Вы находитесь на странице: 1из 13

CONFIDENCE TO MANAGE LEARNING:

THE SELF-EFFICACY FOR SELF-REGULATED


LEARNING OF EARLY ADOLESCENTS WITH
LEARNING DISABILITIES
Robert M. Klassen

Abstract. This study examined the self-efficacy for self-regulated learning of 146 early adolescents with and without learning
disabilities (LD). Results from the study showed that a 7-item selfregulatory efficacy measure demonstrated factorial invariance for
the adolescent sample and also for a validation sample of 208
undergraduates with and without LD. Adolescents with LD rated
their self-regulatory efficacy and reading self-efficacy lower than
their NLD peers. Hierarchical multiple regression showed that selfregulatory efficacy made a significant contribution to end-of-term
English grade after controlling for sex, SES, reading self-efficacy,
and reading score. Finally, students with LD who scored low on
self-regulatory efficacy were significantly more likely than their
higher-scoring LD peers to have a low end-of-term English grade,
although there was no difference on a reading performance score.
Several suggestions for teachers working with adolescents with LD
are provided, along with directions for future research.

ROBERT M. KLASSEN, Ph.D., Department of Educational Psychology, University of Alberta, Edmonton, Alberta.

This study examined the self-efficacy for self-regulated learning of adolescents with learning disabilities
(LD). Social cognitive theory presumes that people have
a measure of agency over important aspects of their
lives and that they exercise this agency in a productive
manner in the pursuit of important goals. In selfefficacy theory, people evaluate their skills and abilities
and convert their beliefs about their capabilities into
purposive action (Bandura, 1997). Self-efficacy, popularly known as confidence, refers to beliefs in the capabilities to carry out the courses of action needed for
desired goals; self-regulated learning refers to the ability
to regulate cognition, motivation, affect, and behavior
in a learning context.

As people enact their self-efficacy beliefs, they demonstrate a degree of control over (a) the activities they
choose to pursue, (b) the persistence they display in the
pursuit of goals, and (c) their reactions to challenges and
failures (Bandura, 1997). Qf course, desired outcomes are
not brought about only by the strength of one's beliefs.
In academic settings, specific skills are needed to master
difficult tasks, and important factors like educational
opportunities, quality of teaching, and learning ability
influence student success. In order to be successful, students must exercise control over their learning by activating and regulating behaviors, thoughts, and
emotions, and they must learn to manage their learning
environment in a planful way (Zimmerman, 2000).

Volume 33, Winter 2010

19

A student's confidence to manage his or her learning


environment - self-efficacy for self-regulated learning {ox

self-regulatory efficacy, for brevity's sake) - has been


shown to be a key factor that inuences the academic
success of children, adolescents, and adults (e.g.,
Caprara et al., 2008; Pajares & Vahante, 2002).
Successful students regulate their learning by setting
goals, retaining their focus on difficult tasks in the face
of seductive distractions, organizing their workspace
and workload, and engaging in effective learning
strategies (Zimmerman, 2000). Academic self-regulation demands knowledge of learning strategies, the
capability to engage these strategies, and the will to
persist with the strategies when obstacles are faced.
Unlike fixed abilities like intelligence, self-regulation is
a dynamic process involving cognition, emotion,
behavior, and context, with forethought (pre-learning
processes like goal setting), leading to performance or
volitional control (learning processes like attention
focusing), and finally to self-refiection (post-learning
processes like self-evaluation) (Schunk & Zimmerman,
1998). Knowledge of self-regulated learning strategies is
not sufficient for success; a critical factor that determines whether or not students engage in self-regulatory processes is the beliefs they hold about their
capabilities to regulate their learning (Usher & Pajares,
2008).
Beliefs about learning are especially critical during
the adolescent years, which are sometimes viewed as a
period of declining academic motivation: The developmental shift from being a bright and interested child to
a more emotionally complex and sometimes academically disinterested adolescent has been attributed to a
host of factors, including poor fit between the adolescent and the school environment, hormonal changes,
and a growing awareness of social and academic competition (e.g., Eccles et al., 1993). The development
of self-regulation skills is infiuenced by (a) cognitive
factors, like metacognitive knowledge awareness and
working memory; (b) motivation and affective factors,
like interest and task value; and (c) behavioral factors,
like time and effort management.
Age-related changes in self-regulation are the result of
increased capacity for knowledge, representation, strategy use, metacognition, and more accurate belief systems (Pintrich & Zusho, 2002). Whereas children in
elementary school use basic organizational and learning
strategies, more sophisticated cognitive and metacognitive learning strategies emerge in adolescence, partly in
response to greater expectations for independent (i.e.,
self-regulated) learning.
Previous studies have shown a drop in adolescents'
academic self-efficacy beliefs and self-regulatory efficacy
(Caprara et al., 2008), and these declines come at a time

when academic achievement becomes critically important, and when future and academic and vocational
opportunities rest on successful navigation of adolescent schooling. In Caprara et al.'s longitudinal study of
self-regulatory efficacy, low levels of self-regulatory efficacy at Time 1 (early adolescence) infiuenced school
grades at Time 3 (end of junior high), which, along with
social-economic status (SES) and self-regulatory efficacy
at Time 4 (secondary school), influenced the decision to
drop out of secondary school at Time 5.
Exploring academic self-regulatory efficacy during
adolescence is crucial, because it not only influences
school performance but also has ramifications for the
life course. For students with learning difficulties, selfefficacy for self-regulated learning may be even more
critical than for typically achieving students, because
the academic doubts stemming from specific skill
deficits may be neutralized by the confidence to engage
in self-regulatory processes like strategic planning, selfmonitoring, and accurate self-evaluation.
Most studies of academic motivation and self-regulation have examined typically achieving students, with
less research attention paid to students with learning
difficulties. Adolescents with LD experience deficits in
specific areas of academic functioning but go through
the physical and psychological changes that characterize the development of all adolescents. Although low
skill levels in students with LD may be associated with
lower levels of self-regulated learning, the two areas (i.e.,
academic skills and self-regulated learning) are theoretically distinct. For example, students with a reading
deficit are capable of learning to organize their learning
environment, plan learning activities, and monitor
their learning progress in ways that are not influenced
by poor reading skills.
Early studies of metacognitive functioning found
that students with LD were less skilled in two forms of
metacognition: problem solving and performance
monitoring (Slife, Weiss, & Bell, 1985). More recent
studies have found that adolescents with LD reported
lower academic and social self-efficacy beliefs (e.g.,
Baird, Scott, Dearing, & Hamill, 2009; Lackaye,
Margalit, Ziv, & Ziman, 2006). Klassen (2007) included
a measure of self-regulatory efficacy in a study of the
spelling and writing of adolescents with and without
LD. He noted that self-regulatory efficacy was lower for
students with LD, but it did not contribute to spelling
and writing performance. In a qualitative follow-up
study, Klassen and Lynch (2007) found that adolescents attributed their performance on learning tasks to
controllable factors such as effort, whereas their teachers attributed LD students' learning performance to
uncontrollable deficits.

Learning Disability Quarterly

20

study Purpose
Research on the self-efficacy beliefs of adolescents
with LD is scarce, but there is potential instructional
benefit to be gained by investigating the learning beliefs
of adolescents who find learning difficult.
There, were four key purposes to this research.
1. To examine the validity of the 7- and 11-item SelfEfficacy for Self-Regulated Learning (SESRL) scale
in a sample of early adolescents with and without
LD. The SESRL scale was originally developed by
Bandura (1990), with two versions currently in
use: a long version with 11 items (occasionally 10
items) and a more succinct 7-item adaptation.
Both versions have been used extensively (e.g.,
Caprara et al., 2008; Pajares & Valiante, 2002;
Usher & Pajares, 2008; Zimmerman, Bandura, &
Martinez-Pons, 1992), although their comparative
validity has not been examined. Recent examples
of use include Caprara et al.'s use of the original
11-item scale in their study of adolescents in Italy,
and Usher and Pajares' use of the 7-item scale in a
study of American students in grades 4 to 11. The
short form of the measure is derived from the long
form, and both measures are hypothesized as unidimensional measures of students' beliefs in their
self-regulatory capabilities. Previous research has
not explored the comparative validity of the 7item and 11-item forms of the scale, nor have prer
vious studies validated the measure for students
with LD.
2. To compare levels of self-regulatory efficacy and
related variables (reading self-efficacy, reading performance, and English grade) in adolescents with
and without LD.
3. To investigate whether self-regulatory efficacy
made a unique contribution to the prediction of
end-of-term English grade beyond the contributions of reading performance and reading selfefficacy.
4. To investigate the reading performance, self-efficacy, English grade, and SES of LD adolescents
with high and low levels of self-regulatory efficacy.
The variables of reading performance, reading selfefficacy, and end-of-term English grade were included
because reading proficiency and the literacy skills
required in English classes are key enablers of learning
across all subject areas and grade levels. Furtherrhore,
they are critical factors in adult employability and
career success. Students who struggle to read confidently and fiuently not only do poorly in English class
but also are at risk for school dropout and lower engagement in the workforce (Daggett & Hasselbring, 2007).
(End-of-term grades are imperfect measures of language-

related achievement and include assessment of a number of activities that vary from teacher to teacher and
school to school. However, high school grades more
accurately predict future academic success in high
school and college than any other factor [Hoffman &
Lowitzki, 2005].)
A measure of SES (father's educational attainment)
was included because SES influences academic achievement through access to resources that promote intellectual development (Caprara et al., 2008). For example,
Eamon (2002) found that income level influenced early
adolescent reading achievement through its association
with level of cognitive stimulation in the home environment.
Four hypotheses pertaining to the self-regulatory
efficacy of adolescents with and without LD were
proposed:
Hi: Based on social cognitive theoretical principles
and recent research (Usher & Pajares, 2008), the
SESRL measure will form a unidimensional construct that shows an equivalent factor structure
for students with and without LD. (No hypothesis
about the preference of the 7-item over the 11-item
measure was proposed.)
H2: Based on previous research (e.g., Klassen, 2007)
and on theory suggesting that learning disabilities
are associated with metacognitive deficits (e.g.,
Butler, 2003; Slife et al., 1985), students with LD
will report lower levels of self-regulatory efficacy
than their peers without learning disabilities
(NLD).
H3: Because the capacity to regulate learning is of
critical importance in the academic performance
of all students, levels of self-regulatory efficacy
will significantly contribute to the prediction of
academic achievement for students with and without LD.
H4: Because self-regulatory efficacy influences
learning in a variety of domains, adolescents with
LD who report low self-regulatory efficacy will
score lower on reading performance, reading selfefficacy, and attainment on end-of-term English
grades.

METHOD
Participants
Participants were 146 early adolescents in grades 8
and 9 at three high schools in Western Canada.
Participants were recruited by the author and by special
education teachers in each of the schools. The LD group
consisted of 73 early adolescents with LD (Mage = 13.89
years) who received pullout learning support in a
resource room one class period out of eight each school
day due to identification of a diagnosed learning dis-

Volume33, Winter 2010

21

ability. The 73 NLD (control) participants (Mgge =13.93


years) were matched for gender and age with the students with LD.
There were 25 females and 48 males in each group.
The control participants did not receive service for
learning difficulties, spoke English as a first language,
and were recruited from social studies classes in the
three schools. NLD students with a history of receiving
service for mild learning difficulties or English as a second language were not included in the study.
The three schools participating in the study were
located in neighborhoods with similar socio-economic
characteristics (median income between 113% and
127% of provincial median income) according to
income data from Statistics Canada (Statistics Canada,
2009), and scored in the middle 50% of schools on
provincial achievement tests (Fraser Institute, 2009).
There were no significant differences on any of the variables according to school membership for LD or NLD
participants.
Student ethnicity was over 90% European Canadian
in both groups, with modest proportions of students
from South and East Asian backgrounds. Students with
LD were identified by school psychologists using school
district criteria that stipulate a two-standard-deviation
discrepancy between IQ (from Wechsler or StanfordBinet scales) and standardized achievement scores, with
IQ scores in the average range or higher. The provincial
and school district's diagnostic criteria for LD are based
on the official definition of the Learning Disabilities
Association of Canada (2002), which states that:
The term leaming disability refers to a number of
disorders which may affect the acquisition, organization, retention, understanding or use of verbal or
nonverbal information. These disorders affect
learning in individuals who otherwise demonstrate
at least average abilities essential for thinking
and/or reasoning, (paragraphs 1-3)
Teachers reported that all, or almost all, of the LD participants were identified as disabled in literacy areas
(reading and writing). Mean IQ and standardized
achievement scores from student files were supplied for
students in the LD group, and end-of-term English
grades were supplied for all students. Mean full-scale IQ
of the LD participants was 97.67 {SD = 12.64), with
mean standardized reading and writing scores at 80.8
(SD = 17.4) and 76.2 {SD = 14.1), respectively.
Psychometric data were not available for students in the
NLD group; it is assumed that IQ scores and literacy and
numeracy skills were normally distributed. There was
no difference on the SES measure (father's educational
attainment) between the LD and NLD groups, F{1, 143)
= 2.75, p = .10. Students in the LD group scored signifir
cantly lower on end-of-term English grade (M= 3.64, SD

= 1.56 on a 1 = D [or Incomplete] to7 =A scale) than students in the NLD group (M = 4.79, SD = 1.76), f(l,145)
= 17.37, Cohen's d=.69.
Procedure
Ethics approval for the research was provided by the
university and the school district, and parental permission was received for all students who participated. The
measures were pilot-tested in one class of grade 8 students in one of the target schools, resulting in minor
changes to the oral directions to improve student
understanding.
Data were collected over a three-week period by the
author, with assistance from the resource room teachers
in the pullout classes for students with LD. The measures were administered to groups of 8-15 students with
LD in the resource room setting and to class groups of
15-25 students in the control setting. Directions were
read aloud, and the researcher and teachers provided
clarification for students who needed it. The participation rate was estimated at over 75% in both groups.
Measures
The 11-item SESRL measure assesses students' beliefs
that they hold the capability to successfully employ selfregulatory strategies (Zimmerman et al., 1992). The
measure has shown reliability and validity in recent
studies of students with LD (e.g., Klassen, 2007; Klassen,
Krawchuk, Lynch, & Rajani 2008) and without LD (e.g.,
Caprara et al., 2008). Cronbach's alpha reliability coefficients in previous research were typically aboye .80. The
7-item version of the SESRL scale is a recently used adaptation of the 11-item scale, with a Cronbach's alpha of
.83 in a recent scale validation (Usher & Pajares, 2008).
Example items that are identical across the 11- and 7item scale include. How well can you finish study when
there are other interesting things to do? and How well can
you finish assignments by deadlines? Scores were measured on an 11-point scale, anchored at 0 by Not well at
all and at 10 by Very well. The range of possible scores
was 0 to 110 for the 11-item scale, and 0 to 70 for the 7item scale.
Reading skill was measured using an adaptation of a
reading test from the Woodcock-Johnson III Tests of
Achievement (Woodcock, McGrew, & Mather, 2001), in
which students are given 50 brief sentences to read in a
2-minute period. Students must read each sentence and
indicate whether the statement is True or False (e.g., A
mountain is usually made of water). The total score of correct responses was summed to give the reading score,
with a possible range of 0 to 50. Median reliability on
the subtest is listed as .90 by the test authors (Woodcock
et al.).
Reading self-efficacy was measured using procedures
outlined in Bandura's (2006) guide to constructing self-

Leaming Disability Quarterly

22

efficacy measures. Consistenf with Bandura's dictums,


fhe measure was domain specific (reading) and offered
gradations of challenge linked fo the outcome measure.
Students were given detailed oral and written instructions, followed by examples of the task, and were then
asked to rate their degree of confidence (from 0 =
Cannot do at all through 5 = Maybe can do, to 10 =
Certain can do) of getting 30%-100% correct for the
reading test. Scores for the eight levels of confidence
(30%-100% with increments of 10%) were summed for
a total self-efficacy score, with a possible range of 0-80.
The reliability coefficient (Cronbach's alpha) for the
reading self-efficacy measure was .95.
End-of-term English grades, provided by the teachers
at the end of the school year, were used as a measure of
overall language-related achievement; possible range of
1 = D or / (Incomplete) through to 7 = A (Excellent).
Classroom-level English grade data were not collected,
but there were no significant differences in end-of-term
English grade among the schools (p = .15). SES was
assessed by students' ratings of their fathers' educational
attainment on a scale from 1 "some high school" to 5
"graduate degree." Early adolescents' ratings of parental
education level, and especially ratings of father's education level, are validly associated with parents' SES (Lien,
Friestad, & Klepp, 2001).

RESULTS
The first research question to be answered related to
the choice of the standard 11-item or adapted 7-item
SESRL measure and the relative invariance of the two
versions for LD and NLD samples. Table 1 presents fit
indices for the two versions of the SESRL measure for
LD, NLD, and combined groups. Goodness-of-fit was
evaluated by the conventional use of change in x^ (^X^),
in which a significant Ax^ indicates invariance, and by
changes in the comparative fit index (ACFI), wherein a
ACFI less than or equal to .01 indicates invariance
between models (Cheung & Rensvold, 2002). The ACFI
is viewed as superior to the use of Ax^ as a test of invariance, because it is not affected by sample size. However,
reporting Ax^ is conventional and is included in the current article to enable comparison with other research
(see Cheung & Rensvold, for further explanation). As
reported in Table 1, the 7-item SESRL measure provides
a significantly better fit for the LD, NLD, and combined
groups, with significant differences in x^ and CFI (range
.11-.21) for each comparison.
The 7-item model showed good fit for the LD, NLD,
and combined groups, with all x^/df ratios below the
upper threshold proposed by Byrne (2004), ranging
from 1.18 for the LD group to 2.04 in the combined

Table 1

Fit Indices for 7- and 11-Item Measures of Self-Efficacy for Self-Regulated Learning

j2ldf

CFI

RMSEA

<.OO1

2.37

.11

.14

44

<.OO1

2.06

.85

11

150.49

44

<.OO1

3.42

.83

.13

LD (n = 73)

16.50

14

.28

1.18

.97

.05

87.96(30)**

.21

NLD ( = 73)

20.18

14

.12

1.44

.96

.07

70.28(30)**

.11

Combined (N = 146)

28.56

14

.01

2.04

.96

.08

121.93(30)**

.13

Model

x'

df

104.46

44

90.46

ACFI

Model 1: 11 Items
LD (n = 73)
NLD ( = 73)
Combined (N = 146)

Model 2: 7 Items

*p< .05. **p<.01.

Volume 33, Winter 2010

23

Table 2
Means and Standard Deviations for Key Variables for Early Adolescents with and
Without LD

NLD

LD
(w = 73)

in = 73)

SD

SD

SESRL (7 items)

.81

42.17

11.54

49.92

10.25

.71

Reading SE

.91

52.89

16.53

62.09

12.61

.63

Reading Score (50 items)

.95

33.79

9.77

42.49

7.65

.99

3.64

1.56

4.79

1.76

.69

English Grade

Note. SESRL = Self-Efficacy for Self-Regulated Learning. Possible range for SESRL = 0-70. SE = self-efficacy. Possible range for reading SE = 0-80.
English grade is reported on a 7-point scale. All comparisons were significantly different, ps < .001.

group. Similary, the CFIs ranged from .96 in the NLD


and combined groups to .97 in the LD group, all above
the minimum CFI threshold proposed by Byrne (2004).
The root mean square errors of approximation (RMSEAs)
ranged from .08 (lower is preferable) in the combined
groups to .05 in the LD group, with all RMSEA levels
below the .09 upper limit suggested by Byrne.
As a validity check, data from a previous study that
used the SESRL with a sample of 208 undergraduates

with ( = 101) and without (n = 107) LD (Klassen et al.,


2008) were re-analyzed using CFA. Results showed that
the 7-item SESRL measure was preferable to the 11-item
measure for undergraduates with LD (A%2 = 62.17, ACFI
= .07, p < .001) and for the whole sample (A%2 = 91.57,
ACFI = .10, p < .001). This finding suggests that measuring SESRL using the 7-item measure is preferable for stu. dents of diverse ages and that the measure is invariant

Table 3
Correlations for Key Variables for Early Adolescents with and Without LD

1.

2.

3.

4.

.37**

.18

.42**

.34**

.64**

.33**

3. Reading Score

.18

.48**

.43**

4. English Grade

.50**

.37**

.51**

1. SESRL
2. Reading SE

Note. SESRL = Self-Efflcacy for Self-Regulated Learning. SE = self-efflcacy. English grade is reported on a 7-point scale. Correlations for the LD sample
are above the diagonal; correlations for the NLD sample are below the diagonal.
*p < .05. "p < .01.

Learning Disability Quarterly

24

for individuals with and without LD. Thus, Hj - that the for LD status, f(4, 138) = 14.18, p < .001,1. = .71; and for
SESRL measure would form a unidimensional structure sex, f (4, 138) = 5.44, p < .001, A, = .86; but not for the
for students with and without LD - was confirmed. (For interaction between LD status and sex, f(4, 138) = 1.19,
the remainder of this article, references to the SESRL p = .32, X = .97.
Univariate tests revealed significant differences for all
measure refer to the 7-item version.)
Table 2 presents reliability indices, means, and stan- the dependent variables according to LD status: Mean
dard deviations for the SESRL measure, reading self-effi- scores for the NLD group were significantly higher than
cacy, reading score, and end-of-term English grade for mean scores for the LD group for each variable (ps <
the LD and NLD samples. The measures showed ade- .001), with Cohen's d effect sizes ranging from medium
quate internal reliability: Cronbach's alpha reliability (.63 for reading self-efficacy) to large (.99 for reading
indices ranged from .81 for SESRL to .95 for the reading score) using Cohen's effect size descriptors (Cohen,
measure. A multivariate analysis of variance (MANOVA) 1988). Mean scores for SESRL and English grade were
was conducted with LD status and sex as independent significantly higher for girls across the LD and NLD
variables, and SESRL (7-item), reading self-efficacy, read- groups ip = .007 for SESRL, and p = .005 for English
ing score, and English grade as dependent variables. grade), but reading self-efficacy and reading score were
Comparisons showed a significant multivariate effect not significantly different according to sex. Because of

Table 4
Hierarchical Multiple-Regression Analyses Predicting English Grade from Reading
Performance, Reading Self-Efficacy, and Self-Efficacy for Self-Regulated Learning for LD and
NLD Adolescents

Step 1
Control Variables
Step 2
Control Variables
Reading Score
Reading SE
Step 3
Control Variables
Reading Score
Reading SE
SESRL
Total
n
Note. Control variables included sex of student and father's educational level.
*p<.05. " p < .

Volume 33, Winter 2010

25

Table 5
Means and Standard Deviations for Key Variables for LD Adolescents with High and Low
Self-Efficacy for Self-Regulation

High SESRL
(n == 27)

Low SESRL
(n == 25)

SD

SD

Reading SE

56.11

15.66

47.01

18.89

.52

Reading Score (50 items)

33.74

10.96

32.32

9.87

.14

English Grade**

4.30

1.64

3.00

1.71

.78

SES*

3.60

1.34

2.71

1.16

.70

Note. SE = self-efficacy. English grade is reported on a 7-point scale. SES is measured as father's highest educational attainment on a 5-point scale.
*p< .05. **p< .01.

the similarities in patterns between sexes in the key variables, the decision was made to combine males and
females for further analyses. The second hypothesis,
H2 - that early adolescents with LD would report lower
levels of self-regulatory efficacy - was confirmed.
Bivariate correlations for the key variables between
the LD and NLD groups were similar (see Table 3). The
SESRL variable was significantly and moderately correlated with reading self-efficacy (r = .37, r = .34, ps < .01)
and English grade (r = .42, r = .50, ps < .01) for the LD
and NLD groups, respectively. But SESRL was not significantly correlated with reading score in either group.
The reading score was significantly correlated with reading self-efficacy in each group, ps < .01.
Table 4 presents the results from hierarchical multiple
regression for the LD and NLD groups, with end-of-term
English grade as the dependent variable. Control variables of SES (parent education level) and sex were
entered at Step 1, followed by reading score and reading
self-efficacy at Step 2, and finally SESRL at Step 3.
For the LD group, the entry of the control variables at
Step 1 did not significantly predict English grade. The
entry of reading score and reading self-efficacy at Step 2
significantly increased explained variance (Ai^^ = .21), as
did the entry on the final step of SESRL (M^ = .08), with
a final K^ of .35, f(5, 69) = 6.81, p < .001. At the final
step, reading score was the strongest predictor of
English grade ( = .44), followed by SESRL ( = .34).
Reading self-efficacy did not significantly contribute to

the prediction of English grade. For the NLD group, the


entry of the control variables at Step 1 significantly predicted English grade {R^ = .11). The entry of reading
score and reading self-efficacy contributed significantly
to English grade {AR^ = .22), as did the entry on the final
step of SESRL (M^ = .10), with a final R^ of .45, ^(5, 72)
= 10.96, p < .001. At the final step, reading score was the
strongest predictor of English grade ( = .37), followed
by SESRL ( = .36). The third hypothesis, 3- that selfregulatory efficacy would be associated with academic
achievement - was confirmed.
The final analysis (presented in Table 5) consisted of a
comparison between students with LD who scored in
the top third (n = 27, M = 53.57, SD = 4.91) and bottom
third (n = 25, M = 29.30, SD = 7.03) on the SESRL variable. The purpose of the analysis was to highlight the
contrasts between students reporting high and low confidence to regulate their learning.
Students scoring high on the SESRL measure scored
significantly higher on English grade (M = 4.3) than the
low-SESRL group (M = 3.00, p < .01) and came from
higher SES backgrounds (p = .02, d = .70). There were no
significant differences for high and low scorers on
SESRL for reading self-efficacy (p = .07) or reading score
(p = .63). Thus, the fourth hjqjothesis, ^- that LD participants low in self-regulatory efficacy would display
lower reading performance, reading self-efficacy, and
lower English grades - was partially confirmed.

Learning Disability Quarterly

26

DISCUSSION
The purpose of this study was to examine the validity of two versions of the SESRL scale in samples of
early adolescents with and without LD, and to examine
the role of self-regulatory efficacy and its relationship
with reading self-efficacy, reading performance, and
English grade for adolescents with and without LD.
The 7-item version of the SESRL measure fit the data
significantly better than the 11-item measure for early
adolescents with and without LD. A second CFA using
an undergraduate sample confirmed the same pattern
of a better fit for the 7-item measure compared to the
11-item measure.
As hypothesized, the CFA showed that the items in
the 7-item measure formed a unidimensional structure
that was invariant for students with and without LD.
The first finding from this study, then, was that the 7item version of the SESRL measure showed construct
validity with samples of adolescents and undergraduates with LD. These results pave the way for future studies of self-efficacy for self-regulated learning in samples
of students with LD, with evidence that the modified 7item measure is internally valid and shows structural
integrity across samples.
As predicted, early adolescents with LD rated their
self-regulatory efficacy lower than did their NLD peers.
Furthermore, girls rated the variable higher than boys,
and this difference held true for adolescent boys and
girls with and without LD. These results are consistent
with previous research showing that female students
have the edge in terms of confidence to regulate their
learning over their male peers. That this relationship
holds true for adolescent girls and boys with LD is not
surprising, given that past research with NLD samples
has revealed that comparatively more girls than boys
employ strategies that optimize management of the
learning environment (e.g., Ablard & Lipschultz, 1998;
Zimmerman & Martinez-Pons, 1990). Past studies have
shown that adolescent girls show higher levels of
metacognitive self-awareness and accuracy in calibrating their academic self-efficacy beliefs with performance, and tend to be less self-congratulatory than
adolescent boys (e.g., Klassen, 2007). Results from this
study revealed that adolescent girls with and without
LD showed higher levels of confidence to regulate and
. monitor their learning, hinting at better developed
metacognitive skills with resultant higher levels of selfregulatory efficacy.
The results from the hierarchical multiple regression
showed that self-regulatory efficacy contributed significantly to the prediction of end-of-term English grades
for early adolescents, and that this contribution held
true for LD and NLD adolescents even after controlling

for SES and sex. In fact, for both groups, self-efficacy for
self-regulated learning contributed almost as much to
the prediction of English grade as did reading ability.
The importance of reading skill to academic performance in English class is expected: Better readers tend to
do better in literacy-oriented classes. The significant
contribution of self-regulatory efficacy for students
with ahd without LD is of critical importance.
Although reading instruction typically ends after the
elementary grades, and reading difficulties are notoriously difficult to remediate in adolescence (Daggett &
Hasselbring, 2007), self-regulation skills can be readily
taught and learned through strategy instruction
schemes such as Butler's Strategic Content Learning
(Butler, 2003). Self-efficacy to self-regulate can be developed in students with LD if they are allowed opportunities for successful experiences, offered verbal
persuasion, and provided with appropriate models.
The relationship between self-efficacy for self-regulated learning and school achievement is highlighted
in the results shown in Table 5, and underlines the
need for further research and attention to practice. Two
critical findings emerge by examining adolescents with
high and low self-regulatory efficacy. First, and contrary to expectations, there was no difference in reading ability between the two groups. In other words,
early adolescents with high and low self-regulatory efficacy do not differ in their basic reading comprehension
and fluency; yet, there is a marked difference in English
grade.
This is an important finding for researchers and practitioners alike. For students with LD, performance in
literacy-based subjects like English is only partially
dependent on reading ability (as shown in the multiple
regression in Table 4). Of almost equal importance is a
student's belief in his or her capabilities to manage the
learning environment. The implication of this finding
is that remedial instruction for adolescents with LD
should include a focus on building self-regulatory skills
and the confidence to use these skills. For some students with LD, poor academic performance may be the
result of low confidence to manage their learning, not
merely low skill levels.
The second key finding is the significant difference in
SES found for students who were high and low in selfregulatory efficacy. Previous studies have found links
between reading achievement and income level (e.g.,
Eamon, 2002), but the results from the current study
showed that students who are low in self-regulatory
efficacy come from families with lower levels of paternal educational attainment. This finding is consistent
with that of Caprara et al. (2008), who proposed that
low SES influences students' confidence to use selfregulated learning tools.

Volume 33, Winter 2010

27

The implications for practitioners are clear: Students


from lower-SES families may need additional support to
develop the tools needed for academic self-regulation,
and; importantly, may lack the confidence to implement the skills that they do possess. Whereas previous
social policies and programs have focused on developing literacy and numeracy skills for students from
lower-SES settings, policies and programs in the future
should provide an additional emphasis on building students' confidence to use academic self-regulation skills.
Implications for Practice
The results from this study have encouraging implications for practitioners who work with adolescents with
LD. Students with LD need enough guidance, strategy
instruction, repetition, and practice to develop confidence that they are able to manage their own learning.
The confidence, or self-efficacy, to regulate learning
may be built from Bandura's four hypothesized sources
of self-efficacy (1997). The first and most powerful
source is individuals' interpretations of their mastery
experiences, or previous successful experiences. In educational settings, the successes students experience
typically build their self-efficacy, whereas failure experiences undermine it.
The second source, observation of successful others, or

vicarious experience, also builds self-efficacy beliefs.


Immediate performance feedback is not always readily
available in school settings, and observation of others'
attainments provides valuable information about one's
own relative competence. Adolescents are particularly
sensitive to the performances of social models that they
perceive share similar characteristics with them (e.g.,
learning ability, gender, race, age). Unfortunately, selfregulated learning practices are difficult to observe
directly, and successful managing of the learning environment at home or at school is not readily observable.
Teachers of students with LD would do well to highlight
students' successful self-regulated learning practices and
offer students opportunities to discuss what does and
what does not work in reg:ulating their learning environment.
A third source of self-efficacy is social persuasion.
Students who are persuaded verbally that they are capable of carrying out self-regulated learning behaviors display effort and perseverance, at least temporarily, when
faced with challenges and distractions. Students will
begin to focus on self-regulatory capabilities when their
teachers regularly discuss the importance of self-regulated learning, and when teachers offer verbal encouragement for students to practicing self-regulatory
thinking and practices.
Finally, students', confidence in their capabilities to
self-regulate is infiuenced by their physiological and

affective states when undertaking self-regulated learning


activities. Individuals interpret various indicators of
stress, anxiety, and fatigue when considering their
capabilities to handle everyday challenges. For example, a student practicing a new self-regulated learning
behavior (e.g., self-questioning while reading) may
interpret bodily states of frustration or nervousness as
indicative of a lack of competence to carry out the task
and, therefore, be less likely to engage in that activity.
Information from the four sources of self-efficacy is
not inherently enlightening without cognitive interpretation of the information from each of the sources.
Teachers and parents can help guide adolescents to
cognitively process the self-efficacy sources by emphasizing students' successes (i.e., mastery experiences), by
highlighting examples of competent self-regulated
practices in peers (i.e., vicarious experiences), by offering encouragement to adopt self-regulatory practices
(i.e., social persuasion), and by helping students anticipate and manage physiological and affective reactions
to anxiety-producing learning situations.
Interventionists have focused on providing students
with LD domain-specific academic, cognitive, and selfregulatory strategies to develop academic skills (e.g.,
Graham & Harris, 2000), and these efforts have ofi:en
improved academic performance (e.g., Garca-Snchez
& Fidalgo-Redondo, 2006). Although post-intervention
measures of self-regulatory efficacy were included in
these studies, it is plausible to conclude that intervention targeted at boosting students' self-regulatory
processes also boost their confidence to regulate these
processes.
Garcia-Snchez and Fidalgo-Redondo (2006) conducted a study designed to promote self-regulated writing strategies in students with LD. Their intervention
included elements of self-instruction, problem definition, focusing of attention, self-evaluation, and selfreinforcement, all in an attempt to improve students'
writing performance. Students who engaged in the selfregulatory interventions showed significant improvements in the quality of their writing work and also
increased time spent on writing and revising activities.
Writing self-efficacy was also improved by the intervention, but no measure of self-regulatory efficacy was
administered. It is advised that future studies of interventions that focus on building self-regulated learning
in students with LD include measures of self-regulatory
efficacy.
Teachers regularly evaluate students' skills and
knowledge in content areas through informal and formal tests, but self-regulatory efficacy is an area that
could also profitably be evaluated through regular
checks by teachers. Schunk and Usher (in press) proposed administering self-efficacy assessments as diag-

Leaming Disability Quarterly

28

In addition, the self-regulatory efficacy of children


nostic tools, whereby teachers develop and ask quesand
older adolescents with LD may not mirror the pattions that assess aspects of self-regulatory efficacy.
terns
discovered in this study. Self-regulated learning
Questions may include items adapted from self-regulaskills
are
thought to develop through childhood and
tory efficacy measures such as. How confident are you
adolescence
(Pintrich & Zusho, 2002), and the relationthat you can finish your homework on time? and How well
ship
between
self-regulatory efficacy and academic
can you arrange a place to study at home where you won't
functioning
(reading
skills and grades) may change
get distracted? Responses to these and other questions
over
time.
would provide diagnostic information that would
Finally, other variables not included in this study
allow teachers to target interventions designed to build
students' self-regulated learning capabilities and confi- also contribute to academic performance, as suggested
by the modest English grade variance explained by the
dence in these capabilities.
Early adolescents with LD are at a critical develop- study variables. Students' reported self-regulated learnmental juncture where self-regulated learning processes ing practices, as opposed to their self-efficacy for selflike strategy development, study skills, and the ability regulated learning practices, may have contributed
to focus on learning tasks become a critical aspect of additional variance to end-of-term English grade and
academic functioning. Most adolescents become more should be measured in subsequent studies.
Future research should explore the effectiveness of
strategic and more capable of using learning strategies
self-regulation
interventions provided to adolescents
as they progress through junior high and high school
with
LD,
with
a
particular focus on how self-regulation
(Pintrich & Zusho, 2002).
strategies
infiuence
their confidence to use those strateExpectations for this self-regulated learning developgies.
Teachers
and
parents
can teach students how to
mental trajectory are held for all children and adolesset
goals,
evaluate
their
learning
progress, and seek
cents,, including those with LD, but the results from the
help
to
manage
the
learning
process.
Low levels of selfcurrent study suggest that early adolescents with LD lag
regulatory
efficacy
for
early
adolescents
lead to poor
behind their same-age peers in their confidence to selfperformance
in
school,
school
dropout,
and
may conregulate. Early adolescents with LD face a dual acatinue
to
hamper
development
during
a
person's
work
demic burden: First, their specific academic skills and
life
(Caprara
et
al.,
2008).
self-efficacy are lower, resulting in greater challenges to
Practitioners providing service to adolescents with
successfully complete grade-level academic work.
Second, they may experience deficits in their confi- LD need to focus not only on remediating and comdence to regulate their learning environment. This self- pensating for academic deficits, but also on building
regulatory efficacy deficit has implications for the students' confidence to manage their own learning, .
phases of self-regulated learning; that is, for planning, because self-regulated learning plays an important part
monitoring, and controlling cognition, motivation and in human functioning long past the adolescent years.
affect, and behavior. The end result of lower levels of
REFERENCES
confidence to regulate learning is lowered persistence
Ablard,
K.
E.,
&
Lipschultz,
R. E. (1998). Self-regulated learning in
and less engagement in academic endeavors.
high achieving students: Relations to advanced reasoning,
Limitations and Future Research
achievement goals, and gender. Journal of Educational
This study was based on cross-sectional self-report
Psychology, 90, 94-101.
data, and the results are subject to the biases associated Baird, G. L., Scott, W. D., Dearing, E., & Hamill, S. K. (2009).
Cognitive self-regulation in youth with and without learning
with mono-method research. Future studies should
disabilities: Academic self-efficacy, theories of intelligence,
supplement self-reported self-regulatory efficacy data
learning vs. performance goal preferences, and effort attribuwith observational data from parents and teachers.
tions. Journal of Social and Clinical Psychology, 28, 881-908.
Other research designs that would build insight include Bandura, A. (1990). Multidimensional scales of perceived academic
efficacy. Stanford, CA: Stanford University.
. the use of daily logs in which students keep track of
their confidence to self-regulate over the course of a Bandura, A. (1997). Self-efftcacy: The exercise of control. New York:
Freeman.
semester or academic year.
Bandura, A. (2006). Guide for constructing self-efficaq? scales. In
Further, the LD and NLD samples were geographiF. Pajares & T. Urdan (Eds.), Adolescence and education, Vol. 5:
Self-efcacy and adolescence (pp. 307-337). Greenwich, GT:
cally restricted and may not represent similar populaInformation Age Publishing.
tions in other settings and countries. For example, the
two groups of students in our sample did not differ on Bong, M. (2006). Asking the right question: How confident are
you that you could successfully perform these tasks? In F.
SES and had similar ethnic backgrounds, whereas some
Pajares & T. Urdan (Eds.), Self-efficacy beliefs of adolescents (pp.
studies have shown lower mean SES in LD populations
287-305). Greenwich, GT: Information Age.
(e.g., Melekian, 1990).
Butler, D. L. (2003). Structuring instruction to promote self-

Volume 33, Winter 2010

29

regulated learning by adolescents and adults with learning disstudents with learning disabilities and their non-LD-matched
abilities. Exceptionality, 11, 39-60.
peers. Learning Disabilities Research and Practice, 21, 111-121.
Byrne, B. M. (2004). Testing for multigroup invariance using
Learning Disabilities of Ganada. (2002). Official definition
AMOS graphics: A road less traveled. Structural Equation
of learning disabilities. Retrieved December 8, 2009, from
Modeling, 11, 272-300.
http://www.ldac-taac.ca/Defined/defined_new-e.asp
Caprara, G. V., Fida, R., Vecchione, M., Del Bove, G., Vecchio, G.
Lien, N., Friestad, G., & Klepp, K-I. (2001). Adolescents' proxy
M., Barbaranelli, G., et al. (2008). Longitudinal analysis of the
reports of parents' socioeconomic status: How valid are they?
role of perceived self-efficacy for self-regulated learning in acajournal of Epidemiological Community Health, 55, 731-737.
demic continuance and achievement, journal of Educational Melekian, B. A. (1990). Family characteristics of children with
Psychology, 100(3), 525-534.
dyslexia, journal of Learning Disabilities, 23, 386-391.
Gheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-ofPajares, F., & Vahante, G. (2002). Students' self-efficacy in their
fit indexes for testing measurement invariance. Structural
self-regulated learning strategies: A developmental perspective.
Equation Modeling, 9, 233-255.
Psychologia, 45, 211-221.
Gohen, J. (1988). Statistical power analysis for the behavioral sciences Pintrich, P. R., & Zusho, A. (2002). The development of academic
(2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates.
self-regulation: The role of cognitive and motivational factors.
Daggett, W. R., & Hasselbring, T. S. (2007). Adolescent reading
In A. Wigfield & J. S. Eccles (Eds.), Development of achievement
whitepaper: What we know about adolescent reading. Rexford, NY: motivation (pp. 249-284). San Diego, GA: Academic Press.
International Genter of Leadership in Education.
Schunk, D. H., & Usher, E. L. (in press). Assessing self-efficacy for
Eamon, M. K. (2002). Effects of poverty on mathematics and readself-regulated learning. In B. J. Zimmerman & D. H. Schunk
ing achievement of young adolescents, journal of Early
(Eds.), Handbook of self-regulation of learning and performance.
Adolescence, 22, 49-74.
New York: Routledge.
Eccles, J. S., Midgley, G., Wigfield, A., Buchanan, G. M., Reuman,
Schunk, D. H., & Zimmerman, B. J. (1998). Self-regulated learning:
D., Elanagan, G., & Iver, D. M. (1993). Development during
From teaching to self-refiective practice. New York: Guilford.
adolescence: The impact of stage-environment fit on young
Slife, B. D., Weiss, J., & Bell, T. (1985). Separability of metacogniadolescents' experiences in schools and in families. American
tion and cognition: Problem solving in learning disabled and
Psychologist, 48, 90-101.
regular students, journal of Educational Psychology, 77, 437-445.
Fraser Institute. (2009). School performance. Retrieved October
Statisfics Ganada. (2009). 2006 census dictionary. Retrieved October
21, 2009, from http://www.fraserinstitute.org/reportcards/
21, 2009, from http://wwwl2.statcan.ca/english/censusO6/
schoolperformance/
reference/dictionary/geoOSO.cfm
Garcia-Snchez, J-N., & Fidaigo-Redondo, R. (2006). Effects of two
Usher, E. L., & Pajares, F. (2008). Self-efficacy for self-regulated
types of self-regulatory instruction programs on students with
learning: A validation study. Educational and Psychological
learning disabilities in writing products, processes, and self-effiMeasurement, 68, 443-463.
cacy. Learning Disability Quarterly, 29, 181-211.
Woodcock, R. W., McGrew, K. S., & Mather, N. (2001). WoodcockGraham, S., & Harris, K. R. (2000). The role of self-regulation and
Johnson III Tests of Achievement. Itasca, IL: Riverside Publishing.
writing: Where do we go from here? Educational Psychologist, Zimmerman, B. J. (2000). Attaining self-regulation: A social cog35, 3-12.
nitive perspective. In M. Boekaerts, P. R. Pintrich, & M. Zeidner
Hoffman, J. L., & Lowitzki, K. E. (2005). Predicting college success
(Eds.), Handbook of self-regulation (pp. 13-39). San Diego, GA:
with high school grades and test scores: Limitations for minorAcademic Press.
ity students. The Review of Higher Education, 28, 455-474.
Zimmerman, B. J., Bandura, A., & Martinez-Pons, M. (1992). SelfKlassen, R. M. (2007). Using predictions to learn about the selfmotivation for academic attainment: The role of self-efficacy
efficacy of early adolescents with and without learning disabilbeliefs and personal goal setting. American Educational Research
ities. Contemporary Educational Psychology, 32, 173-187.
journal, 29, 663-676.
Klassen, R. M., Krawchuk, L. L., Lynch, S. L., & Rajani, S. (2008).
Zimmerman, B. J., & Martinez-Pons, M. (1990). Student differProcrastination and motivation of undergraduates with learnences in self-regulated learning: Relating grade, sex, and gifteding disabilities: A mixed methods inquiry. Learning Disabilities
ness to self-efficacy and strategy use. journal of Educational
Research and Practice, 23, 137-147.
Psychology, 82, 51-59.
Klassen, R. M., & Lynch, S. L. (2007). Self-efficacy from the perspective of adolescents with learning disabilities and their spePlease address correspondence about this article to: Robert M.
cialist teachers, journal of Learning Disabilities, 40, 494-507.
Klassen, Department of Educational Psychology, University
Lackaye, T., Margalit, M., Ziv, O., & Ziman, T. (2006).
of Alberta, Edmonton, Alberta, Ganada T6G 2G5;. e-mail:
Gomparisons of self-efficacy, mood, effort, and hope between
robert.klassen@ualberta.ca

Learning Disability Quarterly

30

Copyright of Learning Disability Quarterly is the property of Council for Learning Disabilities and its content
may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express
written permission. However, users may print, download, or email articles for individual use.

Вам также может понравиться