ArticlePDF Available

Abstract and Figures

The authors explored different aspects of children's reading motivation and how children's motivation related to the amount and breadth of their reading. The reading motives assessed included self-efficacy, intrinsic–extrinsic motivation and goals, and social aspects. Fourth- and 5th-grade children ( N = 105) completed a new reading motivation questionnaire twice during a school year. Children's reading amount and breadth were measured using diaries and questionnaires. Children's reading motivation was found to be multidimensional. Their motivation predicted children's reading amount and breadth even when previous amount and breadth were controlled. An intrinsic motivation composite predicted amount and breadth of reading more strongly than did an extrinsic motivation composite. Some aspects of girls' reading motivation were more positive than boys'. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Content may be subject to copyright.
Journal of Educational Psychology
1997,
Vol.
89, No. 3,420-432Copyright 1997 by the American Psychological Association, Inc.
0022-0663/97/53.00
Relations of Children's Motivation for Reading
to the Amount and Breadth of Their Reading
Allan Wigfield and John
T.
Guthrie
University of Maryland College Park
The authors explored different aspects of children's reading motivation and how children's
motivation related to the amount and breadth of their reading. The reading motives assessed
included self-efficacy, intrinsic-extrinsic motivation and goals, and social aspects. Fourth- and
5th-grade children (N = 105) completed a new reading motivation questionnaire twice during
a school year. Children's reading amount and breadth were measured using diaries and
questionnaires. Children's reading motivation was found to be multidimensional. Their
motivation predicted children's reading amount and breadth even when previous amount and
breadth were controlled. An intrinsic motivation composite predicted amount and breadth of
reading more strongly than did an extrinsic motivation composite. Some aspects of girls'
reading motivation were more positive than boys'.
Students' reading amount and breadth contribute substan-
tially to several valued aspects of their achievement and
performance, such as reading achievement, world knowl-
edge,
and participation in society. Anderson, Wilson, and
Fielding (1988) found that the amount of independent
out-of-school reading accounted for 16% of the variance in
the reading comprehension of fifth graders, after general
activity levels were controlled. Similarly, Stanovich and
Cunningham (1992) found that amount and breadth of
reading predicted reading achievement, as indicated by
standardized vocabulary tests, even after previous general
intelligence was controlled. Both Anderson et al. and
Cipielewski and Stanovich (1992) found that the amount of
reading predicted growth of reading achievement during
elementary school on different measures of reading compre-
hension (Cipielewski & Stanovich, 1992).
Children's reading amount and breadth contribute to their
knowledge of the world, including knowledge of informa-
tion such as the size of the National Aeronautics and Space
Administration budget, who were allies in World War
II,
and
the distinctions among various religious beliefs (Stanovich
& Cunningham, 1993). Individuals who read frequently also
Allan Wigfield and John T. Guthrie, Department of Human
Development, University of Maryland College Park.
Portions of this paper were reported at the 1994 meeting of the
American Educational Research Association, New Orleans, Louisi-
ana. The work reported herein is a National Reading Research
Project of the University of Georgia and University of Maryland. It
was supported under the Educational Research and Development
Centers Program (PR/AWARD 117A2007) as administered by the
Office of Educational Research and Improvement, U. S. Depart-
ment of Education. The findings and opinions expressed here do
not necessarily reflect the position or policies of the National
Reading Research Center, the Office of Educational Research and
Improvement, or the U. S. Department of Education-
Correspondence concerning this article should be addressed to
Allan Wigfield, Department of Human Development, University of
Maryland, College Park, Maryland 20742. Electronic mail may be
sent via Internet to aw44@umail.umd.edu.
participate more in their communities. Guthrie* Schafer, and
Hutchinson (1991), using a national database, found that
amount of reading predicted participation in community
organizations, after home background and level of schooling
were controlled.
We thus were interested in developing a theoretical and
empirical account of amount and breadth of reading, and
focused on motivation for reading as an important contribu-
tor
to
amount and breadth of reading.
We
took
a
motivational
focus because motivation determines why individuals do (or
do not) choose to do different activities (see Eccles, Wig-
field, & Schiefele, in press). Because researchers still do not
know a great deal about the nature of motivation specifically
for reading, our first task was to conceptualize and then
measure reading motivation. We began with the notion that
there are a variety of motives relevant to engaging in reading
activity (see Baker, Afflerbach, & Reinking, 1996; Guthrie,
McGough, Bennett, & Rice, 1996; Oldfather & Wigfield,
1996).
We examined the motivation literature to find con-
structs pertinent to engaging in reading activity. We then
developed a questionnaire to measure these aspects of
motivation, and assessed their relations to the amount and
breadth of children's reading.
There are many different theories of motivation and
different motivational constructs within them (for recent
reviews, see Eccles et al., in press; Pintrich & Schunk,
1996).
In this study we adapted to reading three sets of
constructs currently prominent in the motivation field. First
are individuals' beliefs about their efficacy to achieve.
Second are constructs dealing with the purposes individuals
have for doing different tasks; these constructs include
valuing of achievement, intrinsic and extrinsic motivation,
and goals for achievement. Third are social aspects of
motivation.
Self-Efficacy Beliefs
Many researchers interested in motivation focus on stu-
dents'
sense of efficacy and beliefs about their ability (e.g.,
420
CHILDREN'S MOTIVATION FOR READING421
Bandura, 1977; Eccles et al., 1983; Nicholls, 1990). AbiUty
beliefs are children's evaluations of their competence in
different areas. Bandura (1977) defined self-efficacy as a
generative capacity where different subskills are organized
into courses of
action.
He proposed that
individuals*
efficacy
expectations for different achievement tasks are a major
determinant of activity choice, willingness to expend effort,
and persistence. Schunk and his colleagues demonstrated
that children's sense of efficacy relates to their academic
performance, and that training students both to be more
efficacious and to believe they are more efficacious im-
proves children's achievement in different subject areas such
as math and reading (see Schunk, 1991; Schunk & Zimmer-
man, 1997). An important implication of this work for
motivation for reading is that when children believe they are
competent and efficacious at reading they should be more
likely to engage in reading.
Intrinsic and Extrinsic Motivation
and Goals for Learning
Motivation researchers also have studied different pur-
poses children have for doing various activities. Relevant
constructs include children's valuing of achievement, intrin-
sic and extrinsic motivation, and achievement goals. These
constructs are crucial to motivation. Even if individuals
believe they are competent and efficacious at an activity they
may not engage in it if they have no purpose for doing so.
Eccles et al. (1983) defined different components of task
values, including interest value (defined as how much the
individual likes the activity), attainment value (defined as
the importance of the activity), and utility value (the
usefulness of an activity). Eccles and her colleagues found
that students' ability beliefs and expectancies for success
predict their performance in mathematics and English,
whereas their subjective task values predict both intentions
and actual decisions to keep taking mathematics and En-
glish, even when previous performance is controlled (Eccles
et al., 1983; Meece, Wigfield, & Eccles, 1990; see Wigfield
& Eccles, 1992, for review).
A construct related to the interest value component is
intrinsic motivation. Intrinsic motivation refers to choosing
to do and then doing an activity for its own sake, rather than
for "extrinsic" reasons such as receiving recognition or
grades (see Deci & Ryan, 1985). One aspect of intrinsic
motivation is becoming totally involved in the activity one is
doing. Many readers have experienced what Csikszentmi-
halyi (1978) described as the flow experience, losing track of
time and self-awareness when becoming completely in-
volved in an activity such as reading a book. Nell (1988)
referred to this state as an "intense and highly energized
state of concentrated attention" (p. 263), which he believed
many individuals seek to obtain. An important implication
of these theorists' work for reading is that readers* engage-
ment in reading will be greatly facilitated when they are
intrinsically motivated to read.
Achievement goals is another construct related to the
purposes children have for achievement (see Ames, 1992;
Dweck & Leggett, 1988; Nicholls, Cheung, Lauer, &
Patashnick, 1989). Dweck and Leggett (1988) denned two
major kinds of goal orientations children can have: perfor-
mance goals and learning goals. Individuals with a perfor-
mance goal orientation seek to maximize favorable evalua-
tions of their ability and minimize negative evaluations of
ability. Questions like "Will I look smart?" and "Can I beat
others?" reflect performance goals. In contrast, with a
learning goal orientation, individuals focus on mastering
tasks and increasing competence at different tasks. Ques-
tions such as "How can I do this task?" and "What will I
learn?"
reflect learning goals. Nicholls and his colleagues
described two similar achievement goal orientations; they
used the terms ego-involved goals and task-involved goals
(e.g., Nicholls et al., 1989). Nicholls et al. also discussed
work-avoidant
goals,
which concern students' attempts to do
as little schoolwork as possible. These and other researchers
have demonstrated that children who have learning goals are
more likely to maintain positive motivation in school.
Work-avoidant students are disengaged from school.
Social Aspects of Motivation
Researchers studying motivation have focused primarily
on academic aspects of motivation. Recently, however,
scholars have become interested in social motivation. For
example, Wentzel (1989) found that high and low achievers'
achievement goals differ; high achievers combined strong
social and academic goals in school, whereas lower achiev-
ers focused more on social goals. Wentzel (1991) reviewed
evidence showing that children with prosocial goals and
who were socially responsible in the classroom tended to do
better in school than children who do not have these kinds of
social goals (see also Wentzel, 1996). We therefore included
social aspects of reading in our conceptualization of reading
motivation.
In sum, motivation researchers have shown that when
individuals have positive ability beliefs about an activity and
think they can do the activity efficaciously, value the activity
for intrinsic reasons, and have learning and prosocial goals,
they should do better at the activity and choose to do it more
frequently. However, these researchers have not looked
specifically at whether these predictions apply to individu-
als'
engagement in reading.
Attitudes About Reading and Interest in Reading
Reading researchers have looked primarily at two motiva-
tion-related constructs. The first is children's attitudes to-
ward reading, which are defined generally as individuals'
feelings about reading (see J. E. Alexander & Filler, 1976).
Alexander and Filler stated that these feelings about reading
should influence how much individuals involve themselves
in reading; thus attitudes about reading should relate to
individuals* motivation for reading (see also Matthewson,
1994;
McKenna, 1994, for more specific models of how
individuals' attitudes toward reading influence their reading
engagement).
A second construct addressed by reading researchers is
children's interest in reading (see R A. Alexander, Kulikow-
422WIGFIELD AND GUTHRIE
ich, & Jetton, 1994, for a review of the work on interest's
effects on text comprehension). This construct is related to
the work on intrinsic value and motivation discussed above.
Researchers have looked at how interest affects comprehen-
sion. Schiefele (1996) found that college students who were
interested in the text materials used in the study understood
those materials more deeply than did students less interested
in the materials, even when the students' prior knowledge of
the materials and general intelligence were controlled. In
studies of fifth and sixth graders, Renninger (1992) found
that interest in the materials enhanced comprehension, even
of materials that were quite difficult for the children
(although there were some gender differences in these
patterns). Thus interest in reading appears to be an important
motivational variable influencing different aspects of read-
ing performance.
We formulated several research questions for this study,
on the basis of the work on children's motivation and on the
importance of reading amount and breadth. The first con-
cerns the nature of children's reading motivation: What are
the aspects of children's reading motivation? The second
concerns relations of reading motivation to amount and
breadth of reading: Does children's reading motivation
relate to the amount and breadth of their reading? If indeed
there are different aspects of reading motivation, there likely
are differences (in an absolute sense) in their levels.
Consequently, our third research question was which aspects
of reading motivation do children hold most strongly?
We also investigated some individual differences in
children's reading motivation. These issues are important
because recent research on children's beliefs and values
regarding reading show that, in general, younger students
have more positive beliefs than older students and girls are
more positive about reading than are boys (Eccles, Wigfield.
Harold, & Blumenfeld,
1993;
Gambrell, Codling, & Palmer,
1996;
Marsh, 1989). Thus our final research question
is:
Are
there grade, time, and gender differences in children's
reading motivation?
Method
Sample
One hundred five Grade
4
and 5 children at an elementary school
in
a
mid-Atlantic state participated
in the
study. There were
59
fourth graders and
46
fifth graders; 47 of the children were girls and
58 were boys.
The
children were from mixed socioeconomic
backgrounds
and
were
a
racially
and
ethnically mixed group.
Approximately 70%
of
the children were European American
and
30%
were African American.
In the
school district from which
these children came,
the 1994
median reading comprehension
percentile
on the
Comprehensive Test
of
Basic Skills
was 65 for
Grade
3
and 55
for
Grade 5. A total
of
14%
of
the students receive
free or reduced cost meals
in
school. Permission
to
participate
in
the study
was
obtained from
the
participants themselves
and
their
parents.
Measures
The Motivation
for
Reading Questionnaire (MRQ). Wigfield
and Guthrie (1995) developed the MRQ
to
assess different aspects
of reading motivation.
On the
basis
of a
previous study
in
which
children were interviewed about their motivation
and
observed
in
classrooms during reading instruction (Guthrie, Van Meter,
et al.,
1996)
and our
review
of
motivation theory,
we
identified
11
possible aspects. They
are
grouped into
the
three categories
of
motivation constructs discussed earlier.
The first
two
aspects
of
motivation
are
based
on the
work
on
self-efficacy. These aspects are reading
efficacy,
the belief that one
can be successful at reading, and reading
challenge,
the satisfaction
of mastering
or
assimilating complex ideas
in
text. The next set of
aspects
are
based
in the
work
on
intrinsic-extrinsic motivation,
subjective values, and achievement goals. The intrinsic motivation
and learning goals aspects include reading curiosity,
the
desire
to
learn about
a
particular topic
of
interest
to the
child,
and
reading
involvement,
the
enjoyment
of
experiencing different kinds
of
literary
or
informational texts.
The
notion
of
involvement
in
reading refers
to the
pleasure gained from reading
a
well-written
book
or
article
on a
topic
one
finds interesting.1 Importance
of
reading
is an
aspect taken from Eccles'
and
Wigfield's work
on
subjective task values (Eccles
et al.,
1983; Wigfield
&
Eccles,
1992).
Another aspect concerns what students
say
they do not tike
about reading;
we
called this dimension reading work avoidance.
This aspect relates
to
Nicholls
et
al.*s (1989) work
on
work-
avoidant goals.
Extrinsic motivation
and
performance goals aspects include
competition
in
reading, the desire
to
outperform others
in
reading;
recognition
for
reading,
the
gratification
in
receiving
a
tangible
form
of
recognition
for
success
in
reading; and reading for grades,
the desire
to be
evaluated favorably
by the
teacher. Because
children often read
in
school where they
are
evaluated
and
compared with others, competition, recognition,
and
grades
may
figure prominently in their motivation
for
reading.
The final aspects concern social motivation
for
reading.
One
proposed aspect
is
social reasons
for
reading,
the
process
of
sharing the meanings gained from reading with friends and family;
another
is
compliance, reading because
of an
external goal
or
requirement. These aspects are based on the work on social goals in
the motivation literature (Wentzel, 1996).
We wrote items
to
assess each
of
these aspects, pilot tested
the
items
on a
group
of 10
fourth-grade children,
and
revised some
of
the items. The version
of
the MRQ used in this study contained
82
items,
with
7 or 8
items measuring each
of
the proposed aspects
(with the exception
of
reading importance, which was measured by
Eccles
et
al.'s, 1983, existing two-item scale). The original MRQ,
with items organized into
the
proposed aspects
of
motivation
for
reading,
is
presented in
the
Appendix.
The MRQ was administered twice during the school year,
in
the
fall (October
and
early November)
and
spring (March
and
early
April).
The questionnaire was administered
by
Allan Wigfield and a
graduate assistant to small groups
of
children (10
to
15 per group).
The administration was done
in
the media center
in
the participat-
ing school. Children were told they were going to answer questions
about their reading,
and
that
the
questions
had no
right
or
wrong
answers. Children answered each item
on a 1 to 4
scale, with
answer choices ranging from very different from me to a lot like me.
They were given three practice items before beginning
the
actual
questionnaire. Children were allowed to read the questions on their
1 In previous reports (e.g., Oldfather & Wigfield, 1996; Wigfield
& Guthrie, 1995), reading involvement was called reading topics
aesthetically enjoyed. We now believe the involvement label better
captures this aspect, because
the
items assess chilifaien's involve-
ment with different kinds
of
reading, as described by Schallert and
Reed (1997).
CHILDREN'S MOTIVATION FOR READING423
own;
the
questionnaire administrators were available
to
answer
questions the children
had
about wording
of
the items.
It
took
the
children approximately 15
to
20 min to complete the MRQ.
The Reading Activity Inventory (RAI).
The RAI
(Guthrie,
McGough,
&
Wigfield,
1994) is a
measure
of the
breadth
and
frequency
of
students*
reading. Questions
on
the RAI
ask
students
whether they read during
the
last week different lands
of
reading
material both
in and out of
school (e.g., different kinds
of
books,
newspapers, comics,
as
well
as
books
in
general).
If
the child says
he
or
she read
a
given kind
of
book
in
the last week, he
or
she then
is asked
to
give
its
title.
The
child then
is
asked
to
indicate
how
often he or she reads that kind of book, responding on
a
1
to 4 scale
from almost never to almost every day.
The RAI was administered directly after the MRQ,
by
the same
administrators. A shortened version was used
in
this study, asking
children about
the
following kinds
of
reading materials: comics,
magazines, newspapers, books, mystery books, sports books,
adventure books,
and
nature books.
The
children were told that
they were going
to
answer some questions about what they read
and how often they read
for
fun. They
did one
practice question,
and then completed
the RAI. It
took children
5 to 10 min to
complete the RAI. To gauge the breadth
of
children's book reading
for fun, we created a composite scale
of
the five items asking about
book reading (books, mystery books, sports books, adventure
books,
and
nature books).2 Although there
is no
traditional
reliability
for
this measure,
the
fall
and
spring administrations
of
the measure correlated
.54 (p <
.001), suggesting
a
substantial
level
of
stability
in
the measure.
Out
of
School Reading Amount.
We
obtained
the
number
of
hours each child
in the
study read outside
of
school
for the
1991-1992
and
1992-1993 school years. This information
was
provided
by the
media specialist
in the
school. Children
at the
school participated
in a
special reading program geared toward
increasing how much they read outside
of
school. When students
read
30
hours outside
of
school over
the
course
of a
year their
names
are
placed
on a
large
map
displayed prominently
in the
school.
All
students reading between
30 and 100
hours
get
recognized
at
the
end of
the year
at a
school assembly. They also
receive
a
free paperback book. Students reading 100 hours or more
outside
of
school receive additional books
and
prizes. The 16
top
students received
a
personalized tour
of
the White House, and
the
student reading
the
most received
a
hard back book valued
up to
$20.
Students
at
the participating school volunteered
to
participate
in
the special reading program; during the year
of
the study approxi-
mately 85%
of
the students participated.
All the
students
in
this
study were
in the
program.
All
participating students took home
logs on which parents recorded the number
of
minutes (recorded
in
quarter hour increments)
the
student read each
day
outside
of
school, up to
a
maximum
of
2 hours per day on school nights and
4
hours on Saturday and Sunday. All kinds of reading materials could
be included in the reading logs, as could the time parents or siblings
read to the child. However, homework time did not count as outside
reading
time,
nor did reading
at
school. The parents signed the logs
to ensure their accuracy. The times were recorded
for
all
7
days
of
the week, beginning
in the
middle
of
September
and
ending
in
mid-May. Parents returned
the
logs
to the
school every
2
weeks,
and
the
media specialist recorded them.
A
summary
of
this
information provided
our
measure
of the
amount
of
children's
reading
(we
were
not
given
the
individual biweekly logs).
In the
absence
of
other indicators
of
measurement reliability,
we
corre-
lated
the
diary data from
the 2
years.
The
correlation
was .59
(p
<
.001), indicating substantial stability
for
this indicator.
The average number
of
hours children spent reading during
the
1991-1992 school year
(the
year before
the
study)
was
58.88
(SD
=
54.83),
and in the
1992-1993 school year the average was
73.59 (SD
=
84.14).
We
converted
the
hours
per
school year
figures into minutes
per
day, using
the
8-month time period over
which the special reading program occurred (and assuming 30 days
per month).
For
1991-1992, children's mean minutes read
per
day was 14.72 min per day (SD
=
13.71); the median minutes
per
day was 11.12.
For
1992-1993, children's mean minutes read
per
day was 18.40 (SD
21.03); the median minutes read per day was
12.81.
These means
and
medians
are
higher than
the
means
of 10.1
(median
of 4.6X
minutes
per day of
book reading reported
by
Anderson
et
al. (1988),
and
10.2 (median
of
5.0) minutes per day
reported
by
Allen, Cipielewski,
and
Stanovich (1992)
in
diary
studies. Different measurement procedures likely explain
the
difference.
Our
information
was
obtained
in
15-min intervals,
whereas Anderson
et
al. and Allen
et
al. used 1-min intervals.
We
relied
on
parent report, whereas they used student report. Students
in our study participated in
a
reading incentive program to increase
their reading, whereas students
in
the other studies did not. Any
or
all
of
these differences could have produced
the
higher mean
and
median levels
of
reading
in
our study.
Results
The results presented are organized to correspond to the
research questions: (a) What are the aspects of children's
reading motivation; (b) Does children's reading motivation
relate to the amount and breadth of their reading; (c) Which
aspects of reading motivation do children hold most strongly;
and (d) Are there grade, time, and gender differences in
children's reading motivation?
Aspects of Children's Motivation for Reading
Various analyses were done to determine whether the
proposed aspects of reading motivation could be identified
empirically. First, unit-weighted scales were created for each
of the proposed aspects, by averaging across all the items
assessing each proposed dimension. The internal consis-
tency reliabilities of these scales then were computed, at
both the fall and spring times of measurement. The reliabili-
ties are presented in the top half of Table
1.
The most reliable
scales included Challenge, Curiosity, Involvement, Social,
Competition, and Compliance. The reliabilities for the other
scales were poorer, suggesting that the items proposed to
form these scales in fact did not cohere as well together.
Item-total correlations. Most of the items written to
assess a given aspect correlated moderately to strongly with
the total score on the scale that included all the items
assessing the proposed aspect of reading motivation. How-
ever, in several cases these item-total correlations were less
than .40. This occurred on the following scales in the fall
administration of the questionnaire: Efficacy (one item),
Curiosity (one item), Recognition (two items), Grades (one
item),
Challenge (one item), and Work Avoidance (two
items).
For the spring administration, the following scales
had some items with lower item-total correlations: Efficacy
2
It should be noted that this set of topics actually may favor
boys.
As discussed
below,
however, girls and boys did not differ in
their reading breadth.
424WIGFIELD AND GUTHRIE
Table 1
Reliabilities for the Theoretical and Factor-Based
Reading Motivation Scales
Scale
Theoretical
Efficacy
Challenge
Curiosity
Involvement
Importance
Recognition
Grades
Social
Competition
Compliance
Work Avoidance
Factor-based
Efficacy
Challenge
Curiosity
Involvement
Importance
Recognition
Grades
Social
Competition
Compliance
Work Avoidance
Fall
.51
.66
.68
.77
.59
.57
.63
.77
.77
.71
.40
.63
.68
.70
.72
.59
.69
.59
.78
.75
.62
.44
Spring
.55
.72
.80
.81
.52
.66
.47
.72
.79
.70
.56
.68
.80
.76
.76
.52
.69
.43
.72
.81
.55
.60
(one item), Recognition (one item), Grades (one item),
Challenge (one item), and Work Avoidance (one item). The
items with low item-total correlations at each time of
measurement generally were the same.
Factor analyses. We did factor analyses to assess fur-
ther the different proposed motivation aspects. The items
first were analyzed to see if any were badly skewed. In the
fall administration, five items were badly skewed. Two of
these items were from the Grades scale, and one each from
the Efficacy, Compliance, and Involvement scales. In the
spring administration, the same two items from the Grades
scale were skewed. Additionally, one other item from the
Grades scale, the same Efficacy item, and one Compliance
item were badly skewed. These items were not included in
the factor analyses.
Because the sample size was not large enough to analyze
all the items at once, we did the factor analyses separately on
the items from each individual motivation aspect to deter-
mine whether those items indeed did define that dimension.
This determination was made by examining the factor
loadings for the different items, using a value of .40 to
indicate that an item loaded on a given factor. Many of the
items loaded on the aspects they were proposed to assess;
however, some did not. In the fall administration, the items
loading at less than .40 on the appropriate dimension
included Social (Item 56), Compliance (Item 13), Efficacy
(Items 9, 15, 16, 18, and 58), Curiosity (Items 20, 23, and
45),
Recognition (Items 40, 41, and 67), Grades (Items 18
and 33), Challenge (Items 50 and 66), Competition (Item
76),
and Work Avoidance (Items 8,10, and
71).
In the spring
administration, the items loading less than .40 on the
appropriate aspects included Social (Item 56), Compliance
(Item 73), Efficacy (Items 9, 15, 16, 18, and 58), Recogni-
tion (Items 41 and 67), Grades (Item 33), Challenge (Items
50,63,
and
66),
Competition (Item 75), and Work Avoidance
(Items 8, 53, and 71). Many of these items were ones that
showed poor item-total correlations. Next, these items with
weaker loadings were eliminated, and the factor analyses on
each separate motivation dimension were re-run. All of the
remaining items loaded on the dimension they were pro-
posed to assess.
On the basis of these different analyses, we deleted 27 of
the items from the original list of items; these items are
indicated by an asterisk in the Appendix. We computed
unit-weighted scales from the reduced item set. Reliabilities
for these scales are presented in the bottom half of Table 1.
Several of the new scales were more reliable man the original
theoretical
scales.
These included Efficacy, Challenge, Recogni-
tion, and Work Avoidance (although the reliability for this
scale was still low, especially in the fall). Reliability of the
following scales stayed the same: Curiosity, Importance,
Social, and Competition. The Involvement, Grades, and
Compliance scales showed slight decreases in reliability. We
used the factor-based scales in the subsequent analyses.
We also factor analyzed the fall and spring motivation
scales to determine if there were higher order dimensions of
reading motivation. A three-factor solution was best for both
the fall and spring scales. The factor loadings are presented
in Table 2. The pattern of loadings was relatively similar at
Table 2
Factor Loadings of the Reading Motivation Scales
Scale
Social
Compliance
Efficacy
Curiosity
Involvement
Recognition
Grades
Challenge
Competition
Importance
Work Avoidance
Social
Compliance
Efficacy
Curiosity
Involvement
Recognition
Grades
Challenge
Competition
Importance
Work Avoidance
1
Fall
.81
.78
.62
.48
.81
.41
Spring
.67
.64
.77
.79
.54
.83
-.43
2
.80
.50
.68
.61
.68
.76
.56
.74
.72
3
-.41
.74
.74
.85
.60
Note. Factor loadings less than .40 are not presented.
CHILDREN'S MOTIVATION FOR READING425
the fall and spring times of measurement. At each time of
measurement, the first factor consisted of the following
scales:
Social, Efficacy, Curiosity, Involvement, Recogni-
tion, and Challenge. Thus this factor includes some of the
scales measuring more intrinsic aspects, those measuring
efficacy and challenge, social reasons for reading, and the
extrinsic motivation recognition. The second factor con-
sisted of Compliance, Grades, Recognition (at the spring),
and Importance. It therefore is a combination of scales
measuring extrinsic motivation, one social aspect of reading,
and one scale designed to measure more intrinsic motiva-
tion. The third factor consisted of Competition and Work
Avoidance; Involvement loaded negatively on this factor at
the fall time of measurement.
Correlations of the motivation scales. Correlations of
the motivation scales are presented in Table 3. Most of the
relations were positive, and ranged from low to moderately
high. The strongest relations included those between Social
and Recognition (r = .62 in the fall, .50 in the spring),
Social and Involvement (r = .52 in the fall, .50 in the
spring),
Efficacy and Recognition (r = .60 in the fall, .53 in
the spring), Curiosity and Involvement (r = .52 in the fall,
62 in the spring), Curiosity and Challenge (r = .52 in the
fall,
.61 in the spring), Involvement and Challenge (r = .54
in the fall, .62 in the spring), Recognition and Grades
(r ~ .52 in the fall, .51 in the spring), and Grades and
Importance (r = .50 in the fall, .48 in the spring). The Work
Avoidance scale related negatively to all of the scales except
to Competition.
Relations of Children's Reading Motivation to
the Amount and Breadth of Their Reading
The correlations of the factor-based motivation scales to
the number of minutes per day children read outside of
school in each of the two school years, and the composite
breadth of reading variable taken from the Reading Activity
Inventory, are presented in Table
4.
The relations were in the
low to moderate range. The fall motivation scales showing
the strongest positive correlations with the amount of time
students read outside of school and the breadth of their book
reading included Efficacy, Involvement, and (to a slightly
lesser extent) Challenge, Recognition, Grades, and Social.
Fall Work Avoidance was significantly and negatively
related to amount of reading during the 1992-1993 school
year. The spring motivation scales most strongly and posi-
tively correlated to reading amount and breadth included
Curiosity, Involvement, Recognition, Grades, and, to a
lesser extent, Efficacy, Challenge, Importance, and Social.
Work Avoidance related significantly and negatively to
children's reports of the breadth of their book reading.
We looked at the relations of intrinsic and extrinsic
aspects to reading amount and breadth in several other ways.
First, we computed composite scales tapping broader as-
pects of motivation, for both the fall and spring scales. We
created these composites on the basis of theoretical distinc-
tions in the motivation literature, and on the results of the
factor analyses of the motivation scales in this study. The
first composite scale included scales measuring efficacy and
Table 3
Correlations of the Factor-Based Motivation Scales
Scale
1. Social
2.
Compliance
3.
Efficacy
4.
Curiosity
5.
Involvement
6. Recognition
7. Grades
8. Challenge
9. Competition
10.
Importance
11.
Work Avoidance
1. Social
2.
Compliance
3.
Efficacy
4.
Curiosity
5.
Involvement
6. Recognition
7. Grades
8. Challenge
9. Competition
10.
Importance
11.
Work Avoidance
1
.26**
.55**
.48**
.52**
.62**
.37**
.39**
.06
.36**
-.26**
.14
.33**
.52**
.50**
.50**
.41**
.49**
-!oi
.32**
—.27**
2
.25*
.32**
.39**
.27**
.46**
.35**
.11
.42**
-.17
.29**
.26*
.28*
.31**
.40**
.21*
.09
.37**
-.13
3
.52**
.42**
.60**
.43**
.51**
.26*
.41**
-.14
.47**
.51**
.53**
.35**
.49**
.24*
.35**
-.26**
4
Fall
.52**
.53**
.49**
.52**
.15
.49**
-.21*
Spring
.62**
.43**
.37**
.61**
.15
.42**
-.30**
5
.41**
.40**
.54**
.05
.45**
-.42**
49**
.34**
.62**
.09
.37**
-.33**
6
.52**
.34**
.34**
.44**
-.14
.51**
.41**
.28**
.52**
-.29**
7
.50**
.24*
.50**
-.19*
.30**
.19
.48**
-.14
8
.25*
.44**
-.32**
.22*
.27*
-.25*
9
.24*
.19*
.22*
.16
10
'
-.22*
-.10*
*/?<.O5.
**p<.01.
426WIGFIELD AND GUTHRIE
Table 4
Relations of Children's Reading Motivation
to Their Reading Amount and Breadth
Motivation scale
Reading Efficacy
Challenge
Curiosity
Involvement
Importance of Reading
Recognition
Grades
Social
Competition
Compliance
Work Avoidance
Reading Efficacy
Challenge
Curiosity
Involvement
Importance of Reading
Recognition
Grades
Social
Competition
Compliance
Work Avoidance
Reading amount
1991-1992
Fall
.31**
.04
.13
.26**
.11
.14
.12
.23*
-.15
.07
-.18
Spring
.19
.21*
.29**
.37**
!l4
.23*
.27*
.21*
-.09
.08
-.08
1992-1993
.36**
.11
.24*
.24*
.20
.24*
.21*
.18
.01
.23*
_
29**
.13
.22*
.27**
.31**
.21*
.32**
.32**
.13
.15
.06
-.13
Reading
breadth
.30**
.27**
.22*
.35**
.20
.25*
.18
27**
.06
.14
-.19
.36**
,33**
.50**
.51**
.36**
.39**
.23*
.31**
.15
.19
-.28**
Note. Reading amount is measured in minutes read per day.
*p<.05.
**/><.01.
aspects of intrinsic motivation, and included the Efficacy,
Curiosity, and Involvement scales—we call this the Intrinsic
composite. The second composite scale consisted of more
extrinsic aspects of motivation and performance goals; it
included the Recognition, Grades, and Competition
scales—we call this the Extrinsic composite.
We then examined whether the reading motivation com-
posites predicted variance in reading amount and breadth,
beyond the amount of variance predicted by previous
reading amount and breadth. In one set of analyses, the fall
Intrinsic composite was added after the previous reading
amount or breadth variable; followed by the fall Extrinsic
composite, the spring Intrinsic composite, and finally the
spring Extrinsic composite. In a second set the order of entry
of the Extrinsic and Intrinsic motivation composites was
reversed. Beginning with the analyses of 1993 reading
amount, children's amount of reading in 1992 was a
significant predictor of the amount they read in 1993
(R2
= .32, p < .001). Neither the fall Intrinsic nor the fall
Extrinsic composite resulted in a significant R2 change (no
matter which order they were entered). Adding the spring
Intrinsic composite to the equation after the two fall
composites did not significantly change the R2. But adding
the spring Extrinsic motivation composite after the other
three motivation composites increased the R2 significantly
(p < .001), with the overall R2 reaching .39. When the
spring Extrinsic composite was added to the equation after
previous reading amount and the two fall motivation compos-
ites,
the R2 increased significantly (p < .001) to
.39.
Adding
the spring Intrinsic composite did not increase the R2 further.
Children's breadth of reading in the fall predicted signifi-
cantly their breadth of reading in the spring (R2 = .27,
p < .001). Neither the fall Intrinsic nor Extrinsic composites
explained additional variance in spring reading breadth
beyond that explained by fall reading breadth, regardless of
the order in which they were entered. Adding the spring
Intrinsic composite resulted in a significant R2 change, with
the overall R2 increasing to
.42.
Adding the spring Extrinsic
composite last did not increase the R2 further. When the
order was reversed and the spring Extrinsic composite was
added immediately after the fall composites, the R2 in-
creased significantly (p < .05) to .35. Adding the spring
Intrinsic composite further increased the R2 to .42 (p < .001).
Further, when both the spring Extrinsic and Intrinsic compos-
ites were in the equation the Extrinsic composite no longer
was a significant predictor.
Next, we created three levels of each of the composite
variables by dividing the sample in thirds on the basis of
their score on each combined scale. We then ran one-way
ANOVAs on the following reading amount and breadth
variables: average minutes read per day during 1991-1992,
average minutes read per day during 1992-1993, fall book
reading breadth, and spring book reading breadth. Means for
each of the different groups and p values are presented in
Table 5.
Children with higher intrinsic motivation read more, and
with more breadth, than students with lower intrinsic
motivation. This pattern occurred on all of the variables, and
the differences were significant for all the variables except
for the amount of time spent reading in the 1991-1992
school
year.
The differences are most pronounced on amount
of time spent reading for the 1992-1993 school year. During
that year the highly intrinsically motivated children spent
nearly three times as much time reading outside of school
(29.80 min per day) than did the group lowest in intrinsic
motivation (10.52 min per day).
By contrast, for the groups high, medium, and low in
extrinsic motivation, only two of the amount and breadth
variables differed across the groups. One was the amount of
time spent reading in 1992-1993, where the children highest
in extrinsic motivation in the spring read more than twice as
many hours outside of school than did the group lowest in
extrinsic motivation. The second was on the fall reading
breadth variable; children highest in extrinsic motivation in
the fall said they read more books than did children in the
other two groups.
Are the differences across the different motivation groups
stronger for concurrent and future reading amount and
breadth variables than for past amount and breadth vari-
ables? Differences on the concurrent and future measures are
one indication that children's motivation predicts differences
in amount and breadth of their reading. Significant differ-
ences favoring the more intrinsically motivated children
occurred for the all the amount and breadth variables
measured concurrently or later than their motivation (see the
CHILDREN'S MOTIVATION FOR READING427
Table 5
Relations of Motivation and Amount and Breadth of Reading Over Time
Time
Low
M SD
Medium
M
Intrinsic motivation level
Concurrent and future
Intrinsic fall, breadth fall
Intrinsic fall, breadth spring
Intrinsic fall, amount 1992-1993
Intrinsic spring, breadth spring
Intrinsic spring, amount 1992-1993
Past
Intrinsic fall, amount 1991-1992
Intrinsic spring, amount 1991-1992
Intrinsic spring, breadth fall
2.09.
2.08.
12.19.
1.82.
10.52.
9.80.
9.18.
2.08.
0.53
0.52
10.80
0.47
10.42
10.62
10.66
0.56
2.22a,b
1.98a
14.44a
2.29b
18.05a,b
13.74a
15.74a
2.33b
Extrinsic motivation level
Concurrent and future
Extrinsic fall, breadth fall
Extrinsic fall, breadth spring
Extrinsic fall, amount 1992-1993
Extrinsic spring, breadth spring
Extrinsic spring, amount 1992-1993
Past
Extrinsic fall, amount 1991-1992
Extrinsic spring, amount 1991-1992
Extrinsic spring, breadth fall
2.18.
2.18.
13.98.
2.09.
11.61.
13.70.
11.59.
2.17.
0.54
0.58
12.65
0.55
12.63
12.59
11.12
0.48
2.18a
2.05.
17.98a
2.19a
17.48a,b
17.69a
16.52.
2.32.
SD
0.47
0.56
12.40
0.51
21.65
11.62
14.09
0.41
0.48
0.56
16.27
0.61
13.98
16.55
15.41
0.42
High
M
2.48b
2.48b
28.60b
2.48b
29.80b
19.34.
19.62.
2.49b
2.47b
2.39.
23.83.
2.38.
29.23b
12.80.
17.62.
2.39.
SD
0.40
0.49
30.21
0.57
27,24
14.71
14.87
0.41
0.42
0.53
29.31
0.52
31.48
11.05
14.64
0.54
P
.004
.001
.003
.000
.008
.020
.060
.007
.017
.060
.180
.130
.001
.360
.240
.180
Note. Means with different subscripts differ at the .05 levels.
top portion of Table 5). There was only one significant
difference for previous reading amount and breadth. For the
extrinsic motivation groups, there were two significant
differences (favoring the more extrinsically motivated chil-
dren) for the concurrent and later measured amount and
breadth of reading variables. There were no significant
effects for extrinsic motivation on the previous amount and
breadth of reading.
Further regression analyses were done to examine the
effects of the motivation composites on reading amount and
breadth controlling for the effects of the other motivation
composite. That is, we first entered one of the motivation
composites, and then entered the other one to see if it
explained variance beyond that explained by the first
composite. Beginning with the fall composites, when en-
tered first the fall Intrinsic motivation composite predicted
spring amount and breadth (R2 - .10, p < .05, and .12,
p < .01, respectively); the fall Extrinsic composite did not
add significantly to the equation. When entered first, the fall
Extrinsic composite predicted neither amount nor breadth;
the fall Intrinsic composite increased the R2 significantly
(overall R2 -
.10,
p <
.05,
and
.12,
p <
.01,
respectively).
When entered first, the spring Intrinsic composite pre-
dicted spring 1993 amount (R2 = .08, p < .01); the spring
Extrinsic composite added significantly to the equation
(overall R2 = .12, p < .001). Further, when the spring
Extrinsic composite was added, the spring Intrinsic compos-
ite no longer was a significant predictor. When entered first
the spring Extrinsic composite predicted significantly spring
1993 reading amount (R2 - .12, p < .001); the spring
Intrinsic composite did not add to the equation.
T\irning to reading breadth, when entered first the spring
Intrinsic motivation composite predicted spring breadth
significantly {R1 = .29,p <
.001).
Adding the spring Extrin-
sic composite did not increase the R2 significantly. When
entered first the spring Extrinsic composite also predicted
significantly spring reading breadth (R2 = .12, p < .001).
When added to this equation the spring Intrinsic composite
provided a significant R2 increase (p < .001) to the equation
(overall R2 = .29). Also, when the Intrinsic composite was
added, the spring Extrinsic composite no longer was a
significant predictor.
Levels of Reading Motivation
We computed means and standard deviations for the
motivation scales to see which children rated most highly.
These are presented in Table 6, along with an indication of
which means are significantly different. At both the fall and
spring times of testing children's mean scores were highest
on the Grades and Importance scales and lowest on the
Competition, Social, and Work Avoidance scales. The means
for the other six scales were relatively similar to one another
in both the fall and the spring. Their ordering in the fall and
spring also was relatively similar. Overall, these results
showed that children rated both more intrinsic and extrinsic
motivation highly and rated Competition, Social, and Work
Avoidance relatively low.
428WIGFIELD AND GUTHRIE
Table 6
Means and Standard Deviations for
the Factor-Based Motivation Scales
Table 8
Means and Standard Deviations for
the Statistically Significant Gender Differences
Motivation
scale
Grades
Importance
Efficacy
Involvement
Curiosity
Challenge
Recognition
Compliance
Competition
Social
Work Avoidance
Grades
Importance
Involvement
Challenge
Efficacy
Recognition
Compliance
Curiosity
Competition
Social
Work Avoidance
M
Fall
3.26a
3.22^
3-12^
3.08b
cd
3.04b
cd
3.01bcd
2.99d
2.95d
2.59e
2.5 U
2.22,
Spring
3.35a
3.14b
3.10b
3.06b
3.05b
3.04b
c
2-97b,c
2.92C
2.48d
2.40d
2.09e
SD
.70
.83
.63
.68
.73
.72
.71
.49
.78
.74
.72
.64
.88
.73
.76
.76
.66
.44
.77
.80
.72
.74
Note. Means with the same subscript do not differ; means with
different subscripts differ atp < .05.
Grade,
Time,
and Gender Differences
in Children's Motivation for Reading
We ran 2 (grade) X 2 (gender) ANOVAs on the various
motivation scales. The means for the significant grade and
gender differences are presented in Tables 7 and 8. In the
fall,
there were significant grade differences on three of the
scales: Efficacy,F(l, 94) =
8.33,p
<
.01;
Recognition, F(l,
94) =
4.86,
p <
.05;
and Social, F(l, 94) = 7.36,p <
.01.
In
all cases, the fourth graders had higher mean scores than did
the fifth graders. In the spring there were no significant grade
Motivation scale
Reading Efficacy
Importance
Social
Competition
Fall
Girls
M
3.29
3.48
2.78
2.31
SD
.46
.65
.72
.81
Boys
M
2.99
3.07
2.32
2.78
SD
.72
.96
.68
.69
Spring
Girls
M
2.63
2.15
SD
.71
.76
Boys
M
2.20
2.74
SD
.68
.75
Note. For the motivation scales, scores range from
1
to 4,
differences on the motivation scales. Fourth graders also
read significantly more minutes per day than did the fifth
graders in the school-based reading program during the
1992-1993 school year, F(l, 96) = 5.27, p < .05. There
were no grade differences on the fall or spring book reading
breadth measures.
Paired t tests were run on each pair of scales given in the
fall and the spring to assess whether the mean level of
children's responses on the different motivation scales
changed over time. There were no significant differences.
We also correlated the fall and spring motivation scales to
see how stable they were. The stability correlations ranged
from .28 for the Compliance scales to .68 for the Involve-
ment scales. The median stability correlation was .55.
In the fall administration there were gender differences on
four of the scales: Efficacy, F(l, 94) = 4.47, p < .05;
Importance, F(l, 99) = 4.22, p < .05; Social, F(lf 94) =
8.38, p <
.01;
and Competition, F(l, 94) = 7.84, p < .01;
see Table 8 for the means. In all cases but the Competition
scale, girls' mean scores were higher than boys' means. In
the spring there were significant gender differences on two
of these scales, Social, F(l, 91) = 8.57, p < .01, and
Competition, F(l, 92), p < .01. Girls had higher mean
scores than did boys on the Social scale, and boys had higher
mean scores than girls did on the Competition scale. Boys
and girls did not differ in the number of minutes per day read
outside of school, nor were there gender differences on the
fall or spring book reading breadth measures.
Table 7
Means and Standard Deviations for
the Statistically Significant Grade Differences
Fourth grade Fifth grade
Motivation scale
Fall Reading Efficacy
Fall Recognition
Fall Social
Average minutes per day spent
reading, 1992-1993
M
3.29
3.13
2.70
22.28
SD
.55
.70
.78
24.01
M
2.90
2.81
2.27
12.53
SD
.68
.70
.58
13.83
Note. For the motivation scales, scores range from 1 to 4. The
reading amount measure is in minutes per day. All the grade
differences occurred in the fall, with the exception of
the
amount of
reading difference, which was a year-end measure.
Discussion
The Multifaceted Nature of Reading Motivation
A major conclusion from our results is that reading
motivation indeed is multifaceted. The aspects of reading
motivation can be grouped conceptually according to impor-
tant constructs in the motivation literature, and the grouping
received empirical support from the factor analyses of the
scales. The Efficacy and Challenge aspects reflect work on
self-efficacy and competence beliefs (see Bandura, 1977;
Schunk, 1991; Wigfield & Eccles, 1992). Bandura and
Schunk both have argued that perceived efficacy is one of
the strongest predictors of achievement, and we find support
for these claims in our results.
CHILDREN'S MOTIVATION FOR READING429
Certain of the identified aspects (e.g., Curiosity, Involve-
ment) are tied clearly to the construct of intrinsic motivation
(see Deci & Ryan, 1985). Involvement adds an interesting
new aspect to the intrinsic motivation construct. This
dimension may be unique to areas like reading or art, where
the kind of book (or artwork) and the way the work is
composed can engage some individuals more than others
(see Schallert & Reed, 1997). The Recognition and Grades
aspects are like the extrinsic motivation construct in the
motivation literature, because they refer to doing an activity
for what it brings you rather than for its own sake.
Competition is tied to the notion of performance goals in the
motivation literature. From a theoretical perspective, Impor-
tance is a more intrinsic motivation. In our factor analyses of
the motivation scales, however, Importance loaded with the
more extrinsic scales. This may have occurred because
students felt the importance of participating actively in the
school reading program, a program that emphasized extrin-
sic aspects of motivating students to read. More research is
needed to clarify the nature of the Importance construct in
reading.
The Social and Compliance aspects concern something
relatively new in the motivation literature, social motivation
(see Wentzel, 1996). Social interactions in particular are
emphasized in recent school-based programs designed to
enhance children's reading engagement (e.g., Guthrie, Van
Meter, et al., 1996). More research is needed on this aspect
of motivation.
A number of the identified aspects were correlated
relatively highly with one another. Further, in the factor
analyses of the motivation scales the individual scales
grouped together in more intrinsic and extrinsic clusters. We
interpret these results to mean that children are motivated to
read by different but interrelated constructs. We believe that
it is useful to posit the 11 different aspects of reading
motivation, despite the correlations of some of the aspects.
The contexts of home and school afford many opportunities
for differentiation of reading motivation. Reading serves
multiple roles in school; it is a subject in
itself,
and a tool for
learning in all other subjects. It can be easy or challenging,
individual or social, rewarded or unrewarded, and competi-
tive or cooperative; therefore, there can be many motives for
reading. In future work, researchers should assess further the
dimensionality of these aspects of reading motivation.
Relations of Reading Motivation to Amount
and Breadth of Reading
Children's reading motivation (measured at the spring)
predicted the amount and breadth of their reading when
earlier amount and breadth of reading were controlled.
Children's previous reading amount and breadth themselves
were important predictors. Thus children who read more,
and more broadly, are likely to continue to do so, whereas
children reading less frequently are less likely to increase
their reading. But knowing whether or not children are
motivated to read adds predictive power to this equation.
Indeed, students highest on the Intrinsic motivation
composite read nearly three times as many minutes per day
as did the group lowest on this composite. This is a major
advantage for the group highest in intrinsic motivation. This
group also had a major advantage over the lowest group in
breadth of