' causal attributions for theirchildren's success at school (“If my child doeswell in a school assignment, it is probably because...” and “If my child doeswell at school,it is probablybecause...”), and twoassessedattributions for failure at school (“Ifmychilddoesnot dowell in his/her school assignments,it is probably because...”, and “If my child does not know how to do some school assignment, it is probably because...”). Parentsanswered each question separately, by rank-ordering four alternatives according to their importance (e.g., “mostly describes mythoughts”; “least describes my thoughts”). The alternatives for success at school were “The child is talented”, “The child tries hard”,“The teaching has been good”, and “The tasks have been too easy for the child”. The alternatives for failure at school were: “The childlacks the required talent”, “The child has not invested enough effort”, “The teaching has not been good enough”, and “The tasks havebeen toodifficult for the child”. On thebasis of theparents' rank-ordered answers, ameanscorewascreated foreach typeof attribution,that is, one score ranging from 1–4 for each causal attribution of ability, effort, teaching, and task difficulty, tallied separately for thesuccess and failure situations, and separately for mothers and fathers. Mean scores were created for the ability and effort attributionsfor successes and for failures for fathers and formothers. Although the present study included questions onparents' attributions aboutability and effort, task difficulty and teaching competence, attributions about ability and effort were the focus of the present study.Cronbach's alpha ranged between .67 and .86 for mothers', and between .75 and .93 for fathers' causal attributions for success, andbetween .68 and .92 for mothers', and between .75 and .86 for fathers' causal attributions for failure.

3. Results

3.1. Parents' ability and effort attributions

The mean (and SD) ability and effort attribution scores are presented in Table 1 for mothers' and father's views of academicsuccesses and for failures at each time of measurement. As Table 1 shows, these causal attributions appear to be rather stable over

Page 5: Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study

Table 1Mean (SD) ability and effort attribution scores for successes and failures at each measurement time (range possible = 1–4, based on importance ranking)

Success Failure

Type of Attribution Time 2 Time 5 Time 8 Time 2 Time 5 Time 8

Ability 2.89 (.76) 3.10 (.73) 3.01 (.76) 1.77 (.74) 1.78 (.75) 1.76 (.61)Effort 2.58 (.76) 2.65 (.74) 2.55 (.72) 3.06 (.70) 3.17 (.73) 3.11 (.65)

18 K. Natale et al. / Journal of Applied Developmental Psychology 30 (2009) 14–22

the assessment period. Consistent with the attribution literature, it appears that both mothers and fathers weight ability asimportant in successes (see upper left portion of Table 1) and weight effort, specifically low levels of effort, as important inexplaining failures (see lower right portion of Table 1).

3.2. Relations between parental causal attributions and children's academic performance

The correlations between parents' causal attributions of ability and effort and children's academic performance are shown inTable 2. In success situations (see upper portion of Table 2), parents' ability attributions and children's performance showedmoderate positive correlations, whereas parents' attributions to effort showed moderate negative correlations with children'sperformance.

To investigate whether the associations between parental causal attributions and children's academic performance weredifferent for mathematics and reading, the correlations between parental causal attributions and children's performance inmathematics and in reading were calculated separately. Using Fischer's z-test, none of the differences between attribution-readingand attribution-mathematics correlations was statistically significant, with one exception: the correlations between parents'ability attribution at Time 2 and children's mathematical performance at Time 6 (r = .39, p b .01) was stronger than the correlationbetween parents' ability attribution and children's reading performance (r = .22, p b .01) in the same time frame (z = 1.85, p b .05).Full correlation and covariance matrices are available from the first author.

There were no significant sex differences in parental attributions (all tsN .15), showing that parents made similar attributionsabout the causes of sons' and daughters' academic successes and failures. There were sex differences in children's performance atthe beginning of children's kindergarten and 1st grade (Time 1 and Time 4), t(205) = 2.10, p b .05 and t(194) = 2.53, p b .05, favoringgirls over boys at both Time 1 and Time 4 (M = .14, SD = .87 vs. M = −.12, SD = .89) and (M = .17, SD = .82) vs. (M = −.15, SD = .94) forgirls vs. boys, respectively. No other statistically significant sex differences were found.

3.2.1. Parents' causal attributions about ability and children's academic performanceTo further examine the research questions, path models were constructed separately for parents' ability and effort attributions

concerning their children's success and failure at school. The tested models included stability coefficients for children's academicperformance and parents' causal attributions. Further, cross-lagged paths from children's previous academic performance tosubsequent parental causal attributionwere included, as were paths from parents' attributions to children's subsequent academicperformance. In all the tested models the children's performance variable and parents' causal attributions showed substantialstability.

Table 2Pearson correlation coefficients between parent' attributions of success and failure to ability and effort and children's academic performance

Children's performance

Time 1 Time 3 Time 4 Time 6 Time 7 Time 9 M Var.

Ability attribution for successTime 2 .29⁎⁎ .34⁎⁎ .37⁎⁎ .34⁎⁎ .30⁎⁎ .39⁎⁎ 2.89 .57Time 5 .37⁎⁎ .41⁎⁎ .47⁎⁎ .49⁎⁎ .44⁎⁎ .41⁎⁎ 3.10 .56Time 8 .39⁎⁎ .40⁎⁎ .47⁎⁎ .43⁎⁎ .40⁎⁎ .41⁎⁎ 3.01 .56

Effort attribution for successTime 2 − .14⁎ − .16⁎ − .20⁎⁎ − .23⁎⁎ − .16⁎ − .23⁎⁎ 2.58 .57Time 5 − .28⁎⁎ − .28⁎⁎ − .29⁎⁎ − .24⁎⁎ − .23⁎⁎ − .25⁎⁎ 2.65 .55Time 8 − .23⁎⁎ − .18⁎ − .20⁎⁎ − .11 − .12 − .20⁎ 2.55 .51

Ability attribution for failureTime 2 − .09 − .06 − .04 − .05 − .11 − .07 1.77 .54Time 5 0.06 − .07 − .09 .01 − .06 − .02 1.78 .57Time 8 − .18⁎ − .18⁎ − .23⁎⁎ − .16⁎ − .16⁎ − .10 1.76 .39

Effort attribution for failureTime 2 .02 − .00 .02 .05 − .01 .00 3.06 .49Time 5 10 .06 .02 − .01 − .02 .00 3.17 .54Time 8 .08 .07 .11 .20⁎⁎ .13 .12 3.11 .43

Note. Var. = Variance.

Page 6: Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study

19K. Natale et al. / Journal of Applied Developmental Psychology 30 (2009) 14–22

The research questions were analyzed by path modelling using the missing data method (Muthén & Muthén, 1998–2003).Because the distributions of the variables were skewed, themodel parameters were estimated using the MLR estimator (Muthén &Muthén, 1998–2003). Goodness-of-fit was evaluated using five indicators: χ2, Bentler's (1990) comparative fit index (CFI), theTucker–Lewis Index (TLI), Root mean square error of approximation (RMSEA), and the standardized root mean square residual(SRMR). According to Hu and Bentler (1999) values above .95 for TLI and CFI, values below .06 for RMSEA, and a value close to .08 forSRMR can be considered as indicating good fit between the hypothesized model and the observed data.

The model fit the data moderately, χ2 (23, N = 207) = 85.70, p b .001, CFI = .96, TLI = .94, RMSEA = .12, SRMR = .06). Themodification indices suggested that the fit of the model would be increased by estimating paths of children's performance (a) fromTime 6 to Time 9, and (b) from Time 1 to Time 4, and (c) estimating the path from parents' ability attribution at Time 2 to Time 8.After these specifications, the model fit the data well, χ2 (20, N = 207) = 26.25, p = .16, CFI = 1.00, TLI = 0.99, RMSEA = .04, SRMR =0.03. The final model is presented in Fig. 1.

As can be seen in Fig. 1, the results showed that children's academic performance at the beginning of kindergarten (Time 1)predicted parents' ability attribution later on during the kindergarten year (Time 2). Similarly, performance at the beginning of thefirst (Time 4) and second grades (Time 7) prospectively predicted parents' attribution to ability in the first and second grades(Times 5 and 8), even after controlling for the previous level of parents' attributions. The better the children performed at school,the more their parents attributed their success to ability. Further, parents' ability attributions during kindergarten (Time 2)prospectively predicted children's academic performance at the end of kindergarten (Time 3). Similarly, parents' ability attributionduring the first grade (Time 5) also prospectively predicted children's performance at the end of the first grade (Time 6). The moreparents attributed their children's success to ability, the better the children's subsequent performance. Conversely, the less theparents attributed their children's success to ability, the worse the children subsequently performed.

3.2.2. Parents' causal attributions about effort and children's academic performanceAn analogous path model was constructed for parents' effort attributions for success. There was moderate fit to the model, χ2

(23, N = 207) = 76.94, p b .001, CFI = .96, TLI = .94, RMSEA = .11, SRMR = 0.05. After estimating the paths of children's performance (a)from Time 6 to Time 9, and (b) from Time 1 to Time 4, the model fit the data well, χ2 (21, N = 207) = 26.58, p = .19, CFI = 1.00, TLI =.99, RMSEA = .04, SRMR = .03.

The results showed that children's performance at the beginning of the first grade (Time 4) prospectively predicted(standardized estimate = − .21, p b .01) parents' effort attribution during the first grade (Time 5): the poorer the performance thechildren showed, the more the parents attributed their success to effort. Furthermore, parents' effort attributions during thesecond grade (Time 8) prospectively predicted (standardized estimate = − .08, p b .05) children's performance at the end of thesecond grade (Time 9): the more the parents attributed their children's success to effort, the poorer the performance the childrenshowed. No other statistically significant cross-lagged paths were found between parents' effort attribution for success andchildren's performance.

3.2.3. Parents' causal attributions about children's failureAnalogous path models were tested for parents' attributions of ability and effort concerning their children's failure at school.

However, examination of the path coefficients and correlations suggested that there were no cross-lagged associations betweenparents' ability attributions for failure and children's performance, or between parents' effort attributions for failure and children'sacademic performance.

3.3. Sex differences

In order to examine whether these models would show the same fit for both boys and girls, all analyses were also carried outusing the multisample procedure (Jöreskog & Sörbom, 1993). In these analyses, the data were divided into two samples, boys andgirls. Path analyses were carried out assuming that all paths would be equal for these two groups. If the fit of the model was goodand no significant modification indices were found, the model was assumed to fit for both boys and girls equally. Thesemultisample analyses showed that inmost cases the samemodel fitted both boys' and girls' data. The only exceptionwas themodel

Fig. 1. The standardized estimates for the model of parents' ability attribution for success and children's performance. Note. ⁎ p b .05, ⁎⁎ p b .01, ⁎⁎⁎ p b .001.

Page 7: Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study

20 K. Natale et al. / Journal of Applied Developmental Psychology 30 (2009) 14–22

for parents' ability attribution for failure and children's performance. This model, χ2(61, N = 96) = 87.28, p = .02, CFI = .98, TLI = .98,RMSEA = .07 SRMR = .13) showed a path from parents' attribution to ability during kindergarten (Time 2) to parents' attribution toability during second grade (Time 8) only for parents of girls (standardized estimate = .07, p b .001). This path was not significant forboys.

4. Discussion

Although a few longitudinal studies have reported an association between children's school performance and subsequentparental causal attributions (Dunton et al., 1988; Holloway & Hess, 1985; Rytkönen et al., 2005), little is known about whetherparents' causal attributions concerning their children's academic outcomes also predict children's school performance. The resultsof the present longitudinal study showed that they do: parents' attributions to ability after their children's success predicted highlevels of children's subsequent performance in kindergarten and first grade. However, children's high level of performance alsopredicted their parents' causal attributions to ability for success.

The first aim of the present study was to investigate whether children's academic performance would predict their parents'causal attributions to ability and effort concerning their children's academic outcomes. The results showed that the better thechildren performed, the more often their parents attributed their success to ability. These results are in line with previous findingsin which it has been shown that, if children do well at school, their parents are likely to attribute their success to an internal andstable attributions, such as ability (Dunton et al., 1988; Holloway & Hess 1985; Rytkönen et al., 2005). Only one minor effect wasfound for effort attributions: the less well the children performed at school, the more often their parents attributed their academicsuccess to effort during Grade 1 (see also Dunton et al., 1988; Holloway & Hess 1985).

Overall, these findings suggest that parents' causal attributions originate in their perceptions of how their children actuallyperform (Rytkönen et al., 2005). This was the case particularly for parents' ability attributions after children's success. The presentstudy contributes to understanding the role of children's school performance in the formation of parental causal attributions intwo ways. First, children's performance predicted their parents' subsequent causal attributions after controlling for the previouslevels of parental attributions. Second, the findings found earlier with older children were substantively equivalent for youngerchildren in the transition from kindergarten to primary school, suggesting the importance of this period is in the formation ofparental beliefs.

A second finding was that parental attributions for academic outcomes were predictive of children's subsequent performance.The more parents attributed their children's success to ability, the better the children performed in subsequent assessments, evenafter controlling for the previous level of the child's performance. Several possible mechanisms could be responsible for this result.First, parental positive beliefs and causal attributions could enhance the child's self-concept of ability, which in turn improves thechild's academic performance (Aunola, Leskinen, Onatsu-Arvilommi, & Nurmi, 2002; Parsons, Adler, & Kaczala, 1982). Second,parents who do not attribute their children's success to ability may communicate negative expectations to the child, which in turnmay lower the child's self-concept of ability and engagement in school work. A third possible explanation is that parents' causalattributionsmay be reflected in their expectations and aspirations concerning the child (Bugental & Happaney, 2002; Dix & Grusec,1985). Such expectations, when communicated to the child, may then influence the child's subsequent academic achievement(Aunola, Nurmi, Niemi, Lerkkanen & Rasku-Puttonen, 2002). Another possible explanation is that parents' causal attributions mayactivate certain patterns of affect and emotional responses (Dix & Grusec, 1985; Miller, 1995), which then have consequences forthe ways in which parents encourage, support and guide their children. This, in turn, may influence the child's academicperformance. Finally, parents who attribute their children's performance to ability may have confidence in their children's talents,and therefore, also bemore likely to encourage and support their childrenwith their schoolwork (Yun Dai, 2002). The results of thepresent study also suggest that parents' causal attributions are important for their children's performance as early as the transitionfrom kindergarten to primary school.

Overall, the findings suggest that attributions that represent faith in the child having a high level of ability may enhance thechild's further academic success. The results of the present study complement those of some previous studies which have found infavor of the importance of parents' effort attributions. For example, Stevenson and Lee (1990) suggested that if parentscommunicate to their children that they succeed because of effort, children work harder in the future. Moreover, Dweck and hercolleagues have found that praise for effort on the part of the research assistant or imagined teacher promotes children's mastery-orientation (Kamins & Dweck, 1999; Mueller & Dweck, 1998).

Further, the results of the present study suggest that children's academic performance and their parents' causal attributions toability after the child's success form a cumulative, recursive pattern. There are, however, two issues that must be taken into accountin interpreting these results: first, although parents' causal attribution emphasizing children's abilities predicted children'sperformance, the effect size of the reverse path was substantially larger. Second, recursive paths were found particularly duringkindergarten and Grade 1. There are several possible explanations for this pattern. One is that parents pay more attention tochildren's academic performance at the early stages of their school career. Consequently, children's performance may impact theirparents' causal attributions more, which then contributes to their parenting and eventually to the child's academic performance.Another possible explanation is that when children grow older, other people, such as peers and teachers, begin to play moreimportant role in children's performance. This explanation concerns only the relative impact from parents' causal attributions tochildren's behaviors.

In the present study, no cross-lagged associations were found between children's academic performance and their parents'causal attributions concerning children's academic failures. It is possible that parents are more willing to share their attributions

Page 8: Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study

21K. Natale et al. / Journal of Applied Developmental Psychology 30 (2009) 14–22

with their children after successes than after failures. It is also possible that, because during this age period children tend toconfront successes more than failures (Aunola, Rytkönen & Nurmi, 2005), causal attributions for failures are not spelled out byparents in everyday interactions.

We did not find that parents' causal attributions varied for sons and daughters. This result is contradictory to some previousfindings suggesting that parents attribute their daughters' and sons' successful performance to different causes (Dunton et al.,1988; Eccles et al., 1990; Holloway, Kashiwagi, Hess, & Azuma, 1986; Holloway & Hess, 1985; Yee & Eccles, 1988). One likely reasonfor these conflicting findings is that previous studies have typically examined older children (Dunton et al., 1988; Yee & Eccles,1988) amongwhom sex differences are normally stronger (Eccles, et al., 1990; Holloway, Kashiwagi, Hess, & Azuma,1986; Jaworski& Hubert, 1994). Another possibility is that sex differences in parents' evaluations of their children's achievement vary acrossdifferent cultures (Lummis & Stevenson, 1990).

Overall, the results of the present study may have consequences for parent education as children start their school career. Ourresults suggest that parents of primary school children should communicate to their children that they can do well at schoolbecause they have the abilities required to succeed. Parents' communication of such optimism to their children may be beneficialfor the child's academic performance.

There are several limitations that should be considered in any attempt to generalize the findings of the present study. First,although the results showed that parents' causal attributions predicted their children's academic performance, we did not examinethe mechanisms underlying these associations. Future research is needed to identify the major mechanisms through whichparental causal attributions contribute to children's behavior. One example of such a mechanism is children's self-concept ofability, which may function as a mediator between children's performance and parents' causal attributions. Second, although thepurpose of the present study was to examine cross-lagged predictions between parents' causal attributions and their children'sperformance, one has to be cautious in interpreting these predictions as causal mechanisms. For example, it is possible that there issome non-measured third factor that explains the cross-lagged associations found here. Third, the measurement of parents' causalattributions was limited because parents were asked for causal attributions for overall school performance rather than particularschool subjects, such as mathematics and reading. In addition, parents ranked four causal attributions rather than rating themindependently, whichmay have influenced the results. A final limitation is also a strength of this study— the datawere collected inone particular country, Finland. To the extent that our results replicate earlier results, the generalization across countries andcultures lends strength to them. Previous studies have shown that parents' causal attributions and beliefs related to their children'sacademic achievement are influenced by the surrounding culture (Bugental & Happaney, 2002; Crystal & Stevenson, 1991;Holloway, 1988; Lummis & Stevenson, 1990; Stevenson, Chen, & Uttal, 1990; Stevenson & Lee, 1990).

In conclusion, the results of the present study showed reciprocal relations (e.g., Pomerantz & Eaton, 2001) between children'sgood performance and parents' attributions of their children's success to ability, and parents' causal attributions of academicsuccess to ability and children's subsequent school performance. These results suggest that when parents of primary schoolchildren communicate to their children that they have the required abilities to succeed, such optimism from the parents isbeneficial for their children's later academic performance.

References

Ames, C., & Archer, J. (1987). Mothers' beliefs about the role of ability and effort in school learning. Journal of Educational Psychology, 79, 409−414.Aunola, K., Leskinen, E., Lerkkanen, M. -K., & Nurmi, J. -E. (2004). Developmental dynamics of math performance from Kindergarten to Grade 2. Journal of

Educational Psychology, 96, 699−713.Aunola, K., Leskinen, E., Onatsu-Arvilommi, T., & Nurmi, J. -E. (2002). Three methods for studying developmental change: A case of reading skills and self-concept.

British Journal of Educational Psychology, 72, 343−364.Aunola, K., & Nurmi, J. -E. (2004). Maternal affection moderates the impact of psychological control on a child's mathematical performance. Developmental

Psychology, 40, 965−978.Aunola, K., & Nurmi, J.-E. (1999). Unpublished raw data. University of Jyväskylä.Aunola, K., Nurmi, J. -E., Niemi, P., Lerkkanen, M. -K., & Rasku-Puttonen, H. (2002). Developmental dynamics of achievement strategies, reading performance, and

parental beliefs. Reading Research Quarterly, 37, 310−327.Aunola, K., Rytkönen, K., & Nurmi, J. -E. (2005). Developmental dynamics of motivation and academic skills in everyday learning situations (ELS) study. Ongoing

research project: University of Jyväskylä.Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 99, 229−246.Bugental, D. B., & Happaney, K. (2002). Parental attributions. In M. H. Bornstein (Ed.), Being and becoming a parentHandbook of parenting, Vol. 3 (pp. 509-535)

Mahwah, NJ: Erlbaum.Cashmore, J. A., & Goodnow, J. J. (1986). Parent–child agreement on attributional beliefs. International Journal of Behavioral Development, 9, 191−204.Cote, L. R., & Azar, S. T. (1997). Child age, parent and child gender, and domain differences in parents' causal attributions and responses to children's outcomes. Sex

Roles, 36, 23−50.Crystal, D. S., & Stevenson, H. W. (1991). Mothers' perceptions of children's problems with mathematics: A cross-national comparison. Journal of Educational

Psychology, 83, 372−376.Dix, T. H., & Grusec, J. E. (1985). Parent attribution processes in the socialization of children. In I. E. Sigel (Ed.), Parental belief systems: The psychological consequences

for children (pp. 177−199). Hillsdale, NJ: Erlbaum.Dunton, K., McDevitt, T., & Hess, R. (1988). Origins of mothers' attributions about their daughters' and sons' performance in mathematics in sixth grade. Merrill-

Palmer Quarterly, 34, 47−70.Eccles, J. S., Jacobs, J. E., & Harold, R. D. (1990). Gender role stereotypes, expectancy effects, and parents' socialization of gender differences. Journal of Social Issues,

46, 183−201.Georgiou, S. N. (1999). Parental attributions as predictors of involvement and influences on child achievement. British Journal of Educational Psychology, 69, 409−429.Hess, R. D., Holloway, S. D., Dickson, W. P., & Price, G. G. (1984). Maternal variables as predictors of children's school readiness and later achievement in vocabulary

and mathematics in sixth grade. Child Development, 55, 1902−1912.Holloway, S. D. (1988). Concepts of ability and effort in Japan and the United States. Review of Educational Research, 58, 327−345.

Page 9: Children's school performance and their parents' causal attributions to ability and effort: A longitudinal study

22 K. Natale et al. / Journal of Applied Developmental Psychology 30 (2009) 14–22

Holloway, S. D., Kashiwagi, K., Hess, R. D., & Azuma, H. (1986). Causal attributions by Japanese and American mothers and children about performance inmathematics. Journal of Psycholog, 21, 269−286.

Holloway, S. D., & Hess, R. D. (1985). Mothers' and teachers' attributions about children's mathematics performance. In I. E. Sigel (Ed.), Parental belief systems: Thepsychological consequences for children (pp. 177−199). Hillsdale, NJ: Erlbaum.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural EquationModeling, 6, 1−55.

Ikäheimo, H. (1996). Diagnostic of the central mathematical concepts. Helsinki: Oy Opperi Ab.Jaworski, T., & Hubert, N. (1994). Mothers' attributions for their children's cognitive abilities. Infant Behavior and Development, 17, 265−275.Jöreskog, K. G., & Sörbom, D. (1993). LISREL8: Structural equation modelling with the SIMPLIS command language. Hillsdale, NJ: Erlbaum.Kamins, M. L., & Dweck, C. S. (1999). Person versus process praise and criticism: Implications for contingent self-worth and coping. Developmental Psychology, 35,

835−847.Kinlaw, C. R., Kurtz-Costes, B., & Goldman-Fraser, J. (2001). Mothers' achievement beliefs and behaviors and their children's school readiness: A cultural

comparison. Journal of Applied Developmental Psychology, 22, 493−506.Lindeman, J. (1998). Allu- Ala-asteen lukutesti (Reading test for primary school). Turun yliopisto: Oppmistutkimuksen keskus. (University of Turku. Learning

Research Center.)Lummis, M., & Stevenson, H. W. (1990). Gender differences in beliefs and achievement: A cross-cultural study. Developmental Psychology, 26, 254−263.Miller, S. A. (1995). Parents' attributions for their children's behavior. Child Development, 66, 1557−1584.Mueller, C. M., & Dweck, C. S. (1998). Praise for intelligence can undermine children's motivation and performance. Journal of Personality and Social Psychology, 75,

33−52.Murphey, D. A. (1992). Constructing the child: Relations between parents' beliefs and child outcomes. Developmental Review, 12, 199−232.Muthén, L., & Muthén, B. O. (1998–2003). Mplus. User's guide. Los Angeles, CA: Muthén & Muthén.Normaalikoulu (1985). Beginner's reading test. Normaalikoulu ( Jyväskylä University Teacher Training School (primary level). Unpublished test material. University

of Jyväskylä, Finland.Nurmi, J. -E., & Aunola, K. (1999). Jyväskylä entrance into primary school study (JEPS) (ongoing). Finland: University of Jyväskylä.Parsons, J. E. (1980). Instrumentation. Final report. University of Michigan.Parsons, J. E., Adler, T. F., & Kaczala, C. M. (1982). Socialization of achievement attitudes and beliefs: Parental influences. Child Development, 53, 310−321.Pomerantz, E. M., & Eaton, M. M. (2001). Maternal intrusive support in the academic context: Transactional socialization processes. Developmental Psychology, 37,

174−186.Rytkönen, K., Aunola, K., & Nurmi, J. -E. (2005). Parents' causal attributions concerning their children's school achievement: A longitudinal study. Merrill-Palmer

Quarterly, 51, 494−522.Räty, H., Vänskä, J., Kasanen, K., & Kärkkäinen, R. (2002). Parents' explanations of their child's performance inmathematics and reading: A replication and extension

of Yee and Eccles. Sex Roles, 46, 121−128.Stevenson, H. W., Chen, C., & Uttal, D. H. (1990). Beliefs and achievement: A study of Black, White, and Hispanic children. Child Development, 61, 508−523.Stevenson, H. W., & Lee, S. -Y. (1990). Contexts of achievement: A study of American, Chinese, and Japanese children.Monographs of the Society for Research in Child

Development, 221, 1−119.Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological Review, 92, 548−573.Weiner, B. (1986). An attributional theory of motivation and emotion. New York: Springler-Verlag.Weiner, B. (1992). Human motivation. Metaphors, theories, and research. Newbury Park, CA: Sage Publications.Yee, D. K., & Eccles, J. S. (1988). Parent perceptions and attributions for children's math achievement. Sex Roles, 19, 317−333.Yun Dai, D. (2002). Incorporating parent perceptions: A replication and extension study of the internal–external frame of reference model of self-concept

development. Journal of Adolescent Research, 17, 617−645.