a process model of l2 learners’ motivation: from the perspectives of general tendency and...

9
A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences Tomohito Hiromori * English Education Center, Ehime University, Matsuyama, Japan Received 23 April 2008; received in revised form 17 November 2008; accepted 27 November 2008 Abstract The purpose of this study is to examine a process model of L2 learners’ motivation. To investigate the overall process of motivation, the motivation of 148 university students was analyzed. Data were collected on three variables from the pre- decisional phase of motivation (i.e., value, expectancy, and intention) and four variables from the post-decisional phase of motivation (i.e., selective attention, emotion control, motivation control, and coping with failure). Both structural equa- tion modeling (SEM) and cluster analysis were used in order to examine the learners’ motivational process from the per- spectives of general tendency and individual differences. The results showed that subjective estimates of value and expectancy mediate behavioral intention and, in turn, affect motivational maintenance and control during the enactment of the intention. Furthermore, more detailed analysis focusing on individual differences revealed the possibility that concrete implementation of actions might be promoted if subjective value and intention are high, even if expectancy for success is low. These results suggest that a process model of motivation will be a useful research framework for uncovering various motivational processes of L2 learners. Ó 2009 Elsevier Ltd. All rights reserved. Keywords: Process model; L2 motivation; General tendency; Individual differences; Structural equation modeling (SEM); Cluster analysis 1. Introduction Learner motivation is considered to affect the success of learning, not only in the case of second/foreign language (henceforth abbreviated only to L2) instruction, but throughout all aspects of educational activities. Since both practitioners and researchers have a considerable interest in motivation, numerous studies have been conducted thus far (for reviews, see Do ¨ rnyei (2001, 2005), Do ¨rnyei and Schmidt (2001), and Oxford (1996)). One of the common features of these recent studies is the examination of how motivation affects the concrete learning process within a particular classroom context, i.e., the move towards a more process-ori- ented approach to the study of motivation (Do ¨ rnyei and Otto ´, 1998; Ushioda, 1994, 1996; Williams and Bur- den, 1997). 0346-251X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.system.2008.11.009 * Tel./fax: +81 089 927 8107. E-mail address: [email protected] Available online at www.sciencedirect.com System 37 (2009) 313–321 www.elsevier.com/locate/system

Upload: tomohito-hiromori

Post on 29-Oct-2016

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences

Available online at www.sciencedirect.com

System 37 (2009) 313–321

www.elsevier.com/locate/system

A process model of L2 learners’ motivation: Fromthe perspectives of general tendency and individual differences

Tomohito Hiromori *

English Education Center, Ehime University, Matsuyama, Japan

Received 23 April 2008; received in revised form 17 November 2008; accepted 27 November 2008

Abstract

The purpose of this study is to examine a process model of L2 learners’ motivation. To investigate the overall process ofmotivation, the motivation of 148 university students was analyzed. Data were collected on three variables from the pre-decisional phase of motivation (i.e., value, expectancy, and intention) and four variables from the post-decisional phase ofmotivation (i.e., selective attention, emotion control, motivation control, and coping with failure). Both structural equa-tion modeling (SEM) and cluster analysis were used in order to examine the learners’ motivational process from the per-spectives of general tendency and individual differences.

The results showed that subjective estimates of value and expectancy mediate behavioral intention and, in turn, affectmotivational maintenance and control during the enactment of the intention. Furthermore, more detailed analysis focusingon individual differences revealed the possibility that concrete implementation of actions might be promoted if subjectivevalue and intention are high, even if expectancy for success is low. These results suggest that a process model of motivationwill be a useful research framework for uncovering various motivational processes of L2 learners.� 2009 Elsevier Ltd. All rights reserved.

Keywords: Process model; L2 motivation; General tendency; Individual differences; Structural equation modeling (SEM); Cluster analysis

1. Introduction

Learner motivation is considered to affect the success of learning, not only in the case of second/foreignlanguage (henceforth abbreviated only to L2) instruction, but throughout all aspects of educational activities.Since both practitioners and researchers have a considerable interest in motivation, numerous studies havebeen conducted thus far (for reviews, see Dornyei (2001, 2005), Dornyei and Schmidt (2001), and Oxford(1996)). One of the common features of these recent studies is the examination of how motivation affectsthe concrete learning process within a particular classroom context, i.e., the move towards a more process-ori-ented approach to the study of motivation (Dornyei and Otto, 1998; Ushioda, 1994, 1996; Williams and Bur-den, 1997).

0346-251X/$ - see front matter � 2009 Elsevier Ltd. All rights reserved.

doi:10.1016/j.system.2008.11.009

* Tel./fax: +81 089 927 8107.E-mail address: [email protected]

Page 2: A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences

314 T. Hiromori / System 37 (2009) 313–321

For example, Williams and Burden (1997) separate the motivational process into three stages along a con-tinuum: reasons for doing something ? deciding to do something ? sustaining the effort, or persisting. Theyargue that the first two stages involve initiating motivation, whereas the third stage involves sustaining moti-vation, and these two aspects of motivation should be clearly differentiated. Dornyei and Otto also put for-ward an elaborate model of the situation-specific aspect of motivation (Dornyei and Otto, 1998). Theirmodel separates the motivational process into three distinct phases: pre-actional phase (corresponding to‘‘choice motivation” that precedes the launching of action) ? actional phase (corresponding to ‘‘executivemotivation” that energizes the concrete action) ? post-actional phase (corresponding to ‘‘motivational retro-spection” that involves critical reflection). Each phase is assumed to be closely related to the others and toportray motivational processes as they happen in real time.

As Dornyei (2001) states, the two conceptualizations mentioned above are not without antecedents in psy-chological literature. In particular, one of the most important impacts has come from the work of the Germanpsychologists Heckhausen and Kuhl (Heckhausen, 1991; Heckhausen and Kuhl, 1985). They constructed aprocess model of motivation which assumed that there are distinct temporally ordered phases within the moti-vational process, namely, the pre-decisional phase of motivation and the post-decisional phase of motivation.A schematic representation of their model can be represented in Fig. 1.

The pre-decisional phase, which can be seen as the decision-making stage of motivation, involves complexplanning and goal-setting processes during which initial values and expectancies of a task are articulated andevaluated in terms of their desirability and chance of fulfillment, and subsequently intentions are formed. Thisphase can be considered as ‘‘motivation” in the narrow sense (Heckhausen, 1991). On the other hand, thepost-decisional phase, which is the implementational stage of motivation, involves motivational maintenanceand control mechanisms during the enactment of the intention that determine action initiation and persever-ance, and that help to overcome various internal obstacles to action. In order to maintain and sustain theintention until it is fulfilled, various action control strategies (Kuhl, 1987) are said to be used, such as ‘‘selec-tive attention” (intentionally ignoring attractive alternatives or irrelevant aspects), ‘‘emotion control” (inhib-iting emotional states that may undermine the enacting and protection of the intention), ‘‘motivation control”(enhancing the strength of the motivation tendency on which the intention is based), and ‘‘coping with failure”

(not pondering failure for too long and detaching oneself from the unattained goal). In relation to ‘‘motiva-tion,” this phase can be considered as ‘‘action” or ‘‘volition” (Heckhausen, 1991) since it involves concreteimplementation of actions appropriate to the attainment of a goal chosen in the pre-decisional phase.

As has been described above, the defining of motivation on the basis of two phases provides a beneficialviewpoint in terms of providing more effective learning guidance. This is because the form of necessary edu-cational intervention differs depending on which of the two phases a student is in. For example, in the case he/she is unable to form an intention, the teacher is required to have the student set a suitable goal and supportthe formation of an intention that enables that goal to be achieved. In other words, the support required hererelates to ‘‘qualitative” aspects in the form of a goal or objective of behavior. On the other hand, in the casehe/she is unable to implement an intention, the teacher is required to support the pursuit of concrete behaviorso as to achieve the goal set by the student. In other words, the support required here relates to ‘‘quantitative”

aspects in the form of the intensity of actual behavior (such as the amount of study time). In this manner, bydefining motivation in terms of a process-oriented view, it becomes possible to obtain a perspective for pro-viding more suitable learning guidance corresponding to the actual state of students.

Intention implementation

(concrete implementation of actions)

Post-decisional phase

Intention formation

(planning and goal-setting)

Pre-decisional phase

Fig. 1. Schematic representation of a process model of motivation.

Page 3: A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences

T. Hiromori / System 37 (2009) 313–321 315

So far there have been few studies which have taken the process-oriented point of view but some haveappeared (Julkunen, 1992, 2001; Shoaib and Dornyei, 2005; Tremblay et al., 1995). For example, Shoaiband Dornyei (2005) examined different motivational influences and various temporal patterns in languagelearners from a qualitative research perspective. Their study found a variety of factors that had either a posi-tive or a negative impact on learners’ motivational disposition, suggesting that motivation is not a stable statebut a dynamic process that fluctuates over time. Julkunen (1992) also investigated the relationship of two gen-eral aspects of motivation for L2 learning (a preference for challenge and curiosity) and motivation specific tothe learning situation. The author concluded from the results that in studying L2 motivation, both general andsubject-specific motivation theories should be used.

2. Objective of this study

On the basis of the above, the specific objective of this study is to clarify the following point: To identify therelationship between intention formation to decide specific goal-attainment in the pre-decisional phase ofmotivation, and intention implementation to take concrete actions in the post-decisional phase of motivation.

Although, the previous findings suggest that adopting a process-oriented perspective is a particularly prom-ising future direction for motivation research, there have been very few studies on how situation-specific moti-vation influences learners’ actual actions. Therefore, this study focuses on this point, examining therelationship between the pre-decisional phase of motivation (i.e., intention based on learners’ estimates ofvalue and expectancy) and the post-decisional phase of motivation (i.e., action control strategies). The resultsare analyzed from the perspectives of general tendency and individual differences in order to investigate thelearners’ motivational profiles and their influence on the concrete actions. The study represents a usefulattempt to uncover various motivational processes of L2 learners and, hopefully, to orchestrate desirablelearning environments that can serve every learner.

3. Methodology

3.1. Participants and learning activity

The participants were 148 university intermediate learners of English enrolled in a required English lan-guage course. In this course, students were expected to acquire the skills and attitudes needed to obtain infor-mation and communicate actively in English. Emphasis was placed on listening and speaking activities, butreading and writing activities were also included, with the aim of improving students’ basic English ability.

A target learning activity of this study was a ‘‘shadowing” task (a kind of reading out loud task), in whichstudents listened to a CD and immediately afterwards repeated what they had just heard without looking at ascript. This activity was selected because (1) it was believed to help students become more conscious of Englishpronunciation and rhythm while speaking, and (2) it was frequently used in class, so it was a known activityfor them.

3.2. Survey instruments

Questionnaire scales were used for this study and they were administered right after the students had fin-ished the learning activity so that they would have a specific task on which to base their responses. The ques-tionnaire consisted of two parts: the pre-decisional phase of motivation and the post-decisional phase ofmotivation. The following is a general description of each scale along with Cronbach alphas in this study.

3.2.1. The pre-decisional phase of motivation

This scale is a self-report questionnaire that assesses learners’ perceptions of the decision-making stage ofmotivation. The scale contains three subscales (value, expectancy, and intention), resulting in 12 items in total(i.e., four items for each subscale). Value (a = .93) was assessed by the importance the individual learnerattached to a particular task. Sample items included ‘‘I thought that this type of activity would be usefulfor studying English” and ‘‘I thought that this type of activity would be important in terms of proceeding with

Page 4: A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences

316 T. Hiromori / System 37 (2009) 313–321

the study of English.” Expectancy (a = .77) was measured by the learner’s estimates of success in a given task.Items of this subscale contained ‘‘I thought I would be good at this type of activity” and ‘‘I was concerned thatI would not be able to handle this type of activity well” (reversed). Finally, intention (a = .83) was assessed bythe learner’s appraisal of the target task. Sample items included ‘‘I thought I would do as much as I could do”and ‘‘I thought I would be able to do that which I was previously unable to do well.” All of these items wererated on a seven-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). A high score indicatedstrong agreement with the proposed item.

3.2.2. The post-decisional phase of motivation

The second part was composed of four subscales that measured various action control strategies availableto maintain and protect ongoing behavioral intentions. Items in each subscale included four items to assessselective attention (a = .87; e.g., ‘‘I focused on dealing with the activity I was working on” and ‘‘I made aneffort to deal with the activity at hand while not thinking about other things”), four items concerning emotioncontrol (a = .82; e.g., ‘‘I dealt with the activity while relaxing as much as possible” and ‘‘I dealt with the activ-ity while calming myself down”), four items concerning motivation control (a = .68; e.g., ‘‘I proceeded whiletelling myself that I ought to be able to do it” and ‘‘I proceeded while resigning myself to the fact that it wouldbe impossible with my English ability” (reversed)), four items concerning coping with failure (a = .82; e.g., ‘‘Iproceeded while trying not to think too deeply even if things did not go well” and ‘‘I proceeded while trying toworry as little as possible about failure”), resulting in 16 items in total. As with the first part, all of these itemswere rated on a seven-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). A high score thuscorresponded to a high degree of agreement with the proposed item.

3.3. Analysis procedures

The purpose of this study was to investigate the manner in which the pre-decisional phase of motivation(i.e., value, expectancy, and intention) functions in terms of influencing the post-decisional phase of motiva-tion (i.e., selective attention, emotion control, motivation control, and coping with failure) among L2 learnersfrom the perspectives of general tendency and individual differences.

To assess the purpose of this study, two sets of analyses were conducted. First, to examine the general ten-dency, structural equation modeling (SEM; also referred to as covariance structure analysis) was used toempirically investigate the overall relationship between the pre-decisional phase of motivation and the post-decisional phase of motivation. SEM is a more powerful statistical technique than traditionally-used proce-dures (i.e., multiple regression analysis, path analysis) because it enables us to investigate inter-relationshipsnot only between observed and latent variables (i.e., confirmatory factor analysis), but also among latent vari-ables based on substantive theory or previous empirical research (Kline, 2005; Purpura, 1999).

Second, to examine the individual differences, cluster analysis was used to profile the learners based on thescores in the pre-decisional phase of motivation and to investigate their effects on the scores in the post-deci-sional phase of motivation among each group. The Ward method with the squared Euclidean distance tech-nique was chosen for cluster analysis because this procedure tends to combine clusters with a small number ofobservations, and produce clusters with approximately the same number of observations (Hair and Black,2000; Yamamori et al., 2003). The number of meaningful clusters was decided by considering large changesin clustering distances and the characteristics of the resulting clusters.

4. Results

4.1. Descriptive statistics and correlation coefficients

Table 1 presents the means (Mean), standard deviations (SD), and scale correlations for the variables thatprobe students’ appraisal of and engagement in the learning activity. The mean scores of value and intentionwere the highest among all of the variables (M = 5.52 and 5.30, respectively) while the mean score of expec-tancy was below the midpoint (i.e., 4) of the seven-point scale (M = 3.17). As for correlations in the pre-deci-sional phase, intention had a relatively strong relationship with value (r = .53), but intention and value

Page 5: A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences

Table 1Means, standard deviations, and correlations between the two phases of motivation.

Two phases of motivationMean SD 1 2 3 4 5 6 7

The pre-decisional phase

1. Value 5.52 0.87 –2. Expectancy 3.17 1.11 .12 –3. Intention 5.30 0.90 .53** .28** –

The post-decisional phase

4. Selective attention 4.99 0.94 .46** .29** .69** –5. Emotion control 4.43 1.08 .18* .25** .43** .39** –6. Motivation control 4.58 1.23 .39** .22** .55** .45** .49** –7. Coping with failure 4.77 1.00 .19* .23** .44** .37** .28** .22** –

Note: N = 148; the scale’s range, theoretically, is from 1 to 7.*p < .05, **p < .01.

T. Hiromori / System 37 (2009) 313–321 317

demonstrated a weak correlation with expectancy (r = .28 and .12, respectively). Each variable in the post-decisional phase had weak to moderate relationships (r = .22–.49). As far as correlations between the pre-deci-sional phase and the post-decisional phase are concerned, some variables, such as value and emotion controlor coping with failure had weak relationships (r = .18 and .19, respectively), whereas other variables, such asintention and the post-decisional phase of motivation subscales showed relatively strong relationships(r = .43–.69).

4.2. Relationship between the two phases of motivation from the perspective of general tendency

As preliminary analyses, confirmatory factor analyses using the maximum likelihood method were carriedout to confirm the factorial structure of each scale. In assessing the models, various fit indices such as v2 sta-tistic with degrees of freedom, CFI (comparative fit index which is conceptually related to the proportion ofvariability explained by the model), and RMSEA (root mean square error of approximation which is concep-tually related to an average model residual) were used as comprehensive references (for details about theseindices, see Kunnan (1998)). As a result, the fits of the models were confirmed to be acceptable in both scales(The pre-decisional phase of motivation: v2(51) = 120.18, p < .001, CFI = .94, RMSEA = .08; The post-deci-sional phase of motivation: v2(59) = 143.42, p < .001, CFI = .92, RMSEA = .08), suggesting that these twoscales had valid psychometric properties.

To assess the relationship between the two phases of motivation, a structural equation modeling (SEM)approach using the maximum likelihood method was used. In the hypothesized model (see Fig. 2), learners’subjective estimates of value and expectancy were expected to influence intention formation to decide specific

Expectancy

.61**

.55**

Coping with failure

Motivation control

Selective attention

.49**

.80**

Emotion control

.29**

.57**Value

Intention

Fig. 2. The hypothesized model of the relationship between the two phases of motivation. Note: Path coefficients represent standardizedestimates. (**p < .01).

Page 6: A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences

Table 2Means and standard deviations of the pre-decisional phase of motivation subscales among the groups.

Value Expectancy Intention

Mean SD Mean SD Mean SD

Cluster 1 (n = 59) 5.88 0.60 3.91 0.79 5.75 0.63Cluster 2 (n = 25) 6.15 0.65 2.09 0.70 5.97 0.71Cluster 3 (n = 44) 4.78 0.83 3.44 0.56 4.80 0.45Cluster 4 (n = 19) 5.22 0.59 1.57 0.44 4.20 0.97

318 T. Hiromori / System 37 (2009) 313–321

goal-attainment. In turn, this intention was hypothesized to affect concrete implementation of actions appro-priate to goal-attainment (i.e., action control strategies). In order to evaluate the adequacy of the fit of theproposed model to the data, various fit indices were used as comprehensive references. As a result, the fitsof the model were confirmed to be satisfactory (v2(269) = 538.32, p < .001, CFI = .89, RMSEA = .07). Forsimplicity, only the structural model (paths connecting latent factors) is presented in Fig. 2 and the measure-ment model (paths connecting latent factors with their indicators) has been omitted.

As Fig. 2 shows, the results indicated that subjective estimates of value and expectancy significantly andpositively predicted intention (.57 and .29, respectively). In both the correlation and SEM analyses, intentionformation was strongly associated with learners’ perceptions of value, and to a lesser extent with their percep-tions of expectancy. On the other hand, it was found that intention played an important role in using concreteaction control strategies. In particular, the learners with strong intentions were assumed to be attending tospecific aspects of the learning activity during task execution because intention had a strong influence on selec-tive attention (.80).

4.3. Relationship between the two phases of motivation from the perspective of individual differences

In order to profile the learners based on their scores in the pre-decisional phase of motivation, cluster anal-ysis was conducted. With the aid of the dendrogram (a graphic representation of the clustering process)obtained from the analysis, participants (N = 148) were categorized into four groups. Table 2 representsthe means and standard deviations of the pre-decisional phase of motivation subscales in each group.1 Fur-thermore, Fig. 3 shows a visual representation indicating the means of the pre-decisional phase of motivationsubscales in each group.

Based on Table 2 and Fig. 3, a general description of the four groups follows. In cluster 1, the mean scoresof the three variables were relatively high compared to the other clusters. This indicated that the learners inthis group felt their sense of value, expectancy, and intention were generally fulfilled. In cluster 2, the meanscores of value and intention were almost the same as those in cluster 1, but the mean score of expectancytended to be low. The mean scores of the three variables in cluster 3 were all moderate. Finally, each variablein cluster 4 was relatively low, especially in the case of expectancy.

As described above, participants were divided into four groups with a different pre-decisional phase ofmotivation. Next, Table 3 shows the means and standard deviations of the post-decisional phase of motivationsubscales in each group, indicating the actions actually implemented by each learner group with its differentcharacteristics.2 In addition, Fig. 4 shows a visual representation indicating the means of the post-decisionalphase of motivation subscales in each group.

As for cluster 1 and cluster 2, the mean scores of all the variables were relatively high, suggesting that thelearners in these clusters did determine action initiation and tried to overcome diverse internal obstacles toaction, using various action control strategies. The mean scores of the four variables in cluster 3 and cluster

1 To confirm the validity of the grouping, ANOVAs were conducted. Results indicated significant overall differences among the clusters(value: F(3, 143) = 30.86, p < .001; expectancy: F(3, 143) = 82.74, p < .001; intention: F(3, 143) = 44.85, p < .001).

2 Just as in the case of the pre-decisional phases of motivation subscales, ANOVAs were conducted to confirm the validity of thegrouping. Results indicated significant overall differences among the clusters (selective attention: F(3, 140) = 19.84, p < .001; emotioncontrol: F(3, 143) = 11.49, p < .001; motivation control: F(3, 143) = 14.84, p < .001; coping with failure: F(3, 143) = 6.10, p < .01).

Page 7: A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences

0

1

2

3

4

5

6

7

Attention Emotion Motivation Failure

Cluster 1 (n = 59)

Cluster 2 (n = 25)

Cluster 3 (n = 44)

Cluster 4 (n = 19)

Fig. 4. Visual representation of the post-decisional phase of motivation in the four groups.

0

1

2

3

4

5

6

7

Value Expectancy Intention

Cluster 1 (n = 59)

Cluster 2 (n = 25)

Cluster 3 (n = 44)

Cluster 4 (n = 19)

Fig. 3. Visual representation of the pre-decisional phase of motivation in the four groups.

Table 3Means and standard deviations of the post-decisional phase of motivation subscales among the groups.

Selective attention Emotion control Motivation control Coping with failure

Mean SD Mean SD Mean SD Mean SD

Cluster 1 (n = 59) 5.33 0.80 4.85 0.82 5.00 0.98 5.09 0.96Cluster 2 (n = 25) 5.50 0.81 4.60 1.17 5.22 1.26 4.83 1.20Cluster 3 (n = 44) 4.67 0.66 4.13 0.93 4.07 1.00 4.63 0.69Cluster 4 (n = 19) 3.93 1.05 3.49 1.12 3.55 1.30 4.07 1.10

T. Hiromori / System 37 (2009) 313–321 319

4 generally corresponded to the results of the pre-decisional phase of motivation subscales in that all the vari-ables in cluster 3 were moderate, and those in cluster 4 were relatively low.

Considering both results provided from Figs. 3 and 4, the post-decisional phase of motivation was consid-ered to increase in relevance depending on the learners’ pre-decisional phase of motivation. This result indi-cated the possibility that the learners’ intention to decide specific goal-attainment as well as their subjectiveestimates of value and expectancy play an important role in enhancing their volitional engagement with thelearning activity at hand. In other words, positive appraisal strongly influenced actual quality/quantity oflearning behavior.

Page 8: A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences

320 T. Hiromori / System 37 (2009) 313–321

5. Discussion

This study examined the L2 learners’ motivation from the process-oriented point of view. More specifically,it investigated the manner in which the pre-decisional phase of motivation functions in terms of affecting thepost-decisional phase of motivation from the perspectives of general tendency and individual differences.

As a result, it was found that subjective estimates of value and expectancy mediated behavioral intentionand, in turn, strongly affected concrete action control and motivational maintenance. Furthermore, moredetailed analysis focusing on individual differences revealed that learners approached the learning activityin various ways. For example, as is clear from Fig. 3, even when approaching the same learning activity,the manner in which the activity was approached varied considerably among: learners with high levels ofvalue, expectancy, and intention; learners with high levels of value and intention, but a low level of expectancy;and learners with low levels of value, expectancy, and intention.

These results show that it is necessary to use teaching methods that correspond to the learners’ level ofmotivation in order to support their volitional engagement with the learning activity. To put it differently,in order to support learners in adopting an enthusiastic approach, it is most likely necessary to considerthe individual characteristics of each learner and provide classes with various aspects of diversity, such asthe subject matter being learned, its usefulness, and the degree of difficulty.

It is important to note the significant role intention plays in this study, although there have been few L2motivation studies which deal with this construct. Both the variable-centered approach (i.e., general tendency)and also the person-centered approach (i.e., individual differences) verified strong relationships between inten-tion and action control strategies. It seems to be a fact that many learners have an ideal ‘‘I want to” type ofinternal statement (e.g., ‘‘I want to be a fluent speaker of English” and ‘‘I want to watch foreign movies with-out subtitles”), but the bulk of these statements will not be realized. As Dornyei (2001) states, ‘‘Simply havingthe incentive to strive for a goal does not guarantee that the person will actually undertake the effort that isrequired” (p. 94). Thus, intention formation based on learners’ positive estimates of value and expectancy hasa crucial role in fostering their actual engagement in learning and motivational maintenance.

Furthermore, the viewpoint as described above is extremely beneficial for deploying more effective learningsupport. As in this study, for example, behind the decision of students to perform shadowing activities to thebest of their ability, the value and necessity of those activities or the possibility of their realization and expec-tations of success play an important role. In the case students do not make positive efforts to engage in activ-ities, it is necessary for the teacher to ascertain whether or not such points are fulfilled. However, once theintention to attempt an activity has been formed (or, once the students ‘‘cross the Rubicon” (Heckhausen,1991)), an attempt should be made to provide support so that the intention is maintained and developed.At that time, action control strategies in the form of focusing on the objective, not being overly concernedabout failure, and controlling emotions and motivation effectively play an important role. Defining motivationon the basis of two phases is considered to enable more effective deployment of learning support correspond-ing to the actual state of students.

Last but not least, a comparison between cluster 1 and cluster 2 revealed that although expectancy was lowin the case of cluster 2, each of the factors relating to actual intention implementation demonstrated high aver-age scores quite similar to cluster 1. This result suggests that even if expectancy is somewhat low, if awarenesswith respect to value and intention is sufficiently satisfied, there is a possibility of subsequent implementationof an activity being adequately performed. In addition, the results from the SEM analysis also indicated a con-nection between value and intention rather than expectancy. On the basis of these findings, it is necessary tostudy further whether value and expectancy have a parallel effect on intention formation, or whether there issome form of hierarchical relationship between the two.

6. Conclusions

Before the concluding remarks, two suggestions for future research are provided. First, this study focusedon the relationship between the pre-decisional phase of motivation and the post-decisional phase of motiva-tion. Future research needs to examine how other individual difference factors, such as learning styles, lan-guage aptitude, and anxiety, relate to these, and how they influence language learning outcomes as a

Page 9: A process model of L2 learners’ motivation: From the perspectives of general tendency and individual differences

T. Hiromori / System 37 (2009) 313–321 321

whole. In particular, it would be valuable to add learners’ English proficiency into the models used in thisstudy in order to obtain more educational implications.

Second, although the more rigorous technique of structural equation modeling (SEM) as well as clusteranalysis used in this research provide a better test of hypothesized relationships than the multiple regressionanalysis and path analysis that were used in earlier studies, it does not reduce the need for experimental studiesto verify the proposed models. In addition, the participants in this survey consisted of only 148 university stu-dents, which implies the need for follow-up research with more participants. Thus, more study should be con-ducted that is experimental as well as longitudinal in nature.

Learner motivation has received much more attention than any other individual difference factor and the1990s in particular represent ‘‘an explosion of interest in the role motivation played in language learning”

(Ellis, 2008: 677). Although, numerous studies have been conducted thus far relating to motivation, little isknown about the process by which motivation affects actual learning behavior and its outcome (i.e., how moti-vational factors influence the process and outcome of learning). Under the present circumstances, since thisstudy examined a hypothesized process model of motivation followed by verification, it is believed the studyis an important and significant attempt at elucidating the process and mechanism by which learner motivationdevelops and changes, and it is hoped that it will contribute to the development of more ‘‘situation-sensitive”

motivation research.

References

Dornyei, Z., 2001. Teaching and Researching Motivation. Longman, Harlow.Dornyei, Z., 2005. The Psychology of the Language Learner: Individual Differences in Second Language Acquisition. Lawrence Erlbaum,

NJ.Dornyei, Z., Otto, I., 1998. In: Motivation in Action: A Process Model of L2 Motivation. Working Papers in Applied Linguistics, vol. 4.

Thames Valley University, London, pp. 43–69.Dornyei, Z., Schmidt, R. (Eds.), 2001. Motivation and Second Language Acquisition. University of Hawaii Press, HI.Ellis, R., 2008. The Study of Second Language Acquisition, second ed. Oxford University Press, Oxford.Hair, J.F., Black, W.C., 2000. Cluster analysis. In: Grimm, L.E., Yarnold, P.R. (Eds.), Reading and Understanding More Multivariate

Statistics. American Psychological Association, NY, pp. 147–205.Heckhausen, H., 1991. Motivation and Action. Springer-Verlag, Berlin.Heckhausen, H., Kuhl, J., 1985. From wishes to action: the dead ends and short cuts on the long way to action. In: Frese, M., Sabini, J.

(Eds.), Goal-directed Behavior: The Concept of Action in Psychology. Lawrence Erlbaum, NJ, pp. 134–159.Julkunen, K., 1992. General and Situation-Specific Motivation in FL Learning. Revised Version of a Paper Presented at the Regional

Language Centre Seminar on Language Acquisition and the Second/Foreign Language Classroom, Singapore, April 22–26, 1991[ERIC #: ED345550].

Julkunen, K., 2001. Situation- and task-specific motivation in foreign language learning. In: Dornyei, Z., Schmidt, R. (Eds.), Motivationand Second Language Acquisition. University of Hawaii Press, HI, pp. 29–41.

Kline, R.B., 2005. Principles and Practice of Structural Equation Modeling, second ed. Guildford Press, NY.Kuhl, J., 1987. Action control: the maintenance of motivational states. In: Halish, F., Kuhl, J. (Eds.), Motivation, Intention, and Volition.

Springer-Verlag, Berlin, pp. 279–291.Kunnan, A.J., 1998. An introduction to structural equation modeling for language assessment research. Language Testing 15 (3), 295–332.Oxford, R.L. (Ed.), 1996. Language Learning Motivation: Pathways to the New Century. University of Hawaii Press, HI.Purpura, J.E., 1999. Learner Strategy Use and Performance on Language Tests: A Structural Equation Modeling Approach. Cambridge

University Press, Cambridge.Shoaib, A., Dornyei, Z., 2005. Affect in life-long learning: exploring L2 motivation as a dynamic process. In: Benson, P., Nunan, D. (Eds.),

Learners’ Stories: Difference and Diversity in Language Learning. Cambridge University Press, Cambridge, pp. 22–41.Tremblay, P.F., Goldberg, M.P., Gardner, R.C., 1995. Trait and state motivation and the acquisition of Hebrew vocabulary. Canadian

Journal of Behavioral Science 27 (3), 356–370.Ushioda, E., 1994. L2 motivation as a qualitative construct. Teanga 14, 76–84.Ushioda, E., 1996. Learner Autonomy 5: The Role of Motivation. Authentik, Dublin.Williams, M., Burden, R., 1997. Psychology for Language Teachers. Cambridge University Press, Cambridge.Yamamori, K., Isoda, T., Hiromori, T., Oxford, R.L., 2003. Using cluster analysis to uncover L2 learner differences in strategy use, will to

learn, and achievement over time. International Review of Applied Linguistics in Language Teaching 41 (4), 381–409.