Expectancy – Value Theory of Achievement Motivation

Text-only Preview

Contemporary Educational Psychology 25, 68–81 (2000)
doi:10.1006/ceps.1999.1015, available online at http://www.idealibrary.com on
Expectancy–Value Theory of Achievement Motivation
Allan Wigfield
University of Maryland
Jacquelynne S. Eccles
University of Michigan
We discuss the expectancy–value theory of motivation, focusing on an expec-
tancy–value model developed and researched by Eccles, Wigfield, and their col-
leagues. Definitions of crucial constructs in the model, including ability beliefs,
expectancies for success, and the components of subjective task values, are pro-
vided. These definitions are compared to those of related constructs, including self-
efficacy, intrinsic and extrinsic motivation, and interest. Research is reviewed
dealing with two issues: (1) change in children’s and adolescents’ ability beliefs,
expectancies for success, and subjective values, and (2) relations of children’s and
adolescents’ ability-expectancy beliefs and subjective task values to their perfor-
mance and choice of activities.
© 2000 Academic Press
Achievement motivation theorists attempt to explain people’s choice of
achievement tasks, persistence on those tasks, vigor in carrying them out,
and performance on them (Eccles, Wigfield, & Schiefele, 1998; Pintrich &
Schunk, 1996). As discussed by Murphy and Alexander (this issue), there
are a variety of constructs posited by motivation theorists to explain how
motivation influences choice, persistence, and performance. One long-stand-
ing perspective on motivation is expectancy–value theory. Theorists in this
tradition argue that individuals’ choice, persistence, and performance can be
explained by their beliefs about how well they will do on the activity and
the extent to which they value the activity (Atkinson, 1957; Eccles et al.,
1983; Wigfield, 1994; Wigfield & Eccles, 1992). In this article we discuss
the nature of the expectancy and value constructs, how they develop, and
Portions of this paper were presented at the 1998 meeting of the American Educational
Research Association in a symposium entitled ‘‘A Motivated Look at Motivation Terminol-
ogy’’ (Patricia Alexander, Organizer). The writing of this article was supported in part by
Grant HD-17553 from the National Institute of Child Health and Human Development.
Address reprint requests to Dr. Allan Wigfield, Department of Human Development, Uni-
versity of Maryland, College of Education, College Park, MD 20742.
0361-475X/00 $35.00
Copyright © 2000 by Academic Press
All rights of reproduction in any form reserved.

FIG. 1.
Eccles, Wigfield, and colleagues’ expectancy–value model of achievement moti-
how they relate to children’s and adolescents’ performance and choice. We
focus on the expectancy–value model developed and assessed by Eccles,
Wigfield, and their colleagues (Eccles, 1984; Eccles et al., 1983; Wigfield,
1994; Wigfield & Eccles, 1992).
Eccles et al. (1983) proposed an expectancy–value model of achievement
performance and choice and studied it initially in the mathematics achieve-
ment domain. The most recent statement of this model is presented in Fig.
1; the overall model is presented to provide a sense of its scope. We focus
in this article on a portion of the model; specifically, the constructs contained
in the expectancies and subjective task values boxes, along with some of the
constructs in the box containing goals and self-schemata. As can be seen in
the figure, expectancies and values are assumed to influence directly achieve-
ment choices. They also influence performance, effort, and persistence. Ex-
pectancies and values are assumed to be influenced by task-specific beliefs
such as ability beliefs, the perceived difficulty of different tasks, and individ-
uals’ goals, self-schema, and affective memories. These social cognitive vari-
ables, in turn, are influenced by individuals’ perceptions of their own previ-
ous experiences and a variety of socialization influences (see Eccles et al.,
1983, Eccles et al., 1998, and Wigfield & Eccles 1992 for discussion of these

Items Used to Assess Children’s Ability Beliefs and Subjective Task Values
Ability Beliefs Items
1. How good in math are you? (not at all good
very good)
2. If you were to list all the students in your class from the worst to the best in math,
where would you put yourself ? (one of the worst
one of the best)
3. Some kids are better in one subject than in another. For example, you might be better
in math than in reading. Compared to most of your other school subjects, how good
are you in math? (a lot worse in math than in other subjects
a lot better in math
than in other subjects)
Expectancy Items
4. How well do you expect to do in math this year? (not at all well
very well)
5. How good would you be at learning something new in math? (not at all good
Usefulness, Importance, and Interest Items
1. Some things that you learn in school help you do things better outside of class, that
is, they are useful. For example, learning about plants might help you grow a garden.
In general, how useful is what you learn in math? (not at all useful
very useful)
2. Compared to most of your other activities, how useful is what you learn in math?
(not at all useful
very useful)
3. For me, being good in math is (not at all important
very important)
4. Compared to most of your other activities, how important is it for you to be good at
math? (not at all important
very important)
5. In general, I find working on math assignments (very boring
very interesting [fun])
6. How much do you like doing math? (not at all
very much)
Defining the constructs in models of motivation is an important goal of
this special issue. Eccles et al. (1983) defined and measured expectancies
for success as children’s beliefs about how well they will do on upcoming
tasks, either in the immediate or longer term future. Ability beliefs are de-
fined as the individual’s perception of his or her current competence at a
given activity. Ability beliefs thus are distinguished conceptually from ex-
pectancies for success, with ability beliefs focused on present ability and
expectancies focused on the future. However, empirically these constructs
are highly related (see Eccles & Wigfield, 1995; Eccles et al., 1993; this
issue is discussed in more detail later). The items we use to measure these
constructs appear in Table 1; items from the mathematics domain are used
as examples. As can be seen we primarily have measured these two beliefs
at the domain-specific level, although we occasionally have asked about spe-
cific activities within different domains.
These definitions can be compared to those of related constructs in the
literature. Bandura (1997) included expectancies in his discussion of self-
efficacy. He distinguished between efficacy expectations, or the individual’s
belief that he or she can accomplish a task, and outcome expectancies, or

the belief that a given action will lead to a given outcome (see also Pajares,
1996). He argued that expectancy–value theorists historically have focused
on outcome expectations in their models, and stated further that efficacy
expectations are more predictive of performance and choice than are out-
come expectations. We would not argue with his claim that efficacy expecta-
tions are more strongly predictive of performance and choice. However, we
would argue that in our work we have measured individuals’ own expecta-
tions for success, rather than their outcome expectations. Thus our expec-
tancy construct is more similar to Bandura’s efficacy expectation construct
than it is to the outcome expectancy construct.
Beliefs about one’s ability play a prominent role in different motivation
theories. A complete review of these theories is outside the scope of this
article (see Eccles et al., 1998 and Pintrich & Schunk, 1996, for reviews),
but we briefly discuss definitions of the ability construct in various motiva-
tion and self-concept theories. In his attribution theory, Weiner (1985) pro-
posed that individuals viewed ability as a relatively stable characteristic over
which they had little control. He argued that attributions made to ability
(and lack of ability) have important motivational consequences. Attributing
success to ability has positive motivational consequences, whereas attribut-
ing failure to lack of ability has negative consequences. Covington (1992)
also focused on individuals’ ability beliefs in his self-worth model, arguing
that individuals attempt to maintain a positive sense of ability in order to
preserve their self-worth. Like Weiner, Covington focused on perceived abil-
ity as a relatively stable capacity. However, based on developmental work
on children’s understanding of the ability construct (e.g., Nicholls, 1978,
1990), he noted developmental differences in individuals’ conceptions of
ability. In their self-determination theory, Deci, Ryan, and their colleagues
(e.g., Deci & Ryan, 1985; Ryan, 1992) included the need for competence
as a basic need that individuals have and discussed how this need is a major
reason why people seek out optimal stimulation and challenging activities.
Finally, self-concept researchers often focus on beliefs about how good one
is at different activities as a crucial aspect of self-concept. The measures
these researchers developed include many items assessing individuals’ be-
liefs about their ability in different areas (e.g., Harter, 1982, 1990; Marsh,
Researchers often measure ability-related beliefs in somewhat different
ways. One crucial difference among measures is the level of specificity of
measurement. Bandura (1997) argued that efficacy should be measured spe-
cifically because specific measures of beliefs relate more closely to behavior.
Pajares (1996), in an article comparing self-efficacy with related constructs,
noted that efficacy most often has been measured at the task-specific level.
These measures typically ask individuals about how confident they are they
can accomplish the task. However, at times Bandura has measured self-effi-

cacy rather generally. For instance, Bandura, Barbaranelli, Caprara, and Past-
orelli (1996) measured individuals’ efficacy for different academic subjects
and then combined them into an overall measure of academic efficacy. As
we noted above, our measures of ability beliefs and expectancies have tended
to be domain rather than activity specific, and so our approach has been
somewhat more general than much of Bandura’s research and that of other
researchers studying self-efficacy (e.g., Schunk, 1983).
Another crucial issue is how individuals are asked to judge their abilities.
As seen in Table 1, we ask individuals to rate their individual abilities (‘‘how
good are you’’) and also ask them to compare their abilities across different
subject areas and to other individuals. Self-concept researchers such as
Harter (1990) and Marsh (1989) have not asked the comparative questions,
focusing instead on questions about how good the individual thinks she is
and how well or poorly she can do different activities. Self-efficacy research-
ers also tend to focus on individuals’ beliefs about how confident they are
they can complete different tasks rather than asking them to compare their
efficacy to that of others (Bandura, 1997; Pajares, 1996).
In sum, ability and expectancy beliefs are crucial to the expectancy–value
theory of motivation and are present in other major theories as well. The
definition of these constructs varies some across theoretical perspectives.
Measures of these beliefs also vary across theory, especially with respect to
their specificity and exactly what aspects of ability are asked about. An im-
portant implication of these differences is that when researchers choose mea-
sures for future work on ability-related beliefs, they should carefully consider
how specific they want their measures to be and which aspects of perceived
ability they are most interested in measuring. An important task remaining
for future research is to examine more closely how similar and different these
various measures are.
Turning to the achievement values portion of the model, Eccles et al.
(1983) defined different components of achievement values: attainment
value or importance, intrinsic value, utility value or usefulness of the task,
and cost (see Eccles et al., 1983, and Wigfield & Eccles, 1992, for more
detailed discussion of these components). Building on Battle’s (1965, 1966)
work, Eccles et al. defined attainment value as the importance of doing well
on a given task. Intrinsic value is the enjoyment one gains from doing the
task. When individuals do tasks that are intrinsically valued, there are impor-
tant psychological consequences for them, most of which are quite positive
(see Deci & Ryan, 1985 for further discussion) Utility value or usefulness
refers to how a task fits into an individual’s future plans, for instance, taking
a math class to fulfill a requirement for a science degree. Cost refers to how
the decision to engage in one activity (e.g., doing schoolwork) limits access
to other activities (e.g., calling friends), assessments of how much effort
will be taken to accomplish the activity, and its emotional cost. Most of our

empirical work has focused on the first three of these value constructs, so
we limit our discussion to them.
Other motivation researchers have assessed constructs related to the intrin-
sic and utility value constructs. Interest value is a construct similar to the
construct of intrinsic motivation as defined by Deci and his colleagues
(Deci & Ryan, 1985; Deci, Vallerand, Pelletier, & Ryan, 1991) and by Harter
(1981) because it concerns doing a task out of interest and enjoyment. This
construct also bears some relation to the construct of interest as discussed
by researchers such as Alexander, Kulikowich, and Jetton (1994), Renninger,
Hidi, and Krapp (1992), and Schiefele (1996). Utility value captures more
‘‘extrinsic’’ reasons for engaging in a task, such as doing a task not for its
own sake but to reach some desired end state. This construct thus can be
tied to the construct of extrinsic motivation (see Deci & Ryan, 1985; Harter,
1981 for further discussion of extrinsic motivation). Although there poten-
tially is some overlap in these constructs, it is important to point out that
the values constructs and constructs of intrinsic and extrinsic value and inter-
est come from distinct theoretical perspectives and so have different intellec-
tual roots.
We now have done a number of studies of the development of children’s
and adolescents’ ability beliefs, expectancies for success, and subjective val-
ues. In this section we discuss our findings regarding (1) how children’s
expectancies for success, ability beliefs, and subjective values change across
the school years; and (2) how these beliefs and values relate to children’s
performance and activity choice.
We have addressed these issues in three major longitudinal studies. The
first was a longitudinal study focused on gender differences in achievement
beliefs and values about mathematics and English.1 The sample consisted of
5th through 12th graders who completed questionnaires once each year over
a 2-year period (see Eccles et al., 1983; Eccles & Wigfield, 1995; Meece,
Wigfield, & Eccles, 1990). The second was a study of how the transition
from elementary to junior high school influenced children’s beliefs and val-
ues about different academic subjects, sports, and social activities (see Eccles
et al., 1989; Wigfield et al., 1991). The sample consisted of children in 6th
1 Gender differences in children’s and adolescents’ achievement beliefs and values have
been a central focus of our work. Space limitations preclude detailed discussion of these differ-
ences. In general, our results show that boys and girls’ beliefs and values differ in gender
stereotypic ways, beginning as early as 1st grade. Interested readers can see Eccles et al.
(1993), Eccles et al. (1989), Wigfield et al. (1991), and Wigfield et al. (1997) for discussion
of gender differences in children’s ability-related beliefs and values.

grade in elementary school who subsequently made a transition to junior
high school in 7th grade. The third is a 10-year longitudinal study of how
children’s achievement beliefs and values change through the elementary
and secondary school years (see Eccles et al., 1993; Wigfield et al., 1997).
This study began with a group of children in 1st, 2nd, and 4th grades and
followed them until they graduated from high school. In each study children
completed questionnaires assessing their ability beliefs, expectancies for suc-
cess, and subjective valuing of different activities, along with a variety of
other constructs. Most of the children participating in the studies were Euro-
pean–American, and they came from lower middle class to middle class
How Children’s Expectancies for Success, Ability Beliefs, and Subjective
Values Change across the School Years
We focus in this article on findings regarding two kinds of change: (1)
change in the structure of children’s ability-related beliefs and values; and
(2) mean level change in the level of children’s and adolescents’ ability-
related beliefs and values. In the model expectancies for success, ability be-
liefs, and the different aspects of task values are proposed to be separate
constructs. When studying young children, however, it is reasonable to ask
if these constructs indeed are distinct in children’s minds. It is important to
establish these distinctions before examining mean-level change.
The structure of children’s ability-related beliefs and achievement values.
In her discussion of children’s self-concept development, Harter (1983) dis-
cussed how children first have broad understandings that they are ‘‘smart’’
or ‘‘dumb’’ and later develop a more fine-grained sense of competence for
specific activities. Our work on this issue has focused on the differentiation
of children’s competence and expectancy beliefs and subjective task values.
To address this issue we analyzed data from the first and third study, using
confirmatory factor analyses. Confirmatory factor analysis (CFA) allows the
researcher to test theoretically derived hypotheses about the structure of a
set of variables and allows for the explicit comparison of different alternative
Several major findings have emerged from this work. First, across child-
hood and adolescence children’s and adolescents’ ability beliefs and expec-
tancies for success consistently loaded together in our CFAs (see Eccles et
al., 1993; Eccles & Wigfield, 1995). These theoretically distinct constructs
do not appear to be empirically distinguishable, at least as we have mea-
sured them. However, children’s and adolescents’ ability–expectancy be-
liefs are domain specific. For example, Eccles et al. (1993) assessed chil-
dren’s ability beliefs and expectancies for success in the domains of math,
reading, music, and sports. The CFAs indicated that children’s beliefs in
each domain formed distinct factors and that each of these factors was char-

acterized by the items measuring ability beliefs and expectancies for success
in the domain. This domain differentiation occurred even for the 1st-grade
children in the study. Marsh, Craven, and Debus (1991) also found that
young children’s ability self-concepts were distinct for different achieve-
ment domains.
Further, young children also seem to distinguish between their ability-
related beliefs and subjective task values. Eccles et al. (1993) found that
within the domains of math, reading, music, and sports, children’s ability–
expectancy beliefs and subjective values formed clearly distinct factors. This
distinction was apparent even in the factor structure within each domain for
the 1st graders. This finding is a crucial one for the expectancy–value model.
Even during the very early elementary grades children appear to have distinct
beliefs about what they are good at and what they value in different achieve-
ment domains.
Can the different aspects of subjective values proposed by Eccles et al.
(1983) be empirically identified? Eccles and Wigfield (1995), analyzing data
from the first study which included early adolescents and adolescents, exam-
ined the attainment, interest, and utility value aspects proposed by Eccles et
al. (1983) that could be distinguished empirically in the mathematics domain.
The CFAs showed that the three task values factors were distinguished
clearly, confirming the theoretical distinctions in the Eccles et al. model.
Comparative tests of the factor structure of the younger (5th through 7th)
and older (8th through 12) grade students showed no differences, indicating
that the distinctions were made by the youngest as well as by the oldest
students. Wigfield et al. (1992) found that during the early elementary school
years children’s subjective values are less differentiated, with two factors
(interest and utility-importance) emerging in the CFAs of children’s re-
sponses to items in the math, reading, and sports domains.
In sum, even young children’s ability-related beliefs are differentiated
clearly across various activities, although within a given activity ability be-
liefs and expectancies for success factor together. Different components of
subjective values also have been empirically identified, especially in children
in 5th grade and above. The most important implication of the work just
reviewed for the expectancy–value model (see Fig. 1) is that early in the
elementary school years certain of the constructs proposed in the model have
been shown to be distinct (e.g., ability-related beliefs, subjective values).
However, the ability beliefs and expectancy constructs, while theoretically
distinct, are highly related empirically.
Mean-level change in children’s achievement beliefs and subjective val-
ues. Various researchers have found in cross-sectional studies that younger
children have more positive achievement-related beliefs than do older chil-
dren (see Stipek & Mac Iver, 1989, for review). We have looked at this
question longitudinally and across different school transitions.

We first consider ability-related beliefs. In cross-sectional analyses of the
Study 3 data, Eccles et al. (1993) reported linear decreases in children’s
ability-related beliefs across the elementary school years, particularly in the
academic achievement domains. Recently, in a 3-year longitudinal follow-up
to the Eccles et al. (1993) study, Wigfield et al. (1997) showed that children’s
ability-related beliefs for math, reading, instrumental music, and sports de-
clined across the elementary school years. Analyses of the data from Study
2, the junior high transition study, showed that these declines often continue
into junior high or middle school and that the largest changes occurred imme-
diately after the junior high transition (Eccles et al., 1989; Wigfield et al.,
1991). Recent work with the 10-year longitudinal data from Study 3 show
that these declines continue across the high school years (Jacobs, Hyatt, Ec-
cles, Osgood, & Wigfield, 1999).
It is interesting to compare these findings to work on the development of
self-efficacy. Shell, Colvin, and Bruning (1995) reported that children’s self-
efficacy for reading and writing was higher among 7th- and 10th-grade stu-
dents than among 4th-grade students. They measured self-efficacy by asking
students how sure they were that they could do reading and writing tasks of
increasing difficulty. Such efficacy beliefs could increase, even if children’s
sense of how able they were compared to other students their age decreased,
as we have found in our work.
Children’s subjective values also decline, although these declines vary
across domain. Eccles et al. (1993), analyzing Study 3 data, found that older
elementary school-aged children valued math, reading, and instrumental mu-
sic less than younger children did. Interestingly, children’s valuing of sports
activities was higher among older elementary school-aged children. In a lon-
gitudinal follow-up to these cross-sectional analyses, Wigfield et al. (1997)
looked at changes in children’s beliefs about the usefulness and importance
of these activities and their interest in them. Children’s beliefs about the
usefulness and importance of math, reading, instrumental music, and sports
activities decreased over the 3 years of the study. However, only children’s
interest in reading and instrumental music decreased over time; their interest
in math and sports did not. This finding shows the importance of examining
the separate aspects of children’s subjective valuing of different activities.
Eccles et al. (1989) and Wigfield et al. (1991) found that across the transi-
tion to junior high school, children’s ratings of both the importance of math
and English and their liking of these school subjects decreased. In math,
students’ importance ratings continued to decline across 7th grade, whereas
their importance ratings of English increased somewhat during 7th grade.
During high school, however, adolescents’ valuing of some activities be-
comes more positive. Eccles et al. (1983) and Wigfield (1984), analyzing
Study 1 data, found that in late elementary school children valued math more

highly than did high school students. In contrast, the older students valued
English more.
In summary, children’s ability-related beliefs and values become more
negative in many ways as they get older, at least through early adolescence.2
Children believe they are less competent in many activities and often value
those activities less. These differences are more pronounced in certain activ-
ity areas. The negative changes in children’s achievement-related beliefs and
value have been explained in two major ways. One explanation is that chil-
dren become much better at understanding and interpreting the evaluative
feedback they receive and engage in more social comparison with their peers.
As a result of these processes many children become more accurate or realis-
tic in their self-assessments, so that their beliefs become relatively more neg-
ative (see Stipek & Mac Iver, 1989, for thorough discussion of how chil-
dren’s processing of evaluative information changes). A second explanation
is that the school environment changes in ways that makes evaluation more
salient and competition between students more likely, thus lowering some
children’s achievement beliefs (e.g., see Stipek, 1996; Wigfield, Eccles, &
Pintrich, 1996, for detailed discussion of these explanations).
How Children’s and Adolescents’ Ability-Related Beliefs and Subjective
Task Values Predict Performance and Choice
This issue has been addressed in analyses of the data from Studies 1 and
3. In the first study, two fundamental findings emerged from our analyses
looking at relations between children’s competence beliefs and values and
their performance and choice. First, even when previous performance is con-
trolled, children’s beliefs about their ability and expectancies for success
are the strongest predictors of subsequent grades in math, predicting those
outcomes more strongly than either previous grades or achievement values.
Second, children’s subjective task values are the strongest predictors of chil-
dren’s intentions to keep taking math and actual decisions to do so.
To illustrate these findings, in Study I we assessed how 7th through 9th
graders’ competence beliefs and performance in math Year 1 predicted their
achievement beliefs and performance Year 2 (see Meece et al., 1990, for
full details). The Year 2 achievement beliefs included in this model were
expectancies for success, perceived importance of math, math anxiety, and
intentions to continue taking more math. Meece et al. found (using structural
equation modeling techniques) that Year 1 ability beliefs directly and posi-
2 The general pattern is for children to have optimistic beliefs and values in the early grades,
which decline across the school years. It is important to point out that some children doubt
their abilities quite early on (see Burhans & Dweck, 1995; Heyman, Dweck, & Cain, 1992
for research documenting this point).

Document Outline

  • FIGURE 1
  • TABLE 1