Journal of Educational Psychology 2010 American Psychological Association

2010, Vol. 102, No. 4, 804 816 0022-0663/10/$12.00 DOI: 10.1037/a0021075

Ability Self-Concepts and Subjective Value in Literacy:

Joint Trajectories From Grades 1 Through 12

Isabelle Archambault Jacquelynne S. Eccles and Mina N. Vida

Universite de Montreal University of Michigan

Because literacy skills are critical for most academic subject matters, researchers have become increas-
ingly interested in understanding childrens motivation in this domain as a way to increase academic
success. In this study, we extend previous work by looking at the heterogeneity of childrens motivational
changes in literacy across Grades 112. We used a cross-sequential design based on 3 different cohorts
of children (N 655) coming from 10 public elementary schools. Data were collected over an 8-year
period (1988 1996) starting when children were in 1st, 2nd, and 4th grades. Using multiple-process
growth mixture modeling, we identified 7 groups of children showing distinct trajectories of change in
literacy subjective task value and ability self-concept across school years. For all children, ability
self-concept and subjective task value decreased between Grade 1 and Grade 12; however, this decline
was more apparent for some youths than for others. Our findings suggest the importance of school-based
interventions that aim to prevent decreases in student motivation for literacy over time and that focus on
the early detection of children at risk for dramatic declines.

Keywords: ability self-concepts, task value, literacy, motivation, trajectories

In most industrialized societies, literacy skills are essential for
daily functioning. Good reading and writing abilities are associated
with important life consequences, such as better employment rates
and higher socioeconomic status (SES; Guthrie, Schafer, &
Hutchinson, 1991; Madden, Slavin, Karweit, Dolan, & Wasik,
1993; Raudenbush & Kasim, 1998). Because these skills are
central to most academic subject matters (Wilson & Trainin,
2007), many youths who do not acquire basic reading and writing
abilities become alienated from school, leading to disengagement and
eventually to dropout (Janosz, Archambault, Morizot, & Pagani,
2008). In light of such consequences, researchers have become
increasingly interested in understanding childrens motivation in
literacy as a way to increase their reading comprehension and
writing skills (Guthrie, McRae, & Klauda, 2007). In the present
This article was published Online First October 4, 2010.
Isabelle Archambault, Ecole de Psychoeducation, Universite de Mon-
treal, Montreal, Quebec, Canada; Jacquelynne S. Eccles and Mina N. Vida,
Institute for Social Research, University of Michigan.
The preparation of this article was made possible by a postdoctoral
fellowship awarded to Isabelle Archambault by the Programme Persever-
ance et Reussite Scolaire du Fond Quebecois pour la Recherche sur la
Societe et la Culture (Quebec, Canada). This research was supported by
National Institute for Child Health and Human Development Grant
HD17553 to Jacquelynne S. Eccles, A. Wigfield, P. Blumenfeld, and R.
Harold. We thank the principals, teachers, students, and parents of the
cooperating school districts for their participation in this project. We also
thank the following people for the work they did on this project: Janis
Jacobs, Jennifer Fredricks, Corinne Alfeld, Carol Freedman-Doan, and
Lori Rudy.
Correspondence concerning this article should be addressed to Isabelle
Archambault, Ecole de Psychoeducation, Universite de Montreal, C.P.
6128, succ. Centre-Ville, Montreal, Quebec, Canada. E-mail: isabelle
.archambault@umontreal.ca
804
study, we investigated how different groups of children change
from Grade 1 through Grade 12 along two important components
of motivation: subjective task value and ability self-concepts for
school-based literacy activities. Moreover, we examined whether
specific individual and family characteristics could predict these
distinct motivational pathways.
Ability Self-Concept and Subjective Task Value
in Literacy
Subjective task value and ability self-concepts represent two widely
studied components of literacy motivation (Baker & Wigfield, 1999;
Chapman & Tunmer, 1995; Durik, Vida, & Eccles, 2006; Eccles,
1984; Guthrie, Wigfield, & VonSecker, 2000; Jacobs, Lanza,
Osgood, Eccles, & Wigfield, 2002; Wigfield & Eccles, 1994).
Ability self-concept refers to childrens own assessment of their
skills and capability to perform well in a specific domain, whereas
subjective task value is generally defined in terms of child interest,
perceived utility, cost, and intrinsic motivation (Chapman & Tun-
mer, 1995; Durik et al., 2006; Eccles et al., 1983; Schunk &
Zimmerman, 1997; Wigfield & Guthrie, 1997). Childrens self-
perceptions and beliefs have been shown to predict their selection
of activities and level of commitment to them (Eccles, 1984;
Eccles et al., 1983). Children tend to select classes, participate
actively, and achieve well in domains that they enjoy, find impor-
tant, and believe they can succeed at. Children who enjoy reading
and writing generally believe it is important in their everyday life
and future and report positive beliefs about their own reading and
writing abilities. They also spend more time reading for leisure,
choose language arts as elective courses in high school more often,
and are more likely to engage in language arts activities (Durik et
al., 2006; Guthrie, Wigfield, et al., 2006). In addition, these chil-
dren show higher involvement in deep texts processing (Schiefele,
 TRAJECTORIES OF SELF-CONCEPT AND VALUE IN LITERACY
1999), show better written text comprehension (Guthrie, Wigfield,
et al., 2006), report higher grades in literacy-related activities
(Baker & Wigfield, 1999; Cunningham & Stanovich, 1997; Pin-
trich & Schunk, 1996; Sweet, Guthrie, & Ng, 1998), and aspire for
careers that require strong reading comprehension skills (Baker &
Wigfield, 1999; Durik et al., 2006; Wigfield & Guthrie, 1997).
Joint Development of Ability Self-Concept and
Subjective Task Value
Although children who both feel competent and perceive greater
value in reading and writing appear to be at an advantage, longi-
tudinal studies indicate that these beliefs globally diminish as
children get older (Durik et al., 2006; Eccles, Wigfield, & Schief-
ele, 1998; Jacobs et al., 2002; Stipek & Mac Iver, 1989; Wigfield
& Eccles, 1994; Wigfield et al., 1997). As suggested by many
authors (Higgins & Parsons, 1983; Nicholls & Miller, 1984; Par-
sons & Ruble, 1977; Stipek, 1984; Wigfield, Eccles, Mac Iver,
Reuman, & Midgley, 1991), the nature of this decrease is likely
explained, at least in part, by changes children encounter as they
grow up. Because children in the early elementary years have a
limited experience in age-segregated schools, they tend to be
overly optimist. As such, they may have positive beliefs about
their ability to achieve or unrealistic expectations of how interest-
ing school subjects will be. As child cognitive skills develop,
school environments change and become more focused on perfor-
mance and social comparisons (Ames, 1992; Eccles, Lord, &
Midgley, 1991; Feldlaufer, Midgley, & Eccles, 1988; Ruble, Bog-
giano, Feldman, & Lobel, 1980; Stipek & Mac Iver, 1989). As a
result, children who are experiencing academic difficulties are
likely to develop more nuanced impressions of their own compe-
tencies, especially when they compare themselves with peers.
Globally, these changes in child perceptions of themselves and
others could lead them to feel less competent in a given domain.
To protect their more general sense of self-worth, children may
then begin to reduce the value that they place on tasks associated
with failure (Covington, 1992, 1998; Eccles et al., 1998; Harter,
1982).
The changes operating over time on childrens task value and
ability beliefs are neither similar nor independent. Ability self-
concept is more likely to remain stable across school years than
subjective task value (Durik et al., 2006). Further, both compo-
nents are part of a whole and need to be integrated to understand
childrens motivational experience over time. According to Eccles
et al.s (1983) expectancy-value model, childrens ability self-
concepts in a specific domain directly influence their expectations
of success and task value in this domain. Therefore, children who
feel less competent in literacy should come to value activities in
this domain less. They are also likely to spend less time and effort
developing their reading and writing skills (Deci & Ryan, 1985;
Eccles et al., 1998; Jacobs et al., 2002). Although some studies
(Nurmi & Aunola, 2005; Spinath & Spinath, 2005; Spinath &
Steinmayr, 2008) have failed to find evidence for these predic-
tions, others have highlighted a relationship between ability self-
concepts and subjective task value (Durik et al., 2006; Eccles et al.,
1983; Eccles & Wigfield, 1995; Jacobs & Eccles, 2000; Jacobs et
al., 2002; Marsh, Trautwein, Ludtke, Koller, & Baumert, 2005;
Wigfield & Eccles, 1994). In their research, Jacobs et al. (2002)
studied the longitudinal changes of childrens subjective task val-
805
ues for English, math, and sports before and after controlling for
ability beliefs. Globally, they found that in every domain, both
dimensions of achievement motivation decreased across Grades
112. Further, they demonstrated that perception of ability ex-
plained around 40% of the decrease in childrens subjective task
value in each of these domains. Although task value can have
several other sources within individuals or contexts (e.g., percep-
tions of individual progress, autonomy in the classroom; Assor,
Kaplan, & Roth, 2002; Spinath & Steinmayr, 2008), the findings
from this study suggest that the changes children experience in
terms of their self-concept and beliefs are related and, therefore,
ought to be integrated together in studying child development.
Individual Differences in Trajectories of Change
Although there certainly is a group of children who follow a
trajectory of change characterized by the same general decline of
literacy ability self-concept and subjective task value that is found
in the average population, this general pattern is unlikely to rep-
resent the reality of every child. Some might maintain more
positive and stable subjective task value and ability self-concept
over the years, whereas others might report low motivation very
early in their academic experience. Unfortunately, we know little
about individual and group differences in the development of these
two components. In fact, most research on childrens motivation
fails to consider simultaneously the multidimensionality of chil-
drens motivation and the nature and course of its development for
different groups.
Some research has addressed differences in the way motiva-
tional constructs cluster together. A study by Roeser, Eccles, and
Sameroff (1998) identified four subgroups of adolescents on the
basis of their pattern of academic school and social-emotional
functioning in the seventh grade: a positive adjustment group, a
poor school value group, a poor mental health group, and a
multiple problem group. In the literacy domain, another study by
Baker and Wigfield (1999) identified different clusters of fifth-
and sixth-grade children on the basis of their reading motivation.
They distinguished seven groups characterized by distinct motiva-
tional profiles. One third of youths were characterized by high
scores on different dimensions of reading motivation (e.g., self-
efficacy, importance beliefs, involvement, curiosity), 15% re-
ported low scores, and the remaining half presented mixed profiles
(low or high scores on only some dimensions).
Other studies modeled longitudinal patterns of student engage-
ment and motivation in school. For example, a study by Ratelle,
Guay, Larose, and Senecal (2004) described the trajectories of
different types of school motivation at the time of transition from
high school to college. These authors demonstrated that some
motivational styles were more stable over the years than others.
They also showed that a decrease in motivation was observed after
the school transition but only for a small group of students.
Similarly, Midgley, Eccles, and their colleagues have shown that
the trajectories of change in both ability self-concept and subjec-
tive task value for mathematics vary substantially across the junior
high school transition depending on student classroom experiences
and initial levels of competence (see Eccles, Midgley, et al., 1993,
for summary). Finally, Archambault, Janosz, Morizot, and Pagani
(2009) identified six groups of students presenting distinct trajec-
tories of behavioral, affective, and cognitive engagement in school
806 ARCHAMBAULT, ECCLES, AND VIDA
from 12 to 16 years of age. They showed that although the majority
of students presented stable and positive educational pathways
across high school, one third of adolescents reported important
changes and disconnectedness on multiple facets (behavioral, af-
fective, cognitive) of their school experience over time.
Taken together, these studies highlight the relevance of consid-
ering the multidimensionality of childrens motivation when
studying development across the whole academic experience. Fur-
ther, they also point out the importance of distinguishing motiva-
tion both for individuals and groups to obtain a more integrated
portrait of childrens development (Willett & Sayer, 1996). Build-
ing upon this work, in the present study we modeled the joint
trajectories of literacy ability self-concept and subjective task
value that children follow across Grades 112. Unlike prior re-
search that has focused on the average decline characterizing the
majority of children, we identified distinct groups following het-
erogeneous pathways on these two dimensions. Moreover, we
examined whether different individual and family characteristics
can contribute to predict these trajectories of change over time.
Predictors of Subjective Task Value and Ability
Self-Concept Trajectories
Researchers have identified several individual, familial, social,
and academic factors that have contributed to childrens ability
self-concepts and subjective task value in reading and writing
activities (Deci, Vallerand, Pelletier, & Ryan, 1991; Eccles et al.,
1983; Eccles, Wigfield, Harold, & Blumenfeld, 1993; Guthrie,
Hoa, Wigfield, Tonks, & Perencevich, 2006; Wigfield & Eccles,
2002). In the present study, we only considered early characteris-
tics related to children or families. Among these characteristics,
past history of achievement represents a good predictor of chil-
drens motivational changes (Corpus, McClintic-Gilbert, & Hay-
enga, 2009; Harter & Connell, 1984; Jacobs et al., 2002; Marsh,
Byrnes, & Yeung, 1999; Marsh et al., 2005; Skaalvik & Hagtvet,
1990). Children who achieve well and invest efforts in school
would be more likely to receive positive feedback on their abilities
and, accordingly, would report more positive perceptions of com-
petencies and greater value of academic tasks over time (Harter &
Connell, 1984).
Gender is also an important predictor of childrens motivation
changes in literacy-related activities. Although gender differences
do not appear in every study (Anderman et al., 2001), over time
girls tend to place higher value on literacy activities and, as a
result, report higher confidence in their abilities in this domain
than boys (Baker & Wigfield, 1999; Durik et al., 2006; Eccles,
1984; Eccles et al., 1989; Jacobs et al., 2002; Marsh, 1989;
Wigfield et al., 1991, 1997). Moreover, girls advantage on boys
regarding their self-concept should increase over time (Durik et al.,
2006; Jacobs et al., 2002).
Finally, parents perceptions of their childrens abilities and
family socioeconomic background have also received considerable
attention in the literature (Bradley & Corwyn, 2002; Brooks-Gunn
& Duncan, 1997; Chapman & Tunmer, 1995; Chapman, Tunmer,
& Prochnow, 2000; Eccles et al., 1983; Fredricks & Eccles, 2002;
Jacobs & Eccles, 1992; Parsons, Adler, & Kaczala, 1982; Rimm-
Kaufman, Pianta, Cox, & Bradley, 2003; Wilson & Trainin, 2007).
According to Eccles et al.s (1983) expectancy-value model, when
parents report high positive believes about their childrens com-
petence in a specific domain, children tend to develop high ability
self-concepts in this domain and to achieve better over time.
Otherwise, when parents hold negative estimates of their chil-
drens reading/English competencies, children tend to feel less
competent in this domain and report lower achievement. Further,
combined with the fact that they often are exposed to low quality child
care and attend schools with both large concentrations of children
with special needs and less resources (e.g., extracurricular activities,
professional services), children from low SES families often de-
velop their reading abilities more slowly (Aikens & Barbarin,
2008; Chapman & Tunmer, 1995; Chapman et al., 2000; Evans,
2004; Whitehurst & Lonigan, 1998; Yeung, Linver, & Brooks-
Gunn, 2002). Over the long run, many experience increased dif-
ficulties with academic material because of their limited literacy
and may respond to this increased difficulty by disengaging from
the academic activities at school (Brooks-Gunn & Duncan, 1997;
Duncan & Brooks-Gunn, 1997).
Despite the substantial literature on individual and family pre-
dictors of childrens motivational changes over time, we know
little about the early role of these predictors on childrens distinct
trajectories of change. Hence, on the basis of the natural principle
of multifinality, which stipulates that the same characteristic may
influence different outcomes over time (Cicchetti & Rogosch,
1996), it is important to consider early factors that predict the
distinct motivational pathways that children follow across their
whole school experience. As suggested by Spinath and Steinmayr
(2008), understanding such factors represents an important step for
researchers and practitioners in understanding characteristics that
have the potential to buffer childrens motivational decrease. This
knowledge will support early prevention programs for children
placed at risk of experiencing such declines.
Goals of the Study
Overall, research highlights that for the majority of children,
their ability self-concept and subjective task value tend to decrease
over time. Yet, some students present more stable and positive
motivation pathways, whereas others report low motivation over
time (Archambault et al., 2009; Baker & Wigfield, 1999; Ratelle et
al., 2004; Roeser et al., 1998). Further, such declines appear to be
influenced by the transaction between individual and environment
factors (Sameroff & Fiese, 1990). However, there is a lack of
knowledge on how these two components of motivation develop
for different groups of children across the whole school experi-
ence. Further, despite the substantial literature on individual and
family predictors of child motivational changes over time, we
know little about the role of early predictors that affect distinct
trajectories of change. In this research, we overcome these limi-
tations by studying the heterogeneity of motivational changes in
literacy across Grades 112. Instead of focusing on the general
positive, negative, or declining trajectories found in previous re-
search, we expected to identify more groups of children following
distinct trajectories of subjective task value and ability self-
concepts. Furthermore, we assess the contribution of early specific
individual and family characteristics to explain interindividual
differences in the development of these trajectories. By studying
the joint development of these motivational dimensions and by
identifying factors that predict these trajectories of change, we
hope to enrich our understanding of individual development. We
 TRAJECTORIES OF SELF-CONCEPT AND VALUE IN LITERACY
also believe that the results of the present study can help guide the
elaboration of specific interventions directed to particular groups
of children.
Because there are few comparable studies, it is difficult to
hypothesize the number or shape of the trajectories that might be
expected. Nevertheless, consistent with past research that studied
motivation at various ages or for the population average (Archam-
bault et al., 2009; Baker & Wigfield, 1999; Jacobs et al., 2002;
Ratelle et al., 2004), we were able to formulate the following
hypotheses. First, we predict the existence of at least three trajec-
tories: a high trajectory where children report elevated and stable
literacy subjective task value and ability self-concepts over time, a
low trajectory where children report low subjective task value in
reading and writing activities and negative perception of their
abilities from school entry to the end of high school, and a
declining trajectory where children report a constant decline on
subjective task value and ability self-concepts. In line with the
typical normative declines in motivation during adolescence
(Jacobs et al., 2002), we predict that this last group will charac-
terize the most children. Second, for theoretical reasons (Deci &
Ryan, 1985; Eccles et al., 1983), we also predict that for most
children, a decline in their ability self-concept would precede a
decline in subjective task value. Finally, we predict that compared
with groups of children reporting stable and high competence
beliefs and value over time, those following low or decreasing
trajectories of ability self-concept and subjective task value will
differ in specific ways on early individual and family characteris-
tics that place them at risk. These include being a boy, having a
past history of disengagement and underachievement, coming
from a low income family, and having parents who hold negative
perceptions of their competencies (Eccles et al., 1996).
Method
Sample
This study is part of the Childhood and Beyond Longitudinal
Project (Eccles et al., 1983). The sample was composed of 655
children (49.2% boys) who participated in the first year of the
study and who presented valid answers on at least three of the five
waves of data collection. Children originally came from 10 ele-
mentary public schools located in four middle-class communities
of a large city in the Midwest. Every child at the appropriate grade
level in each school was asked to participate. Of the children,
75% 80% agreed to take part in the study and received parental
consent to do so. These children were followed over their elemen-
tary, middle, and high school years (for more information about
this sample, see Jacobs et al., 2002).
Table 1
Number of Children by Cohort and Grade at Each Year of Data Collection
Time point (year of assessment)
Cohort Time 1 (1988) Time 2 (1989)
1 (n 215) Grade 1 Grade 2
2 (n 232) Grade 2 Grade 3
3 (n 208) Grade 4 Grade 5
807
This study used a cross-sequential design based on three differ-
ent cohorts of children. As showed in Table 1, during the first year
of the study, children of the first, second, and third cohorts were in
first, second, and fourth grades, respectively. Using such a design,
we were able to get information on child subjective literacy value
and ability self-concept from Grade 1 to Grade 12. Data for the
11th grade were missing in every cohort. However, these data
were not problematic given that our analytic method allowed for
the presence of missing data on the repeated measures that were
used to build the trajectories.
Measures and Procedures
Data were collected between 1988 and 1996. The assessment
took place in the participants classrooms during the spring period.
During the first three years of data collection, questions were read
aloud to the children to foster their understanding. During these
waves of data collection, mothers were asked to complete ques-
tions about family characteristics and their own perception of her
childs competency in literacy. Finally, teachers were also asked to
answer questions regarding their perception of individual childs
abilities and efforts in reading. A 3-year interruption in data
collection occurred because of a break in funding. During the next
three waves, which took place in junior and senior high school,
participants read the questionnaires themselves (for more details,
see Jacobs et al., 2002).
Task beliefs. We assessed child literacy beliefs during the five
waves of data collection. Comparable wording was used over
years; however, to be relevant to what children learned in school,
we asked the children in elementary school about their self-beliefs
about reading and the children in secondary school about their
self-beliefs about English. We assessed Reading/English Ability
Self-Concept using five items that measured childrens perception
of their abilities in reading (e.g., Compared to other subjects how
good are you at reading?). We measured Reading/English Sub-
jective Task Value using four items that asked children how
important and useful reading/English was to them (i.e., How
useful is what you learn in reading). Both scales were answered
using 17 Likert response scales (e.g., 1 not useful, 7 very
useful) and showed good internal reliability in all waves (for more
details, see Jacobs et al., 2002). To sharpen the comparison of
individual changes over time and to ensure that changes in child
ability self-concept and value over time are not due to changes in
childrens understanding of the questions, we centered scores on
these two scales, across cohorts, on the grand mean (the mean of
the whole sample for all time points; Bryk & Raudenbush, 1992).
At each grade, the grand mean of 5.2230 was subtracted from
student scores on ability self-concept, and the grand mean of
Time 3 (1990) Time 4 (1994) Time 5 (1996)
Grade 3 Grade 7 Grade 9
Grade 4 Grade 8 Grade 10
Grade 6 Grade 10 Grade 12
808 ARCHAMBAULT, ECCLES, AND VIDA
5.0508 was subtracted from student scores on subjective task
value. This strategy enabled us to the interpret students scores on
ability self-concept and value in relation to the sample mean.
Covariates at the First Year of Assessment
Childrens gender. Information on gender was obtained from
official records. Female participants were coded as 0, and male
participants were coded as 1.
Family income. Family income was self-reported by the
mother on a 9-point scale (1 under $10,000 a year, 9 over
$80,000).
Reading/learning efforts and grades. Childrens grades,
reading, and learning efforts were reported by teachers on the first
year of assessment. Both scales were measured on a 15-point scale
(E minus to A plus).
Parental perception of their childs reading ability. Paren-
tal perception of their childs reading ability was self-reported
by the mother at Time 1. This scale comprises five items ( .94)
that assessed maternal perception of their childs reading compe-
tencies (e.g., How good is your child at reading). Parents class would provide a better fit. Finally, we also relied on the
responses were evaluated on a 7-point Likert scale (1 not at all
good, 7 very good).
Analytic Strategy
Multiple-process growth mixture modeling (GMM; B. O. Muthen,
2001, 2004) was implemented in Mplus to identify subgroups of
children following distinct trajectories of literacy subjective task
value and ability self-concepts. GMM does not assume that the
population of children is homogeneous and can be represented by
a single multivariate normal distribution. Rather, GMM assumes
that individuals follow heterogeneous, quantitatively and qualita-
tively distinct trajectories of change (B. O. Muthen, 2001; Nagin,
1999). On the basis of repeated measures, Mplus permits one to
estimate simultaneously the change parameters (i.e., intercept and
slope) on two different but related outcomes. This kind of joint
development model (see L. K. Muthen & Muthen, 2006) allowed
us to examine simultaneously the developmental course of chil-
drens subjective task value and ability self-concept from Grade 1
through Grade 12. To identify childrens membership in these joint
trajectories, GMM classifies each individual in a probabilistic
manner.
Covariates were included in the different models to predict child
membership in each trajectory (see B. O. Muthen, 2001, 2004).
The relationships identified with the covariates are model-based
rather than post hoc estimations. Because Mplus allows missing
data in all parts of the GMM except for covariates, missing values
on literacy subjective task value and ability self-concept were not
problematic. Yet, we used the PATTERN option in Mplus to
consider the type of missing data associated with the multiple
cohort design. To make the distinction between longitudinal ef-
fects and cohort effects, we tested the assumption of invariance of
growth parameters across cohorts and found no variations. For the
covariates, we imputed missing data (less than 10%) using the
NORM multiple imputations program (Schafer, 1999). NORM
uses an iterative method based on EM algorithm to impute missing
data by drawing values from the conditional distribution of the
variables given their observed data (for technical details, see
Schafer, 1999). To limit the danger of identifying a local solution
that is not the true maximum likelihood, we used the random starts
features available in Mplus and tested 500 random starting value
sets for all of our models (McLachlan & Peel, 2000).
Finally, we selected the best model on the basis of the conver-
gence of different statistical indices and on substantive arguments
(see B. O. Muthen, 2003). We used the log-likelihood ratio (LLR),
the Akaike information criterion (AIC; Akaike, 1987), the Bayes-
ian information criterion (BIC; Schwarz, 1978), and the sample-
size adjusted Bayesian information criterion (ABIC; Sclove,
1987). The smallest absolute values of these indices generally
indicate a better fit model. However, it is preferable to rely on the
BIC and ABIC rather than the AIC (particularly the ABIC), as
these are stricter and usually favor the selection of a more parsi-
monious model with fewer trajectories (see Kass & Wasserman,
1995). To further secure our decision in selecting the best model,
we used the adjusted likelihood ratio test (LMR-LRT; Lo, Men-
dell, & Rubin, 2001) and the bootstrap likelihood ratio test (BLRT;
McLachlan & Peel, 2000). For a given model, a nonsignificant
LMR-LRT and BLRT test suggests that a model with one fewer
Entropy Index that summarizes the quality of the classification
(Ramaswamy, DeSarbo, Reibstein, & Robinson, 1993) on the
basis of pertinent theory.
Results
We estimated the means, correlations, and distribution statistics
for the total sample. The standard deviations and distribution
statistics (i.e., skewness and kurtosis) were not large for ability
self-concept and subjective task value scales, suggesting that the
identification of mixture distributions would not be a direct result
of severe nonnormal distributions. All correlations were in the
expected directions. Further, correlations between the covariates
were all small (around .10), suggesting no multicollinearity prob-
lem.
We tested up to 10 different models. In all models, an intercept
and both linear and quadratic slopes were specified. Time-
invariant covariates were also included in the models and allowed
to influence latent growth parameters (i.e., intercept and slope) as
well as the categorical latent variable (i.e., class membership). The
means of these growth factors were the only parameters allowed to
vary across classes. Correlations between temporally adjacent re-
siduals measurement variances were also estimated.
On the basis of the different fit statistics (see Table 2), we opted
for a seven-class model. Although the information criteria (LLR,
AIC, BIC, ABIC) consistently decreased as the number of classes
increased, this decrease stabilized at the five-class solution. While
testing the BLRT for models from three to 10 classes, results were
inconclusive ( p .05 in every model). We thus decided not to rely
on this test. Otherwise, the LMR-LRT favored a seven-class
model. The Entropy Index of .77 for this model indicated adequate
quality of classification (i.e., not too many classification errors).
The classification of the final model is presented in Table 3. This
classification table is based on the posterior probabilities of being
classified in each latent class and suggests that the seven-class
model provides an acceptable classification, although two latent
classes showed average probabilities slightly lower than .80 diag-
onally. On the basis of these indicators, but also because this
 TRAJECTORIES OF SELF-CONCEPT AND VALUE IN LITERACY 809

Table 2 Students from the three declining trajectoriesEarly Decline

Summary of Fit Statistics for Different Growth Mixture Models (7.8%), Constant Decline (28.1%), and Late Decline (13.3%)
trajectoriesall evidenced substantial motivational declines over
No. of class LLR AIC BIC ABIC years. The main difference between these three groups lies in the
1 11,904 23,929 24,198 24,008 proportion of boys and girls at the moment when their motivational
2 11,445 23,035 23,363 23,131 declines started. Children in the Early Decline and Late Decline
3 11,343 22,859 23,244 22,971 trajectories were mostly boys (70.6%, ASR 3.2; 69.0%, ASR
4 11,262 22,723 23,167 22,852 3.6, respectively); in contrast, there was a slight majority of girls
5 9,565 19,343 19,818 19,481
in the Constant Decline group (57.5%, ASR 2.2). For the
6 9,536 19,309 19,838 19,464
7 9,513 19,286 19,869 19,456 Constant Decline and the Early Decline children, the decrease in
8 9,488 19,259 19,896 19,445 ability self-concept and subjective task value began in the second
9 9,456 19,221 19,911 19,423 grade. Yet, by the sixth grade, the Early Decliners reported a
10 9,303 19,193 19,874 19,397 motivation-level far below the average, whereas the decrease
Note. LLR log-likelihood ratio; AIC Akaike information criterion; showed by the Constant Decliners was slower. Finally, the Late
BIC Bayesian information criterion; ABIC sample-size adjusted Decliners followed a similar trend but in later grades. During the
Bayesian information criterion. transition to secondary school, these children reported high sub-
jective task value and ability self-concept. In the ninth grade, their
motivation was still around the average; however, by the 12th
seven-class solution was theoretically more interesting because it grade, their motivation especially their ability self-concept be-
included the Late Decline Trajectory, we opted for this model. liefswas quite low.
In the last group, the Low Trajectory, children are characterized
The Course of Childrens Literacy Subjective Task by weak perception of their reading abilities and low valuing of
literacy from the very beginning of school entry. They presented
Value and Ability Self-Concept
the lowest scores until ninth grade, when their motivation shifted
Figure 1 illustrates the seven trajectories of literacy subjective task and increased slightly. This Low group represented 2.2% of our
value and ability self-concept. This figure presents child mean pre- sample, mostly boys (63.6%, ASR 2.6).
dicted scores on the two dimensions across Grades 112. Overall,
contrary to our first hypothesis, none of the groupsneither children Predictors of the Different Trajectories
in the highest trajectory nor children in the lowest trajectory
reported stable patterns of ability self-concept and subjective task Membership in each school motivation trajectory was regressed
value. Further, this figure illustrates three declining trajectories instead on a series of theoretically meaningful covariates measured in
of one, and, in contrast to our second hypothesis, children of all early schooling. Because the dependent variable consists of an
groups except the Moderate trajectory reported similar changes on unordered categorical latent variable, the log-odds of motivation
their ability self-concept and subjective task value over time. trajectory membership must be calculated relative to a reference
The first group illustrated in Figure 1, the High Trajectory, is group. Because children in the High Trajectory maintained the
composed of children presenting the most positive profile. Of all highest subjective task value and ability self-concept over all
groups, these children maintained the highest level of literacy years, we used this group as the reference.
subjective task value and ability self-concepts over years. None- The predictive effects of the covariates are reported as odds
theless, they showed a small motivational decrease from Grade 1 ratios (with 95% confidence intervals) in Table 4. These odds
to Grade 7. This group represents only 10% of the sample and, ratios represent the change in odds of being in one trajectory
unsurprisingly, is composed of a majority of girls (66.7%, adjusted compared with the High Trajectory when the value of the predictor
standardized residual [ASR]1 2.7). increases by one unit. Compared with children from the High
In the next group, the Moderate Trajectory (20.3%), students group, the odds of being a boy was 2.16 3.31 times higher for
motivation fluctuated around the mean. Of all trajectories, it is the children in the Transitory, Early, Late Decline, and Low Trajec-
only one in which child ability self-concept and subjective task tory. Compared with children from the High group, the Early,
value do not follow similar patterns of changes over time. In the Late, and Constant Decliners were also more likely to live in a
first years of schooling, members of the Moderate Trajectory lower income family. Further, except for the Late Decliners,
reported higher literacy subjective task value than ability self- youths of all other trajectories earned lower grades in their first
beliefs. Yet, up to Grade 9, their literacy subjective task value years of schooling than the High group. Finally, the odds of having
decreased consistently, whereas their ability perception remained a parent reporting a low perception of his/her childs reading
stable. By the end of 12th grade, both boys and girls (girls ability was higher in the Early Decline group.
52.6%, ASR 0.05) from this group reported mean-level mo-
tivation on the two dimensions. Discussion
Children in the Transitory Decline group (18.3%) reported the In this study, we assessed the joint development of child ability
same mean level of ability self-concept and subjective task value in self-concept and subjective task value in the domain of literacy
Grade 1. Their motivation decreased until the transition to junior high
school when their scores on both scales began to increase. Over all the
years, these children, mostly boys (73.3%, ASR 5.8), maintained 1
ASRs are interpreted as Z scores (e.g., ASR values 1.96 are signif-
higher ability self-concepts than subjective task value. icant at the two-tailed .05 level).
810 ARCHAMBAULT, ECCLES, AND VIDA

Table 3
Classification Table Based on Estimated Posterior Probabilities for the Final Model

Class High Moderate Transitory decline Constant decline Early decline Late decline Low

High .892 .020 .018 .052 .000 .018 .000

Moderate .031 .779 .076 .0816 .013 .012 .009
Transitory decline .018 .063 .808 .049 .049 .000 .013
Constant decline .072 .049 .046 .774 .045 .014 .000
Early decline .000 .011 .060 .054 .861 .000 .015
Late decline .091 .007 .000 .059 .000 .843 .000
Low .000 .019 .045 .000 .038 .000 .899

from the beginning of elementary school through the end of high
school. We identified seven groups of children with distinct tra-
jectories of change over the school years. We further demonstrated
that a confluence of risk factors predicted these different pathways,
suggesting that school-based interventions aimed at promoting
child motivation beliefs should be concerned with individual dif-
ferences over development.
Trajectories of Ability Self-Concept and Subjective
Task Value
On the basis of research that has examined child and adolescent
motivation and engagement patterns over time (Baker & Wigfield,
1999; Jacobs et al., 2002; Janosz et al., 2008), we expected to
identify at least three distinct trajectories of literacy beliefs: stable

High Trajectory Moderate Trajectory Transitory Declining Trajectory

2
2 (10.0%) 2
(20.3%) (18.3%)

1,5 1,5
1,5

1 1
1

0,5 0,5
0,5

0 0 0

-0,5 -0,5 -0,5

-1 -1 -1

-1,5 -1,5 -1,5

-2 -2 -2
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12

2 2 Low Trajectory
2 Early Decline Trajectory 2 Constant Decline Trajectory Late Decline Trajectory
(2.2%)
(7.8%) (28.1%) (13.3%)
1,5 1,5
1,5 1,5

1 1
1 1

0,5 0,5
0,5 0,5

0 0 0
0

-0,5 -0,5 -0,5 -0,5

-1 -1 -1 -1

-1,5 -1,5 -1,5 -1,5

-2 -2 -2 -2
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12

Legend Self-Concept of Ability Subjective Task Value

Figure 1. Mean predicted value and ability self-concept scores by profile.

 TRAJECTORIES OF SELF-CONCEPT AND VALUE IN LITERACY 811

and high, stable and low, and decreasing. We further assumed that

[1.26, 8.68]
[0.56, 1.13]

[0.50, 1.26]
[0.38, 0.96]

[0.35, 1.24]
95% CI
for most children, a decline in competency beliefs would precede
a decline in subjective task value. Overall, these hypotheses were

Low
partially supported. However, contrary to expectations, we found
much more between-child variation than expected in both compe-
tence beliefs and value of literacy. Moreover, of all the groups

3.31

0.62
OR

0.79

0.79

0.66
identified in this study, we found three groups in which child
motivation declined from Grade 1 to Grade 12 and none in which
[1.06, 3.44] children maintained stable beliefs across grades.
[0.48, 0.83]

[0.50, 1.13]
[0.85, 1.62]

[0.46, 1.95]
Most research on group differences in the development of
95% CI

engagement and motivation have identified youths presenting pos-

Late decline

itive and stable attitudes over time. Consistent with those findings

Note. All covariates are significant predictors based on the chi-square test from the total model. The reference group is the high trajectory. OR odds ratio.
(Archambault et al., 2009; Janosz et al., 2008; Ratelle et al., 2004),
we identified by the end of 12th grade a High Trajectory group in
0.64

which students reported high competency beliefs about reading

OR

2.25

0.75
1.17

0.95

and English and positive task value about the usefulness and
importance of this domain for their future. Unlike those in the
general population (Jacobs et al., 2002), these childrens scores
[1.30, 5.14]
[0.32, 0.94]

[0.71, 1.48]
[0.43, 0.87]

[1.05, 6.81]
95% CI

were relatively stable over years.

Results of the Multinomial Logistic Regression: Predictors of Value and Ability Self-Concept Trajectories Membership

Early decline

Researchers have suggested two different but related explana-

tions for this observed small motivational decline in highly moti-
vated children. For one, classroom changes in terms of increased
control by teachers, social comparisons between peers, and focus
2.58

0.61
OR

0.55

2.59

on individual differences in performance become more salient in

1.02
High versus

later elementary school (Deci, Ryan, & Williams, 1996; Eccles,

Wigfield, Midgley, et al., 1993). Second, changes in child cogni-
tive ability makes social comparison more common. As a result,
[0.37, 1.66]
[0.53, 0.85]

[0.61, 1.18]
[0.57, 0.97]

[0.86, 2.34]
95% CI

children are likely to become less motivated in domains in which

Constant decline

they feel less competent (Harter, 1992; Ruble et al., 1980; Stipek
& Mac Iver, 1989). Both of these age-related changes can result in
motivated children becoming more aware of themselves and oth-
ers, which in turn can lead them to process and evaluate the
0.69
OR

0.77
0.79

0.85

1.42

feedback they receive differently. At school entry, most children

like school, participate in activities, and feel competent in the
academic tasks they initiate. During this time, the focus in the
[1.20, 3.88]
[0.65, 1.03]

[0.77, 1.28]
[0.44, 0.88]

[0.77, 1.78]

classroom is more on task mastery than on relative performance

95% CI
Transitory decline

(Freedman-Doan et al., 2000; Patrick, Mantzicopoulos, Samara-

pungavan, & French, 2008). However, as they get older and are
confronted with an increasing number of academic activities, a
greater focus on relative performance levels, and a sharp increase
0.63

in testing, children begin to develop a better sense of which

2.16
OR

0.82

1.00

1.17

activities they do or do not enjoy. Hence, they become better able

to provide sensitive and objective evaluations of themselves, their
interests, and their abilities in different domains. For children from
[0.68, 7.47]
[0.66, 1.12]

[0.79, 1.21]
[0.44, 0.85]

[0.71, 2.28]

the High Trajectory, these new insights only lead to a relative

95% CI

decline in their competency beliefs and the value they attach to

Moderate

school subjects. However, this is not the case for most of the
children: 90% of our children experienced substantial declines in
p .10. p .05. p .01.

their literacy motivation across their primary and secondary school

0.61

years.
OR

2.25
0.86

0.98

1.28

Children in declining trajectories evidenced the sharpest drop in

motivation. As expected, such pathways represented half the chil-
dren in our sample. In the Early, Constant, and Late Decline
of reading ability
Parental perception

trajectories, children evidenced motivation decreases, but the rate

Time 1 predictor

learning efforts

of this decline varied across the school years. All of these children
Reading grades
Family income
Gender (boys)

Reading and

started their first grade year with high motivational beliefs. Most
reported that reading was useful for them and that believed they
Table 4

were good at it. Yet, over time, their beliefs about reading became
more negative. This decline is particularly worrisome for Early

812 ARCHAMBAULT, ECCLES, AND VIDA
Decliners and Constant Decliners because the changes that they
experienced began in early schooling. This premature loss of
confidence and decrease in their valuing of literacy are detrimental
for sustaining engagement in learning. For this group, in the
domain of reading and writing there is a risk of this trend becom-
ing firmly embedded and thus difficult to change. As they disen-
gage from learning, the likelihood that they will acquire the basic
reading and writing requirements necessary to succeed in other
school subjects is also likely to decline, which in turn may lead to
an increase in both their rate of disengagement from school and the
likelihood that they will eventually drop out of school before
graduating (Janosz et al., 2008). Although we did not investigate
the consequences associated with being on these declining trajec-
tories in this article, we hope to do so in the future.
We also identified groups of children who reported both a
decrease and an increase of ability self-concept and subjective task
value at different time periods of their academic experience. Con-
sistent with our expectations, we identified a Low Trajectory group
in which children reported low motivational beliefs from their
early grades. Yet, unexpectedly, the motivational profile of these
children was not stable. It decreased up to the end of elementary
school and then went back to its initial level. A similar pattern was
observed in the Transitory Decline group. The perceptions of
children from this group changed dramatically in eighth grade:
From that point, they felt increasingly better about their ability in
reading and writing and assigned growing importance to these
tasks. This change likely reflects either individual changes or
contextual changes in their experiences in English classrooms. In
American schooling, students experience transition to high school
at this time, and curriculum shifts from a focus on reading and
writing to a focus on literature. Perhaps these changes helped spark
the interest of a subset of our participants. These two trajectories
highlight the importance of investigating the circumstances under
which adolescents can experience motivational increases as they
move into secondary school. A better understanding of these
circumstances could enhance our ability to design effective inter-
vention strategies for children with early signs of low literacy
motivation. Such work would also contribute to our understanding
of the interpersonal and contextual processes underlying develop-
mental shifts in child motivation.
Relations Between Ability Self-Concept and Subjective
Task Value Over Time
In most trajectories, children reported parallel changes in their
feeling of competence and value of reading and writing activities.
Such results were quite surprising. However, they are consistent
with theories proposing that child ability self-concept in a specific
domain directly influences the value given to tasks related to this
domain (Deci & Ryan, 1985; Eccles et al., 1983). Actually, the fact
that both dimensions vary conjointly in most trajectories does not
allow us to disentangle the specific effect of one component on the
other. The changes found on subjective task value could be influ-
enced by changes on ability self-concept (and vice versa) or by a
third variable such as perception of individual progress (Spinath &
Steinmayr, 2008). More research is needed to sort out the effect
that each component has on the other in the development of
specific trajectory of changes.
Children in one group did evidence noticeable differences be-
tween their ratings for subjective task value and ability self-
concept in literacy. In their first years of schooling, children from
this Moderate Trajectory reported lower ability self-concept than
subjective task value. Over time, the importance and usefulness
that they assigned to reading and writing decreased steadily,
whereas their competency beliefs ratings were stable. Around
eighth grade, the two dimensions converged. The early gap that
existed between the two dimensions in this group suggests that
some young children acknowledge the value of reading even
though they do not have great confidence in the reading ability.
Such children may evaluate their abilities with reference to the
ability of peers over time, or they may fail to achieve as well as
they hoped in reading. As a result, these students may begin to feel
less competent and gradually reduce the importance that they
attach to literacy. According to Wigfield and Eccles (1994), such
disengagement could act as a protective mechanism used by chil-
dren to preserve their self-esteem. To prevent such disaffection, it
is important that teachers and practitioners adapt learning activities
to challenges and support student valuing of reading and compe-
tency beliefs.
Predictors Associated With Trajectories of Change
Our findings also highlight important individual and family
predictors associated with child motivational trajectories. Unsur-
prisingly, we demonstrated that student gender and early academic
difficulties remained strong predictors of motivational pathways.
Consistent with previous studies (Baker & Wigfield, 1999; Durik
et al., 2006), we found a higher proportion of girls in the High
Trajectory, in which children maintained the highest and most
stable literacy value and ability self-concepts over time. Con-
versely, the proportion of boys was significantly higher in groups
evidencing a motivational decline, especially during the elemen-
tary school years. Moreover, children who reported a decline in
their competencies beliefs over years demonstrated lower levels of
academic achievement than children who maintained high beliefs
about literacy from early grades on. However, such findings need
to be interpreted while keeping in mind that the relationship
between achievement and motivation is reciprocal (Corpus et al.,
2009; Marsh et al., 1999, 2005; Nurmi & Aunola, 2005). There-
fore, the greater motivational decrease observed among low
achievers might be explained by their motivation prior to the study
instead of by their early achievement.
Children evidencing declines in their literacy motivation were
also more likely to live in low SES families than children who
maintained high levels of ability self-concepts and subjective task
values for reading (Brooks-Gunn & Duncan, 1997). Even though
these children start school with reasonably high ability self-
concepts and attach considerable importance to reading, these
children are less likely to feel competent about their ability to
master reading. One explanation being that disadvantaged families
may be less well equipped to intervene and help children catch up
in their reading competence. Consequently, children from lower
income families are likely to reduce the importance that they attach
to these skills to preserve their self-esteem and to continue having
greater academic difficulties than children from more wealthy
families.
 TRAJECTORIES OF SELF-CONCEPT AND VALUE IN LITERACY
Finally, our finding that parental beliefs were more positive for
children showing the earliest motivational decline and presenting
the greatest number of family risk factors than parental beliefs for
children evidencing the highest motivational level was quite un-
expected. According to Eccles et al. (1983), parental perception of
child ability should influence the development of their child be-
liefs children whose parents hold high estimates of their reading
ability should develop high reading ability self-concepts. Evi-
dence, in general, supports this prediction at the population level.
One potential explanation for our results might be that parents of
the high risk group of children are less educated and acquainted
with the demands of the educational system. Therefore, they might
have an unrealistic perception of their childs actual difficulties
and abilities. If so, this would further exacerbate the dynamics
noted previously: If these parents are not aware that their child
needs help in learning to read, they will not be motivated to seek
out the supplementary assistance that their child might need. Our
results do not allow us to go such in depth in our conclusions, but
more research could help clarify this idea and improve our under-
standing of the family role on childrens motivation over time.
Implications
This study has several implications for research and practice.
First, although some practices used by teachers in many upper
elementary and secondary schools have been criticized for their
harmful consequences on student motivation (Brophy & Evertson,
1978; Midgley, Feldlaufer, & Eccles, 1988), the practices used
with early elementary school children have received less attention.
Considering that 10% of the children in our study reported rela-
tively low reading ability and task value by the third grade, and
that nearly one third reported similarly low self- and task beliefs by
the end of elementary school situation, it is important to identify
how the practices used by elementary school teachers can influ-
ence these patterns. This knowledge could help researchers and
professionals in the development of effective strategies promoting
childrens self-concept of ability and value of literacy.
This study also brings new light on the heterogeneity that exists
in the development of literacy-related beliefs. Although the vast
majority of children evidence motivational declines as they pass
through primary and secondary school, the extent and timing of
such changes vary between individuals. It remains unclear whether
these differences reflect specific factors that operate at different
points in time for different children and what these factors might
be. Moreover, there is certainly a need to understand how children
in the most problematic declining groups compare with each other
and whether characteristics measured later in the academic expe-
rience could influence the nature and course of their motivation.
However, what is clear is that prevention and intervention efforts
need to be developed with these individual differences in mind. A
strategy designed for children who already evidence low motiva-
tional believes in the first grade might not be appropriate for
adolescents whose motivational beliefs decline as they move into
secondary school. Further, different types of interventions might
be needed depending on whether parents view their childs ability
as unrealistically high or realistically low. Finally, different inter-
ventions may be needed for boys and girls. For example, as boys
spent less time reading for pleasure (Baker & Wigfield, 1999;
Coles & Christine, 2002), the kind of books teachers propose may
813
need to be adapted to better stimulate their interest. An important
challenge is to bring about changes in curriculum and the daily
practices of teachers that are likely to help them respond to the
differential needs of students.
Limitations
This study has several limitations that are worth mentioning.
First, the sample was composed primarily of children coming from
middle-class European American families. Thus, our findings can-
not be generalized to other populations. Future work should ad-
dress this limit and study the development of childrens literacy
ability self-concept and subjective task value with more diverse
populations of children. A second limitation concerns the factors
that were included in our model to predict trajectory membership.
Because GMM models only allow the inclusion of a limited
number of factors, we selected only a restricted subset of individ-
ual and family-related characteristics as covariates. Although those
characteristics have been shown to influence child motivation, we
omitted some individual (e.g., classroom attention and behaviors),
family (e.g., number of books available at home, parental invest-
ment in school), teacher (e.g., teacherstudent relationships,
teacher practices), and school (e.g., school neighborhood, children
classroom ratio) characteristics that could have been related to
child motivation trajectories. Finally, the characteristics that we
selected were only assessed once at school entry. Therefore, we
were not able to determine whether these characteristics were
stable and whether changes could have influenced child motivation
variations over time.
Conclusions
By assessing different trajectories of ability self-concept and
subjective task value in the literacy domain, this study demon-
strates that children do not all evidence the same trajectories of
change over time in their motivational beliefs. Because literacy
skills represent the performance standard for most academic sub-
ject matters, findings from this research highlight the necessity of
investing financial resources for screening children who present
early risks of following negative literacy motivational pathways
over time (namely boys), who come from low SES families, and
who present reading difficulties and important motivational de-
clines in the first two years of schooling. These children should be
the targets for interventions that promote intrinsic motivation and
competence for reading while they still perceive themselves as
having good abilities and believe reading and writing is important.
References
Aikens, N. L., & Barbarin, O. (2008). Socioeconomic differences in
reading trajectories: The contribution of family, neighborhood, and
school contexts. Journal of Educational Psychology, 100, 235251.
doi:10.1037/0022-0663.100.2.235
Akaike, H. (1987). Factor analysis and AIC. Psychometrika, 52, 317332.
doi:10.1007/BF02294359
Ames, C. (1992). Classrooms: Goals, structures, and student motivation.
Journal of Educational Psychology, 84, 261271. doi:10.1037/0022-
0663.84.3.261
Anderman, E. M., Eccles, J. S., Yoon, K. S., Roeser, R., Wigfield, A., &
Blumenfeld, P. (2001). Learning to value mathematics and reading:
814 ARCHAMBAULT, ECCLES, AND VIDA
Relations to mastery and performance-oriented instructional practices.
Contemporary Educational Psychology, 26, 76 95. doi:10.1006/
ceps.1999.1043
Archambault, I., Janosz, M., Morizot, J., & Pagani, L. (2009). Adolescent
behavioral, affective, and cognitive engagement in school: Relationship
to dropout. Journal of School Health, 79, 408 415. doi:10.1111/j.1746-
1561.2009.00428.x
Assor, A., Kaplan, H., & Roth, G. (2002). Choice is good, but relevance is
excellent: Autonomy-enhancing and suppressing teacher behaviours pre-
dicting students engagement in schoolwork. British Journal of Educa-
tional Psychology, 72, 261278. doi:10.1348/000709902158883
Baker, L., & Wigfield, A. (1999). Dimensions of childrens motivation for
reading and their relations to reading activity and reading achievement.
Reading Research Quarterly, 34, 452 477. doi:10.1598/RRQ.34.4.4
Bradley, R. H., & Corwyn, R. F. (2002). Socioeconomic status and child
development. Annual Review of Psychology, 53, 371399. doi:10.1146/
annurev.psych.53.100901.135233
Brooks-Gunn, J., & Duncan, G. J. (1997). The effects of poverty on
children. The Future of Children, 7, 5571. doi:10.2307/1602387
Brophy, J. E., & Evertson, C. M. (1978). Context variables in teaching.
Educational Psychologist, 12, 310 316. doi:10.1080/00461527809529185
Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models.
Newbury Park, CA: Sage.
Chapman, J. W., & Tunmer, W. E. (1995). Development of young chil-
drens reading self-concepts: An examination of emerging subcompo-
nents and their relationship with reading achievement. Journal of Edu-
cational Psychology, 87, 154 167. doi:10.1037/0022-0663.87.1.154
Chapman, J. W., Tunmer, W. E., & Prochnow, J. E. (2000). Early reading-
related skills and performance, reading self-concept, and the develop-
ment of academic self-concept: A longitudinal study. Journal of Edu-
cational Psychology, 92, 703708. doi:10.1037/0022-0663.92.4.703
Cicchetti, D., & Rogosch, F. A. (1996). Equifinality and multifinality in
developmental psychopathology. Development and Psychopathology, 8,
597 600. doi:10.1017/S0954579400007318
Coles, M., & Christine, H. (2002). Gendered readings: Learning from
childrens reading choices. Journal of Research in Reading, 25, 96 108.
doi:10.1111/1467-9817.00161
Corpus, J. H., McClintic-Gilbert, M. S., & Hayenga, A. O. (2009). Within-year
changes in childrens intrinsic and extrinsic motivational orientations: Con-
textual predictors and academic outcomes. Contemporary Educational Psy-
chology, 34, 154 166. doi:10.1016/j.cedpsych.2009.01.001
Covington, M. V. (1992). Making the grade: A self-worth perspective on
motivation and school reform. New York, NY: Cambridge University
Press.
Covington, M. V. (1998). The will to learn. New York, NY: Cambridge
University Press.
Cunningham, A. E., & Stanovich, K. E. (1997). Early reading acquisition
and its relation to reading experience and ability 10 years later. Devel-
opmental Psychology, 33, 934 945. doi:10.1037/0012-1649.33.6.934
Deci, E. L., & Ryan, A. M. (1985). Intrinsic motivation and self-
determination in human behavior. New York, NY: Plenum Press.
Deci, E. L., Ryan, R. M., & Williams, G. C. (1996). Need satisfaction and
the self-regulation of learning. Learning and Individual Differences, 8,
165183. doi:10.1016/S1041-6080(96)90013-8
Deci, E. L., Vallerand, R. J., Pelletier, L. G., & Ryan, R. M. (1991).
Motivation and education: The self-determination perspective. Educa-
tional Psychologist, 26, 325346. doi:10.1207/s15326985ep2603&4_6
Duncan, G. J., & Brooks-Gunn, J. (1997). Income effects across the
life-span: Integration and interpretation. In G. J. Duncan & J. Brooks-
Gunn (Eds.), Consequences of growing up poor (pp. 596 610). New
York, NY: Russell Sage Foundation.
Durik, A. M., Vida, M., & Eccles, J. S. (2006). Task values and ability
beliefs as predictors of high school literacy choices: A developmental
analysis. Journal of Educational Psychology, 98, 382393. doi:10.1037/
0022-0663.98.2.382
Eccles, J. (1984). Sex differences in achievement patterns. In T. Sondereg-
ger (Ed.), Nebraska symposium on motivation: Psychology and gender
(Vol. 32, pp. 97132). Lincoln, NE: University of Nebraska Press.
Eccles, J. S., Adler, T. F., Futterman, R., Goff, S. B., Kaczala, C. M.,
Meece, J. L., & Midgley, C. (1983). Expectancies, values, and academic
behaviors. In J. T. Spence (Ed.), Achievement and achievement motiva-
tion (pp. 75146). San Francisco, CA: Freeman.
Eccles, J. S., Flanagan, C., Lord, S., Midgley, C., Roeser, R., & Yee, D.
(1996). Schools, families, and early adolescents: What are we doing
wrong and what can we do instead? Journal of Developmental &
Behavioral Pediatrics, 17, 267276.
Eccles, J. S., Lord, S., & Midgley, C. (1991). What are we doing to early
adolescents? The impact of educational contexts on early adolescents.
American Journal of Education, 99, 521542. doi:10.1086/443996
Eccles, J. S., Midgley, C., Wigfield, A., Buchanan, C. M., Reuman, D.,
Flanagan, C., & Mac Iver, D. (1993). Development during adolescence:
The impact of stage-environment fit on young adolescents experiences
in schools and in families. American Psychologist, 48, 90 101. doi:
10.1037/0003-066X.48.2.90
Eccles, J. S., & Wigfield, A. (1995). In the mind of the actor: The structure
of adolescents achievement task values and expectancy-related beliefs.
Personality and Social Psychology Bulletin, 21, 215225. doi:10.1177/
0146167295213003
Eccles, J. S., Wigfield, A., Flanagan, C. A., Miller, C., Reuman, D. A., &
Yee, D. (1989). Self-concepts, domain values, and self-esteem: Rela-
tions and changes at early adolescence. Journal of Personality, 57,
283310. doi:10.1111/j.1467-6494.1989.tb00484.x
Eccles, J., Wigfield, A., Harold, R. D., & Blumenfeld, P. (1993). Age and
gender differences in childrens self- and task perceptions during ele-
mentary school. Child Development, 64, 830 847. doi:10.2307/
1131221
Eccles, J. S., Wigfield, A., Midgley, C., Reuman, D., Mac Iver, D., &
Feldlaufer, H. (1993). Negative effects of traditional middle schools on
students motivation. The Elementary School Journal, 93, 553574.
doi:10.1086/461740
Eccles, J. S., Wigfield, A., & Schiefele, U. (1998). Motivation to succeed.
Hoboken, NJ: Wiley.
Evans, G. W. (2004). The environment of childhood poverty. American
Psychologist, 59, 7792. doi:10.1037/0003-066X.59.2.77
Feldlaufer, H., Midgley, C., & Eccles, J. S. (1988). Student, teacher, and
observer perceptions of the classroom environment before and after the
transition to junior high school. Journal of Early Adolescence, 8, 133
156. doi:10.1177/0272431688082003
Fredricks, J. A., & Eccles, J. S. (2002). Childrens competence and value
beliefs from childhood through adolescence: Growth trajectories in two
male-sex-typed domains. Developmental Psychology, 38, 519 533. doi:
10.1037/0012-1649.38.4.519
Freedman-Doan, C., Wigfield, A., Eccles, J. S., Blumenfeld, P., Arbreton,
A., & Harold, R. D. (2000). What am I best at? Grade and gender
differences in childrens beliefs about ability improvement. Journal of
Applied Developmental Psychology, 21, 379 402. doi:10.1016/S0193-
3973(00)00046-0
Guthrie, J. T., Hoa, L. W., Wigfield, A., Tonks, S. M., & Perencevich,
K. C. (2006). From spark to fire: Can situational reading interest lead to
long-term reading motivation? Reading Research and Instruction, 45,
91117.
Guthrie, J. T., McRae, A., & Klauda, S. L. (2007). Contributions of
concept-oriented reading instruction to knowledge about interventions
for motivations in reading. Educational Psychologist, 42, 237250.
Guthrie, J. T., Schafer, W. D., & Hutchinson, S. R. (1991). Relations of
document literacy and prose literacy to occupational and societal char-
 TRAJECTORIES OF SELF-CONCEPT AND VALUE IN LITERACY
acteristics of young Black and White adults. Reading Research Quar-
terly, 26, 30 48. doi:10.2307/747730
Guthrie, J. T., Wigfield, A., Humenick, N. M., Perencevich, K. C., Taboada,
A., & Barbosa, P. (2006). Influences of stimulating tasks on reading moti-
vation and comprehension. Journal of Educational Research, 99, 232246.
doi:10.3200/JOER.99.4.232-246
Guthrie, J. T., Wigfield, A., & VonSecker, C. (2000). Effects of integrated
instruction on motivation and strategy use in reading. Journal of Edu-
cational Psychology, 92, 331341. doi:10.1037/0022-0663.92.2.331
Harter, S. (1982). The perceived competence scale for children. Child
Development, 53, 8797. doi:10.2307/1129640
Harter, S. (1992). The relationship between perceived competence, affect,
and motivational orientation within the classroom: Processes and pat-
terns of change. In R. K. Boggiano & T. S. Pittman (Eds.), Achievement
and motivation (pp. 77114). New York, NY: Cambridge University
Press.
Harter, S., & Connell, J. P. (1984). A model of the relationship among
childrens academic achievement and their self-perceptions of compe-
tence, control, and motivational orientations. In J. Nicholls (Ed.), The
development of achievement motivation (pp. 219 250). Greenwich, CT:
JAI.
Higgins, E. T., & Parsons, J. E. (1983). Social cognition and the social life
of the child: Stages as sub-cultures. In E. T. Higgins, D. N. Ruble, &
W. W. Hartup (Eds.), Social cognition and social development (pp.
15 62). Cambridge, MA: Cambridge University Press.
Jacobs, J. E., & Eccles, J. S. (1992). The impact of mothers gender-role
stereotypic beliefs on mothers and childrens ability perceptions. Jour-
nal of Personality and Social Psychology, 63, 932944. doi:10.1037/
0022-3514.63.6.932
Jacobs, J. E., & Eccles, J. S. (2000). Parents, task values, and real-life
achievement-related choices. San Diego, CA: Academic Press.
Jacobs, J. E., Lanza, S., Osgood, D., Eccles, J. S., & Wigfield, A. (2002).
Changes in childrens self-competence and values: Gender and domain
differences across grades one though twelve. Child Development, 73,
509 527. doi:10.1111/1467-8624.00421
Janosz, M., Archambault, I., Morizot, J., & Pagani, L. (2008). School
engagement trajectories and their differential predictive relations to
dropout. Journal of Social Issues, 64, 21 40. doi:10.1111/j.1540-
4560.2008.00546.x
Kass, R. E., & Wasserman, L. (1995). A reference Bayesian test for nested
hypotheses and its relationship to the Schwarz criterion. Journal of the
American Statistical Association, 90, 928 934. doi:10.2307/2291327
Lo, Y., Mendell, N. R., & Rubin, D. B. (2001). Testing the number of
components in a normal mixture. Biometrika, 88, 767778. doi:10.1093/
biomet/88.3.767
Madden, N. A., Slavin, R. E., Karweit, N. L., Dolan, L. J., & Wasik, B. A.
(1993). Success for all: Longitudinal effects of a restructuring program
for inner-city elementary schools. American Educational Research Jour-
nal, 30, 123148.
Marsh, H. W. (1989). Age and sex effects in multiple dimensions of
self-concept: Preadolescence to early adulthood. Journal of Educational
Psychology, 81, 417 430. doi:10.1037/0022-0663.81.3.417
Marsh, H. W., Byrnes, B. M., & Yeung, A. S. (1999). Causal ordering of
academic self-concept and achievement: Reanalysis of a pioneering
study and revised recommendations. Educational Psychologist, 34, 155
167. doi:10.1207/s15326985ep3403_2
Marsh, H. W., Trautwein, U., Ludtke, O., Koller, O., & Baumert, J. (2005).
Academic self-concept, interest, grades, and standardized test scores:
Reciprocal effects models of causal ordering. Child Development, 76,
397 416. doi:10.1111/j.1467-8624.2005.00853.x
McLachlan, G. J., & Peel, D. (2000). Finite mixture models. New York,
NY: Wiley.
Midgley, C., Feldlaufer, H., & Eccles, J. S. (1988). The transition to junior
815
high school: Beliefs of pre- and posttransition teachers. Journal of Youth
and Adolescence, 17, 543562. doi:10.1007/BF01537831
Muthen, B. O. (2001). Second-generation structural equation modeling
with a combination of categorical and continuous latent variables: New
opportunities for latent class/latent growth modeling. In L. M. Collins &
A. Sayer (Eds.), New methods for the analysis of change (pp. 291322).
Washington, DC: American Psychological Association. doi:10.1037/
10409-010
Muthen, B. O. (2003). Statistical and substantive checking in growth
mixture modeling: Comment on Bauer and Curran (2003). Psychologi-
cal Methods, 8, 369 377. doi:10.1037/1082-989X.8.3.369
Muthen, B. O. (2004). Latent variables analysis: Growth mixture modeling
and related techniques for longitudinal data. In D. Kaplan (Ed.), Hand-
book of quantitative methodology for the social sciences (pp. 345368).
Newbury Park, CA: Sage.
Muthen, L. K., & Muthen, B. O. (2006). Mplus users guide (4th ed.). Los
Angeles, CA: Author.
Nagin, D. S. (1999). Analyzing developmental trajectories: A semipara-
metric, group-based approach. Psychological Methods, 4, 139 157.
doi:10.1037/1082-989X.4.2.139
Nicholls, J. G., & Miller, A. T. (1984). Reasoning about the ability of self
and others: A developmental study. Child Development, 55, 1990 1999.
doi:10.2307/1129774
Nurmi, J.-E., & Aunola, K. (2005). Task-motivation during the first school
years: A person-oriented approach to longitudinal data. Learning and
Instruction, 15, 103122. doi:10.1016/j.learninstruc.2005.04.009
Parsons, J. E., Adler, T. F., & Kaczala, C. M. (1982). Socialization of
achievement attitudes and beliefs: Parental influences. Child Develop-
ment, 53, 310 321. doi:10.2307/1128973
Parsons, J. E., & Ruble, D. N. (1977). The development of achievement-
related expectancies. Child Development, 48, 10751079. doi:10.2307/
1128364
Patrick, H., Mantzicopoulos, P., Samarapungavan, A., & French, B. F.
(2008). Patterns of young childrens motivation for science and teacher
child relationships. Journal of Experimental Education, 76, 121144.
doi:10.3200/JEXE.76.2.121-144
Pintrich, P. R., & Schunk, D. H. (1996). Motivation in education: Theory,
research, and applications. Englewood Cliffs, NJ: Prentice Hall.
Ramaswamy, V., DeSarbo, W., Reibstein, D., & Robinson, W. (1993). An
empirical pooling approach for estimating marketing mix elasticities
with PIMS data. Marketing Science, 12, 103124. doi:10.1287/
mksc.12.1.103
Ratelle, C. F., Guay, F., Larose, S., & Senecal, C. (2004). Family correlates
of trajectories of academic motivation during a school transition: A
semiparametric group-based approach. Journal of Educational Psychol-
ogy, 96, 743754. doi:10.1037/0022-0663.96.4.743
Raudenbush, S. W., & Kasim, R. M. (1998). Cognitive skill and economic
inequality: Findings from the National Adult Literacy Survey. Harvard
Educational Review, 68, 3379.
Rimm-Kaufman, S. E., Pianta, R. C., Cox, M. J., & Bradley, R. H. (2003).
Teacher-rated family involvement and childrens social and academic
outcomes in kindergarten. Early Education and Development, 14, 179
198. doi:10.1207/s15566935eed1402_3
Roeser, R. W., Eccles, J. S., & Sameroff, A. J. (1998). Academic and
emotional functioning in adolescence: Longitudinal relations, patterns,
and prediction by experiences in middle school. Development and Psy-
chopathology, 10, 321352. doi:10.1017/S0954579498001631
Ruble, D., Boggiano, A., Feldman, N., & Lobel, J. (1980). A developmen-
tal analysis of the role of social comparison in self-evaluation. Devel-
opmental Psychology, 16, 105115. doi:10.1037/0012-1649.16.2.105
Sameroff, A. J., & Fiese, B. H. (1990). Transactional regulation and early
intervention. In S. J. Meisels & J. P. Shonkoff (Eds.), Early intervention:
A handbook of theory, practice, and analysis (pp. 119 149). New York,
NY: Cambridge University Press.
816 ARCHAMBAULT, ECCLES, AND VIDA

Schafer, J. L. (1999). NORM: Multiple imputation of incomplete multi-
variate data under a normal model (Version 2) [Computer software for
Windows 95/98/NT]. University Park, PA: Pennsylvania State Univer-
sity, Department of Statistics.
Schiefele, U. (1999). Interest and learning from text. Scientific Studies of
Reading, 3, 257279. doi:10.1207/s1532799xssr0303_4
Schunk, D. H., & Zimmerman, B. J. (1997). Social origins of self-
regulatory competence. Educational Psychologist, 32, 195208. doi:
10.1207/s15326985ep3204_1
Schwarz, G. (1978). Estimating the dimension of a model. Annals of
Statistics, 6, 461 464. doi:10.1214/aos/1176344136
Sclove, L. S. (1987). Application of model-selection criteria to some
problems in multivariate analysis. Psychometrika, 52, 333343. doi:
10.1007/BF02294360
Skaalvik, E. M., & Hagtvet, K. A. (1990). Academic achievement and
self-concept: An analysis of causal predominance in a developmental
perspective. Journal of Personality and Social Psychology, 58, 292307.
doi:10.1037/0022-3514.58.2.292
Spinath, B., & Spinath, F. M. (2005). Longitudinal analysis of the link
between learning motivation and competence beliefs among elementary
school children. Learning and Instruction, 15, 87102. doi:10.1016/
j.learninstruc.2005.04.008
Spinath, B., & Steinmayr, R. (2008). Longitudinal analysis of intrinsic
motivation and competence beliefs: Is there a relation over time? Child
Development, 79, 15551569. doi:10.1111/j.1467-8624.2008.01205.x
Stipek, D. J. (1984). Sex differences in childrens attributions for success
and failure on math and spelling tests. Sex Roles, 11, 969 981. doi:
10.1007/BF00288127
Stipek, D., & Mac Iver, D. (1989). Developmental change in childrens
assessment of intellectual competence. Child Development, 60, 521
538. doi:10.2307/1130719
Sweet, A. P., Guthrie, J. T., & Ng, M. M. (1998). Teacher perceptions and
student reading motivation. Journal of Educational Psychology, 90,
210 223. doi:10.1037/0022-0663.90.2.210
Whitehurst, G. J., & Lonigan, C. J. (1998). Child development and emer-
gent literacy. Child Development, 69, 848 872.
Wigfield, A., & Eccles, J. S. (1994). Childrens competence beliefs,
achievement values, and general self-esteem: Change across elementary
and middle school. Journal of Early Adolescence, 14, 107138. doi:
10.1177/027243169401400203
Wigfield, A., & Eccles, J. S. (2002). Students motivation during the
middle school years. San Diego, CA: Academic Press.
Wigfield, A., Eccles, J. S., Mac Iver, D., Reuman, D. A., & Midgley, C.
(1991). Transitions during early adolescence: Changes in childrens
domain-specific self-perceptions and general self-esteem across the tran-
sition to junior high school. Developmental Psychology, 27, 552565.
doi:10.1037/0012-1649.27.4.552
Wigfield, A., Eccles, J. S., Yoon, K. S., Harold, R. D., Arbreton, A. J.,
Freedman-Doan, C., & Blumenfeld, P. C. (1997). Change in childrens
competence beliefs and subjective task values across the elementary
school years: A 3-year study. Journal of Educational Psychology, 89,
451 469. doi:10.1037/0022-0663.89.3.451
Wigfield, A., & Guthrie, J. T. (1997). Relations of childrens motivation for
reading to the amount and breadth or their reading. Journal of Educational
Psychology, 89, 420 432. doi:10.1037/0022-0663.89.3.420
Willett, J. B., & Sayer, A. S. (1996). Cross-domain analyses of change over
time: Combining growth modeling and covariance structure analysis. In
R. Schumacher & G. Marcoulides (Eds.), Advanced structural equation
modeling: Issues and techniques (pp. 125157). Mahwah, NJ: Erlbaum.
Wilson, K. M., & Trainin, G. (2007). First-grade students motivation and
achievement for reading, writing, and spelling. Reading Psychology, 28,
257282. doi:10.1080/02702710601186464
Yeung, W. J., Linver, M. R., & Brooks-Gunn, J. (2002). How money
matters for young childrens development: Parental investment and
family processes. Child Development, 73, 18611879. doi:10.1111/
1467-8624.t01-1-00511
Received June 1, 2009
Revision received July 26, 2010
Accepted August 4, 2010

E-Mail Notification of Your Latest Issue Online!

Would you like to know when the next issue of your favorite APA journal will be available
online? This service is now available to you. Sign up at http://notify.apa.org/ and you will be
notified by e-mail when issues of interest to you become available!