ArticlePDF Available

Development and Initial Validation of the Online Learning Value and Self-Efficacy Scale

Authors:

Abstract

Recently, several scholars have suggested that academic self-regulation may be particularly important for students participating in online learning. The purpose of the present study was to develop a quantitative self-report measure of perceived task value and self-efficacy for learning within the context of self-paced, online training, and to investigate reliability and validity evidence for the instrument. Investigations of this kind are essential because task value and self-efficacy have been shown to be important predictors of students' self-regulated learning competence and academic achievement in both traditional and online contexts. In Study 1 (n = 204), 28 survey items were created for the Online Learning Value and Self-Efficacy Scale (OLVSES) and an exploratory factor analysis was conducted. Results suggested two interpretable factors: task value and self-efficacy. In Study 2 (n = 646), confirmatory factor analysis suggested several survey modifications that resulted in a refined, more parsimonious version of the OLVSES. The resulting 11-item, two-factor scale appears to be psychometrically sound, with reasonable factor structure and good internal reliability. In Study 3 (n = 481), a third sample was collected, and scores from the OLVSES appeared to demonstrate evidence of adequate criterion-related validity. Instrument applications and suggestions for future research are discussed.
J. EDUCATIONAL COMPUTING RESEARCH, Vol. 38(3) 279-303, 2008
DEVELOPMENT AND INITIAL VALIDATION
OF THE ONLIN E LEARNING VALUE
AND SELF-EFFICACY SCALE*
ANTHONY R. ARTINO, JR.
D. BETSY M
CCOACH
University of Connecticut
ABSTRACT
Recently, several scholars have suggested that academic self-regulation may
be particularly important for students participating in online learning. The
purpose of the present study was to develop a quantitative self-report measure
of perceived task value and self-efficacy for learning within the context of
self-paced, online training, and to investigate reliability and validity evidence
for the instrument. Investigations of this kind are essential because task value
and self-efficacy have been shown to be important predictors of students’
self-regulated learning competence and academic achievement in both tradi
-
tional and online contexts. In Study 1 (n = 204), 28 survey items were created
for the Online Learning Value and Self-Efficacy Scale (OLVSES) and an
exploratory factor analysis was conducted. Results suggested two interpret
-
able factors: task value and self-efficacy. In Study 2 (n = 646), confirmatory
*An earlier version of this manuscript was presented at the 2007 annual meeting of the American
Educational Research Association, Chicago, Illinois.
†The first author is a military service member. The views expressed in this article are those of the
authors and do not necessarily reflect the official policy or position of the Department of the Navy,
Department of Defense, nor the U.S. Government.
279
! 2008, Baywood Publishing Co., Inc.
doi: 10.2190/EC.38.3.c
http://baywood.com
factor analysis suggested several survey modifications that resulted in a
refined, more parsimonious version of the OLVSES. The resulting 11-item,
two-factor scale appears to be psychometrically sound, with reasonable
factor structure and good internal reliability. In Study 3 (n = 481), a third
sample was collected, and scores from the OLVSES appeared to demonstrate
evidence of adequate criterion-related validity. Instrument applications and
suggestions for future research are discussed.
INTRODUCTION
With the rapid growth of the Internet, online learning has become a viable
alternative to traditional classroom instruction (Bernard et al., 2004; Tallent-
Runnels et al., 2006). Evolving from previous conceptions of distance education
(Larreamendy-Joerns & Leinhardt, 2006), online learning has emerged as the
format-of-choice for myriad institutions eager to provide students with the
opportunity and convenience of learning from a distance (Moore & Kearsley,
2005). It seems that what was once considered a poor substitute for traditional
classroom instruction has finally entered mainstream education (Moore, 2003;
Moore & Kearsley, 2005).
Evidence of the tremendous growth in online learning is abundant. For instance,
the U.S. Department of Defense, an organization that spends more than $17 billion
annually on military training (United States General Accounting Office, 2003),
recently committed to the development of the Advanced Distributed Learning
(ADL) network. The ADL initiative is designed to capitalize on the capabilities
of computer technology to make education and training available to the military’s
more than three million personnel anytime, anywhere, and online instruction is
considered a critical component of the ADL network (Fletcher, Tobias, & Wisher,
2007). Similarly, postsecondary institutions have recognized the utility of online
learning. A recent survey of 2,200 U.S. colleges and universities by The Sloan
Consortium (2006) found that 96% of large institutions (greater than 15,000 total
enrollments) have some online offerings; 62% of Chief Academic Officers rated
learning outcomes in online education as the same or superior to traditional,
face-to-face instruction; 58% of schools identified online education as a critical
long-term strategy; and overall online enrollment increased from 2.4 million in
2004 to 3.2 million in 2005.
As online learning has grown, so too has interest in academic self-regulation
(Schunk, Pintrich, & Meece, 2008). Academic self-regulation—also known as
self-regulated learning—has been defined as, “an active, constructive process
whereby learners set goals for their learning and then attempt to monitor, regulate,
and control their cognition, motivation, and behavior, guided and constrained by
their goals and the contextual features of the environment” (Pintrich, 2000,
280 / ARTINO AND MCCOACH
p. 453). Self-regulated learners are generally characterized as active participants
who efficiently control their own learning experiences in many different ways,
including establishing a productive work environment and using resources effec
-
tively; organizing and rehearsing information to be learned; regulating their
emotions during academic tasks; and holding positive motivational beliefs about
their capabilities, the value of learning, and the factors that influence learning
(Schunk & Zimmerman, 1994, 1998).
Recently, several scholars have suggested that self-regulated learning skills
may be particularly important for students participating in online courses
(Dabbagh & Kitsantas, 2004; Garrison, 2003; Hartley & Bendixen, 2001). For
example, Dabbagh and Kitsantas (2004) contended that “in a Web-based learning
environment, students must exercise a high degree of self-regulatory competence
to accomplish their learning goals, whereas in traditional face-to-face classroom
settings, the instructor exercises significant control over the learning process and
is able to monitor student attention and progress closely” (p. 40). The authors
recommended that future research on academic self-regulation investigate the
specific learner characteristics that support effective and efficient learning in
these highly autonomous instructional settings.
PURPOSE OF THE STUDY
The purpose of the present study was to develop a quantitative self-report
measure of perceived task value and self-efficacy for learning within the context
of self-paced, online training,
1
and to establish reliability and validity evidence
for the instrument. Investigations of this kind are particularly important because
task value and self-efficacy have been shown to be significant predictors of
students’ use of self-regulated learning strategies and academic achievement in
traditional school settings (Pintrich, 1999). Furthermore, numerous experts have
suggested that these motivational constructs may be even more critical in pre
-
dicting student success in online learning situations (Bandura, 1997; Dabbagh
& Kitsantas, 2004; Garrison, 2003; Hartley & Bendixen, 2001; Schunk &
Zimmerman, 1998). Given that survey research hinges on the legitimacy of the
measurement tools themselves, studying such instruments further informs the
field. The ultimate goal of the present study, then, was to produce a psycho
-
metrically sound instrument that researchers can use to make valid empirical
inferences.
ONLINE VALUE AND SELF-EFFICACY SCALE / 281
1
Self-paced, online courses are a specific type of online training in which students use a Web
browser to access a course management system and complete Web-based courses at their own pace.
While completing these courses, students do not interact with an instructor or other students.
STUDY 1: ITEM DEVELOPMENT AND EXPLORATORY
FACTOR ANALYSIS
Literature Review
Prior to development of the Online Learning Value and Self-Efficacy Scale
(OLVSES), a literature review was conducted to determine if suitable instruments
already existed that could be used to measure task value and self-efficacy for
learning with self-paced, online training.
Task Value
The most current research on the concept of task value comes from expectancy-
value theory and the work of Eccles and Wigfield (1995, 2002). These authors
define task value in terms of four components: attainment value/importance,
intrinsic interest value, extrinsic utility value, and cost. Attainment value (or, more
simply, importance) is defined as the importance of doing well on a task, and is
linked to the relevance of engaging in a task “for confirming or disconfirming
salient aspects of one’s self-schema” (Eccles & Wigfield, 2002, p. 119). Intrinsic
interest is defined as the inherent enjoyment or pleasure one gets from engaging in
an activity, or a person’s subjective interest in the content of a task. Extrinsic
utility value is defined as the usefulness of a task for individuals in terms of their
short- and long-term goals, including academic and career goals. Finally, cost is
conceptualized in terms of the negative aspects of participating in a task, as well as
the amount of effort needed to succeed and the lost opportunities that may result.
Of these four components of perceived task value, attainment, interest, and utility
value are best thought of as “attracting characteristics that affect the positive
valence of the task” (Eccles & Wigfield, 1995, p. 216).
Although the cost component of task value has not been researched as much
empirically, the three positive components have received much attention. In a
study of adolescents in grades 5 through 12, Eccles and Wigfield (1995) examined
the dimensionality of a set of items measuring attainment, interest, and utility
value and found them to be separable in both exploratory factor analysis (EFA)
and confirmatory factor analysis (CFA). In a much earlier study, however,
Parsons (1980) failed to find empirical distinctions between the three positive
components of task value in EFA. It seems that, in this case, participants did not
distinguish between these three theoretically separate constructs.
With the exception of the instrument developed by Eccles and Wigfield
(1995), a review of the literature found very few scales designed to measure the
three positive aspects of perceived task value as separate constructs. And while
the instrument used by Eccles and Wigfield appears to have adequate psycho
-
metric properties, its domain specificity (mathematics) and brevity (the scale
includes only seven items for measuring all three constructs) made the question
-
naire undesirable for the present study.
282 / ARTINO AND MCCOACH
Self-Efficacy for Learning with Self-Paced,
Online Training
Since Bandura’s (1977) seminal article on the self-efficacy component of
social cognitive theory, measures of self-efficacy have become ubiquitous in
educational research. Bandura (1986) defined self-efficacy as, “people’s judg
-
ments of their capabilities to organize and execute courses of action required to
attain designated types of performances” (p. 391). Bandura (1977) hypothesized
that self-efficacy affects an individual’s choice of activities, effort, and per
-
sistence. People who have low self-efficacy for accomplishing a specific task
may avoid it, while those who believe they are capable are likely to participate.
Moreover, individuals who feel efficacious are hypothesized to expend more
effort and persist longer in the face of difficulties than those who are unsure of
their capabilities (Bandura, 1997).
An important aspect of self-efficacy is its domain specificity. That is, people
judge their capability depending on the particular domain of functioning
(Bandura, 2006). Personal efficacy, then, is not a general disposition void of
context, but rather a self-judgment that is specific to the activity domain. As
such, high self-efficacy in one domain does not necessarily indicate high efficacy
in another. Therefore, to achieve predictive power, measures of perceived self-
efficacy must be “tailored to domains of functioning and must represent gradations
of task demands within those domains” (Bandura, 1997, p. 42).
The goal of the efficacy scale included in the OLVSES is to assess the extent
to which students feel confident they can learn effectively using self-paced,
online courseware. A review of the literature failed to uncover any instru-
ments designed to measure perceived self-efficacy for learning with self-
paced, online courseware. However, the review did reveal a number of scales
relating to self-efficacy for using online computer technologies, in general.
For example Brown et al. (2003) developed a technology efficacy scale that
included items associated with the use of various computer technologies in a
synchronous, Web-based learning environment. Unfortunately, this instrument
was quite long, containing more than 35 items, and it did not address many
of the capabilities required of learners in self-paced, online learning environ
-
ments (e.g., logging in to a course management system, navigating through
a computer-based course, and completing a course while dealing with the
distractions of the work environment). Other authors (see, for example, Lynch
& Dembo, 2004; Miltiadou & Savenye, 2003) have developed similar online
technology self-efficacy scales, but again, these instruments were aimed at
assessing efficacy for a slightly different domain of functioning (i.e., Web-based
learning environments that are collaborative in nature and include access to
an instructor and other students). Given the dearth of self-efficacy scales aimed
specifically at self-paced, online training, an original scale was developed in
the present study.
ONLINE VALUE AND SELF-EFFICACY SCALE / 283
Item Development and Content Validation
Based on the results of the literature review, initial items were developed for
each subscale using the following process: a) conceptual definitions were written
for each construct (see Table 1); b) approximately 10 items per construct were
created based on the conceptual definitions; and c) items were compared to similar
scales in the literature and were edited to ensure all aspects of the construct
had been covered. Following initial item development, six content experts were
recruited to participate in a content validation (see recommendations in DeVellis,
2003; McKenzie, Wood, Kotecki, Clark, & Brey, 1999). Each content expert
was provided with the 41 draft items and comprehensive instructions for com
-
pleting the content validation. Content experts were given one week to finish
the validation, which required them to review all items and perform the following
four tasks: a) identify the construct category into which each statement best
fits; b) indicate the certainty of their placement of the statement in the proper
category; c) indicate how relevant they felt each item was for the chosen category;
and d) rate how favorable each item was with respect to the chosen construct.
Additionally, content experts were asked to recommend wording changes for
any items they felt were unclear. Results from the content validation yielded a
28-item instrument designed to measure the four hypothesized latent variables. All
284 / ARTINO AND MCCOACH
Table 1. Construct Categories and Conceptual Definitions for
Each of the Four Subscales
Construct category Conceptual definition
I. Attainment Value/
Importance
II. Intrinsic Interest
Value
III. Extrinsic Utility
Value
IV. Self-efficacy for
Learning with
Self-Paced,
Online Training
Attainment value (or, more simply, importance) is
defined as the importance of doing well on a task in
terms of one’s self-schema and core personal values.
Intrinsic interest value is defined as the inherent
enjoyment or pleasure one gets from engaging in an
activity, or simply a person’s subjective interest in the
content of a task.
Extrinsic utility value is defined as the usefulness of a
task in terms of one’s short- and long-term goals,
including academic and career goals.
Self-Efficacy for Learning with Self-Pased, Online
Training is defined as an individual’s confidence in his
or her ability to successfully learn the material
presented in a self-paced, online learning format.
items employed a 7-point Likert-type response scale ranging from 1 (completely
disagree) to 7 (completely agree).
Participants and Procedures
A convenience sample of 475 personnel from the U.S. Navy were invited
to participate in Study 1. A total of 204 individuals completed the survey
(response rate = 43%); information regarding non-participants was not avail
-
able for analysis. The sample included 150 men (74%) and 53 women (26%);
1 person did not report gender. The mean age of the participants was 39.0 years
(SD = 9.3; range 22-69). Participants reported a wide range of educational
experience, including: High School/GED (n = 21, 10%), Some College (n = 51,
25%), 2-Year College (n = 24, 12%), 4-Year College (B.S./B.A.; n = 25, 12%),
Master’s Degree (n = 48, 24%), Doctoral Degree (n = 15, 7%), and Professional
Degree (n = 16, 8%).
Due to geographical dispersion of the researchers and participants, a Web-based
version of the OLVSES was used in Study 1. Participants were contacted via
e-mail and invited to complete an anonymous, online survey concerning their
experiences with self-paced, online training. Participants were asked to respond
to survey items while keeping in mind what they considered to be the most
effective self-paced, online course they had completed within the last two years.
This approach was necessary because the survey could not be given at the end
of a specific course.
Exploratory Factor Analysis Results
A principal axis factor (PAF) analysis with oblique rotation (Oblimin; delta = 0)
was carried out on the 28 items from the OLVSES using SPSS 13.0 (see factor
analysis recommendations in Preacher & MacCallum, 2003). Oblique rotation
methods allow for factors to be correlated, and the assumption was made that
the four factors thought to be present in the OLVSES were related. Evaluation
of the correlation matrix indicated that it was factorable: Kaiser-Meyer-Olkin
Measure of Sampling Adequacy = .93, which is “marvelous” (> .90) according
to Kasier’s criteria (Pett, Lackey, & Sullivan, 2003). Bartlett’s Test of Sphericity
(!
2
= 4078.48, df = 378, p < .001) was significant, indicating that the correlation
matrix was not an identity matrix, and all measures of sampling adequacy were
deemed sufficient (i.e., > .60; Pett et al., 2003).
The number of factors to extract was determined on the basis of several criteria,
including parallel analysis, examination of the resulting scree plot, and eigen
-
values greater than 1.0 (i.e., the K1 criterion; Hayton, Allen, & Scarpello, 2004).
The parallel analysis suggested that two factors should be retained. Inspection
of the scree plot, although subjective, seemed to suggest two or three factors,
while the K1 criterion suggested four initial factors. Based on these results, it
was determined that three factors would be retained—a reasonable compromise
ONLINE VALUE AND SELF-EFFICACY SCALE / 285
considering the dangers of under-extracting and the tendency for the K1 criterion
to over-extract (Hayton et al., 2004).
The three initial factors extracted accounted for 57.9% of the total variance
in the items. Inspection of the table of communalities revealed that the majority
of the items had high extracted communalities (i.e., > .40; see Table 2), which
indicates that much of the common variance in the items can be explained by
the three extracted factors (Pett et al., 2003). Only four items had low extracted
communalities (i.e., < .40): items SE-2, SE-5, TV-5, and TV-18.
Several rules were used to determine the number of factors and individual
items to be retained in the final solution: a) factors needed to contain at least
three items; b) the absolute value of all factor pattern coefficients needed to be
> .50 on at least one factor; and c) items with factor pattern coefficients (absolute
value) " .30 on more than one factor were dropped (see recommendations in Pett
et al., 2003). The factor pattern and structure coefficients from the PAF analysis
are displayed in Table 2. The rotated pattern and structure coefficients were judged
to have identical factor interpretations (i.e., all items had the strongest coefficients
in the same factor in both rotated matrices), with the pattern matrix generating the
most interpretable simple structure.
The first factor (extraction eigenvalue = 11.86) included 16 items: TV-1, TV-3,
TV-4, TV-6 to TV-17, and TV-19. Although item TV-2 loaded highly on Factor 1,
it also loaded on Factor 3 and was therefore dropped from the final solution.
The second factor (extraction eigenvalue = 3.52) included eight items: SE-1 to
SE-7 and SE-9. Although item SE-8 loaded highly on Factor 2, it also loaded
on Factor 3 and was therefore dropped from the final solution. The third factor
(extraction eigenvalue = 0.83) had no items with pattern coefficients (absolute
value) > .50, and, therefore, Factor 3 was dropped from the final solution. The
correlation between the two remaining factors was .37.
Reliability Analysis
Based on the results of the PAF, a reliability analysis was run on the 16 items
retained in the task value subscale. The Cronbach’s alpha for these 16 items was
.96. However, further inspection of the inter-item correlation matrix revealed
some considerable redundancy in items TV-12 and TV-17. Each of these items
was highly correlated (r > .70) with four other items in the subscale. Therefore,
these two items were deleted. The Cronbach’s alpha for the resulting 14-item
task value subscale was .95 (see Table 3).
Next, a reliability analysis was run on the eight items retained in the self-efficacy
subscale. The Cronbach’s alpha for these eight items was .88. However, further
inspection of the inter-item correlation matrix revealed that item SE-2 had low
correlations (r < .40) with four other items in the subscale. Inspection of the
item-total correlation for the item confirmed this result. Therefore, item SE-2 was
286 / ARTINO AND MCCOACH
deleted. The Cronbach’s alpha for the resulting seven-item self-efficacy subscale
was .89 (see Table 3).
Discussion
Results from the EFA did not reproduce the conceived survey structure. Instead
of four factors, as hypothesized, results suggested only two interpretable factors:
task value and self-efficacy for learning with self-paced, online training. The 14
items that make up the task value subscale (see Table 4) assess the respondent’s
belief that a self-paced, online course is valuable. High scores on this subscale
indicate the person finds the online course interesting, important, and useful.
The seven items that make up the self-efficacy subscale (see Table 4) assess the
respondent’s confidence in his/her ability to learn the material presented in a
self-paced, online course. High scores on this subscale indicate the person is
completely confident he/she can learn the material presented in a self-paced,
online format.
STUDY 2: CONFIRMATORY FACTOR ANALYSIS
In scale development, CFA is used to test whether or not a hypothesized factor
model fits the data (Netemeyer, Bearden, & Sharma, 2003). Confirmatory factor
analysis differs from EFA in that the researcher can “constrain” certain parameters
to predetermined values and “free” others, thereby allowing the analysis to derive
estimates of these model parameters (Thompson, 2004). In Study 2, CFA was used
to test the two-factor solution identified in Study 1.
Participants and Procedures
A convenience sample of approximately 780 undergraduates (sophomores
and juniors) from the U.S. Naval Academy were invited to participate in Study 2.
A total of 646 students completed the survey (response rate = 83%); again,
information regarding non-participants was not available for analysis. The
sample included 514 men (80%) and 113 women (17%); 19 participants (3%)
did not report gender. The mean age of the participants was 20.4 years (SD = 1.0;
range 18-24).
Due to co-location of the researchers and participants, a paper-based version of
the OLVSES was used in Study 2. Following completion of a self-paced, online
course, participants were invited to complete the anonymous survey concerning
their experiences with course. In this case, the online course was the first part of a
two-stage training program in flight physiology and aviation survival training that
was required for all Naval Academy undergraduates. The online course was
composed of four, 40-minute lessons. Each lesson included text, graphics, video,
and interactive activities, as well as end-of-lesson quizzes that consisted of 12 to
15 multiple-choice and true/false, declarative knowledge-type questions.
ONLINE VALUE AND SELF-EFFICACY SCALE / 287
Table 2. Results from the Exploratory Factor Analysis with Oblique Rotation (Oblimin; delta = 0)
Factor
Item Communality 1 2 3
TV-10 I was very intersted in the content of this course.
TV-17 I enjoyed learning the material presented in this online course.
TV-12 The material presented in this course is useful for me to know.
TV-15 It was important for me to learn the material in this course.
TV-11 I felt that doing well in this self-paced, online course was imperative
for me.
TV-16 The knowledge I gained by taking this course can be applied in many
different situations.
TV-3 I will be able to use what I learned in this course in my job.
TV-6 In the long run, I will be able to use what I learned in this course.
TV-7 I really enjoyed completing this self-paced, online course.
TV-4 It was personally important for me to perform well in this course.
TV-2 Understanding the material in this course was important to me.
TV-13 Completing this course moved me closer to attaining my career goals.
TV-9 This course provided a great deal of practical information.
TV-19 Finishing this online course gave me a sense of accomplishment.
TV-8 Performing well in this course made me feel good about myself.
TV-14 This self-paced, online course included many interesting activities.
TV-1 I liked the subject matter of this course.
TV-18 The information I learned in this course has very little use in my daily
life. (REV)
TV-5 This online course was very boring. (REV)
.768
.788
.782
.744
.640
.566
.669
.715
.704
.554
.662
.501
.523
.596
.631
.568
.443
.221
.321
.91 (.87)
.85 (.88)
.85 (.85)
.84 (.83)
.79 (.79)
.78 (.75)
.78 (.79)
.77 (.82)
.75 (.80)
.73 (.74)
.73 (.75)
.72 (.70)
.71 (.72)
.69 (.74)
.68 (.74)
.67 (.71)
.65 (.66)
.50 (.46)
.37 (.45)
–.13 (.21)
.00 (.39)
.00 (.29)
.00 (.27)
.00 (.29)
.00 (.21)
.00 (.31)
.13 (.40)
.13 (.42)
.00 (.30)
.00 (.29)
.00 (.22)
.00 (.30)
.13 (.39)
.16 (.42)
.00 (.35)
.00 (.26)
–.12 (.00)
.20 (.35)
.00 (.00)
–.12 (–.13)
.26 (.25)
.23 (.23)
–.14 (–.15)
.00 (.00)
.22 (.21)
.19 (.817)
–.22 (–.24)
.00 (.00)
.33 (.32)
.00 (.00)
.00 (.00)
–.17 (–.19)
–.25 (–.26)
–.24 (–.25)
.00 (.00)
.00 (.00)
–.28 (–.30)
288 / ARTINO AND MCCOACH
SE-6 I am confident I can do an outstanding job on the activities in a
self-paced, online course.
SE-4 I am confident I can learn without the presence of an instructor to
assist me.
SE-7 I am certain I can understand the most difficult material presented in a
self-paced, online course.
SE-3 Even in the face of technical difficulties, I am certain I can learn the
material presented in an online course.
SE-1 I can perform well in a self-paced, online course.
SE-9 Even with distractions, I am confident I can learn material presented
online.
SE-8 I am confident I can successfully navigate through a self-paced,
online course.
SE-5 I find it difficult to comprehend information presented in a self-paced,
online learning format. (REV)
SE-2 I am confident I can successfully log in to an online course manage-
ment system.
.683
.593
.630
.564
.566
.558
.532
.328
.365
.00 (.36)
.00 (.24)
.00 (.30)
.00 (.24)
.00 (.21)
.00 (.35)
.00 (.30)
.00 (.16)
.12 (.31)
.80 (.82)
.79 (.77)
.78 (.79)
.77 (.75)
.77 (.75)
.68 (.73)
.64 (.64)
.57 (.56)
.54 (.56)
.00 (.00)
.00 (.00)
.00 (–.12)
.00 (.00)
.00 (–.13)
–.15 (–.20)
.35 (.30)
–.11 (–.16)
.19 (.15)
Note: n = 204. Pattern coefficients are presented first, followed by structure coefficients in parentheses. Entries in bold indicate pattern
coefficients >|.50| on at least one factor and pattern coefficients "|.30| on only one factor.
ONLINE LEARNING VALUE / 289
Confirmatory Factor Analysis Results
Using AMOS 7.0, correlations among the 21 OLVSES items were calculated.
Listwise deletion of cases with missing data was used. There were 618 cases
with no missing values on the 21 items. The two latent variables were the two
factors identified by the previous EFA. The 21 observed variables were the
actual OLVSES items. We freely estimated regression weights for 19 of the 21
items (one item per factor served as a marker variable). In addition, we freely
estimated the covariance between the two latent constructs.
The second column of Table 5 provides a summary of the resulting goodness-
of-fit indices for the two-factor model. Chi-square was statistically significant;
Chi-square/degrees of freedom ratio was > 2.0; Tucker Lewis index (TLI) and
comparative fit index (CFI) were < .90; and root mean square error of approxi
-
mation (RMSEA) was > .08; all indicating that the model did not fit the data well
(see recommendations in Hu & Bentler, 1999). Table 6 provides parameter
estimates for the original two-factor model. Overall, the pattern coefficients were
consistent with the hypothesized model. Except for SE-4, which had a pattern
coefficient of –.36, all factor pattern coefficients were moderate (> |.55|) to high
(> |.75|) on their corresponding factor.
In an attempt to improve model fit, standardized residuals and modification
indices (MIs) were examined. Standardized residuals represent differences
between the implied covariance matrix and the observed covariance matrix and
reflect possible sources of model misfit (Netemeyer et al., 2003). Standardized
residuals with absolute values > 2.57 are considered statistically significant
(Netemeyer et al., 2003), and a number of items had standardized residuals greater
than + 2.57. Inspection of the MIs also revealed that these same items would
benefit greatly from the addition of correlated errors (MI > 40.0). However,
because the objective of the present study was to develop a valid and reliable
instrument for measuring task value and self-efficacy, and not simply to produce
the best-fitting model, we chose to trim items based on these results. Therefore,
290 / ARTINO AND MCCOACH
Table 3. Reliability Statistics for Each Subscale in the
Final Solution for Study 1
Cronbach’s
alpha
95% Confidence
Interval
Mean
inter-item
correlations
SD of
inter-item
correlations
Subscale No. items Lower Upper
Task value
Self-efficacy
14
7
.95
.89
.94
.86
.96
.91
.58
.54
.08
.09
ONLINE VALUE AND SELF-EFFICACY SCALE / 291
Table 4. Items Retained in Each Subscale Based
on Results of Study 1
Task Value (TV)
TV-1
TV-3
TV-4
TV-6
TV-7
TV-8
TV-9
TV-10
TV-11
TV-13
TV-14
TV-15
TV-16
TV-19
I liked the subject matter of this course.
I will be able to use what I learned in this course in my job.
It was personally important for me to perform well in this course.
In the long run, I will be able to use what I learned in this course.
I really enjoyed completing this self-paced, online course.
Performing well in this course made me feel good about myself.
This course provided a great deal of practical information.
I was very interested in the content of this course.
I felt that doing well in this self-paced, online course was imperative
for me.
Completing this course moved me closer to attaining my career
goals.
This self-paced, online course included many interesting activities.
It was important for me to learn the material in this course.
The knowledge I gained by taking this course can be applied in many
different situations.
Finishing this online course gave me a sense of accomplishment.
Self-Efficacy for Learning with Self-Paced,
Online Training (SE)
SE-1
SE-3
SE-4
SE-5
SE-6
SE-7
SE-9
I can perform well in a self-paced, online course.
Even in the face of technical difficulties, I am certain I can learn the
material presented in an online course.
I am confident I can learn without the presence of an instructor to
assist me.
I find it difficult to comprehend information presented in a self-paced,
online learning format. (Reverse Coded)
I am confident I can do an outstanding job on the activities in a
self-paced, online course.
I am certain I can understand the most difficult material presented in a
self-paced, online course.
Even with distractions, I am confident I can learn material presented
online.
items TV-1, TV-4, TV-9, TV-14, and SE-4
2
were deleted and the CFA was
run again.
The third column of Table 5 provides a summary of the resulting goodness-
of-fit indices for this revised model. All fit indices improved as a result of this
change, approaching recommended standards. Examination of the MIs again
revealed a number of items with large chi-square change statistics (MI > 20.0),
suggesting the need to add correlated errors between several sets of items. Based
on these results, items TV-2, TV-5, TV-6, TV-11, and SE-1 were deleted and the
CFA was run a final time.
The fourth column of Table 5 provides a summary of the resulting goodness-
of-fit indices for this revised model. Once again, all fit indices improved as a result
of this change. Using the recommended standards of Hu and Bentler (1999),
overall model fit was deemed adequate. The Chi-square/df ratio (3.15) approached
the recommended level of 2.0, TLI (.96) and CFI (.97) were > .95, and RMSEA
(.06) was < .08.
Reliability Analysis
Based on the results of the CFA, reliability analyses were run on the six
items retained in the task value subscale and the five items retained in the
292 / ARTINO AND MCCOACH
Table 5. Fit Indices for Confirmatory Factor Analysis Models Tested in Study 2
CFA Model
Index
Original 2-Factor
Model
2-Factor Model,
TV-1, TV-4, TV-9,
TV14, and SE-4
Deleted
2-Factor Model,
TV-2, TV-5, TV-6,
TV-11, and SE-1
Deleted
Chi-square
df
Probability
Chi-sq/df ratio
TLI
CFI
RMSEA
1292.52
188
.000
6.875
.811
.831
.098
563.43
103
.000
5.470
.882
.899
.085
135.32
43
.000
3.147
.960
.969
.059
Note: TLI = Tucker Lewis Index; CFI = Comparative Fit Index; RMSEA = Root Mean
Square Error of Approximation.
2
In Study 2, the 21 items retained in the OLVSES were renumbered in consecutive order
by subscale.
ONLINE VALUE AND SELF-EFFICACY SCALE / 293
Table 6. Standardized Factor Pattern Coefficients for the
Original 21-Item, Two-Factor Model
Factor
Observed Variables (Items)
Task
Value
Self-
Efficacy
TV-1
TV-2
TV-3
TV-4
TV-5
TV-6
TV-7
TV-8
TV-9
TV-10
TV-11
TV-12
TV-13
TV-14
I liked the subject matter of this course.
I will be able to use what I learned in this course in my job.
It was personally important for me to perform well in this course.
In the long run, I will be able to use what I learned in this course.
I really enjoyed completing this self-paced, online course.
Performing well in this course made me feel good about
myself.
This course provided a great deal of practical information.
I was very interested in the content of this course.
I felt that doing well in this self-paced, online course was
imperative for me.
Completing this course moved me closer to attaining my career
goals.
This self-paced, online course included many interesting
activities.
It was important for me to learn the material in this course.
The knowledge I gained by taking this course can be applied in
many different situations.
Finishing this online course gave me a sense of
accomplishment.
.565
.558
.699
.678
.583
.660
.655
.726
.708
.708
.655
.743
.683
.658
SE-1
SE-2
SE-3
SE-4
SE-5
SE-6
SE-7
I can perform well in a self-paced, online course.
Even in the face of technical difficulties, I am certain I can learn
the material presented in an online course.
I am confident I can learn without the presence of an instructor
to assist me.
I find it difficult to comprehend information presented in a
self-paced, online learning format. (Reverse Coded)
I am confident I can do an outstanding job on the activities in a
self-paced, online course.
I am certain I can understand the most difficult material
presented in a self-paced, online course.
Even with distractions, I am confident I can learn material
presented online.
.713
.713
.760
–.360
.806
.742
.769
Note: Parameter estimates “fixed” to be zero are reported as dashes (—).
self-efficacy subscale; Cronbach’s alphas for the two subscales were .85 and .87,
respectively (see Table 7).
Discussion
Results from the CFA suggested several survey modifications that resulted in
a refined, more parsimonious version of the OLVSES. The resulting 11-item,
two-factor scale appears to be psychometrically sound, with reasonable factor
structure and good internal reliability (see reliability guidelines in Gable &
Wolfe, 1993). Table 8 presents the parameter estimates for the final model, and
Table 9 provides an overall summary of the psychometrics for the two subscales
that make up the OLVSES. The final items retained in each subscale are provided
in Table 10.
STUDY 3: EVIDENCE OF CRITERION-RELATED VALIDITY
To establish the criterion-related validity of the scale, we examined the relation
-
ship among the current scale and several established measures of constructs
known to be critical aspects of academic self-regulation (Schunk et al., 2008).
We felt it was important not only to establish the factorial adequacy of the scale,
but also to establish the relationship of the current scale to other measures of
interest, as well as the utility of the scale for predicting criterion measures,
such as students’ achievement emotions (Pekrun, Goetz, Titz, & Perry, 2002)
and their use of self-regulated learning strategies (Pintrich, Smith, Garcia, &
McKeachie, 1993).
Participants and Procedures
A convenience sample of 481 undergraduates (sophomores and juniors) from
the U.S. Naval Academy
3
were invited to participate in Study 3. All students
completed the survey (response rate = 100%). The sample included 398 men
294 / ARTINO AND MCCOACH
Table 7. Reliability Statistics for Each Subscale in the Final Solution for Study 2
Cronbach’s
alpha
95% Confidence
Interval
Mean
inter-item
correlations
SD of
inter-item
correlations
Subscale No. items Lower Upper
Task value
Self-efficacy
6
5
.85
.87
.84
.85
.87
.89
.50
.57
.06
.06
3
The sample used in Study 3 was composed of different students than the sample from Study 2.
ONLINE VALUE AND SELF-EFFICACY SCALE / 295
Table 8. Standardized Factor Pattern Coefficients for the
Final 11-Item, Two-Factor Model
Factor
Observed Variables (Items)
Task
Value
Self-
Efficacy
TV-3
TV-7
TV-8
TV-10
TV-12
TV-13
It was personally important for me to perform well in this course.
This course provided a great deal of practical information.
I was very interested in the content of this course.
Completing this course moved me closer to attaining my career
goals.
It was important for me to learn the material in this course.
The knowledge I gained by taking this course can be applied in
many different situations.
.650
.716
.745
.669
.797
.675
SE-2
SE-3
SE-5
SE-6
SE-7
Even in the face of technical difficulties, I am certain I can learn
the material presented in an online course.
I am confident I can learn without the presence of an instructor
to assist me.
I am confident I can do an outstanding job on the activities in a
self-paced, online course.
I am certain I can understand the most difficult material
presented in a self-paced, online course.
Even with distractions, I am confident I can learn material
presented online.
.676
.751
.807
.772
.786
Note: Parameter estimates “fixed” to be zero are reported as dashes (—).
Table 9. Subscale Summary Statistics Based on Participant
Scores from Study 2
Subscale name No. items
Cronbach’s
alpha M SD
Subscale
correlation
Task value
Self-efficacy
6
5
.85
.87
5.25
5.16
0.99
1.04
.289*
n = 638
*p < .01
(83%) and 83 women (17%). The mean age of the participants was 20.5 years
(SD = 1.0; range 19-24).
The data-gathering procedures for Study 3 were identical to those for Study 2.
In particular, a paper-based version of the OLVSES was used, and participants
completed the survey after finishing the self-paced, online course in flight physi
-
ology and aviation survival training.
Instrumentation
The survey used in Study 3 was composed of 50 items with a response
scale ranging from 1 (completely disagree) to 7 (completely agree). Along with
the task value and self-efficacy subscales from the OLVSES, four additional
296 / ARTINO AND MCCOACH
Table 10. Items Retained in Each Subscale Based
on Results of Studies 1 and 2
Task Value (TV)
TV-3
TV-7
TV-8
TV-10
TV-12
TV-13
It was personally important for me to perform well in this course.
This course provided a great deal of practical information.
I was very interested in the content of this course.
Completing this course moved me closer to attaining my career
goals.
It was important for me to learn the material in this course.
The knowledge I gained by taking this course can be applied in many
different situations.
Self-Efficacy for Learning with Self-Paced,
Online Training (SE)
SE-2
SE-3
SE-5
SE-6
SE-7
Even in the face of technical difficulties, I am certain I can learn the
material presented in an online course.
I am confident I can learn without the presence of an instructor to
assist me.
I am confident I can do an outstanding job on the activities in a
self-paced, online course.
I am certain I can understand the most difficult material presented in a
self-paced, online course.
Even with distractions, I am confident I can learn material presented
online.
subscales were included in the survey (see subscale descriptions below). All of
the variables derived from this survey were created by computing means of the
items associated with a particular subscale.
Negative Achievement Emotions
Two subscales adapted from Pekrun, Goetz, and Perry (2005) were used to
assess students’ negative achievement emotions: 1) a five-item boredom subscale
intended to assess students’ course-related boredom; and 2) a four-item frustra
-
tion subscale designed to assess students’ course-related frustration, annoyance,
and irritation.
Cognitive and Metacognitive Learning Strategies
Students’ self-reported use of two learning strategies was assessed with items
derived from the Motivated Strategies for Learning Questionnaire (MSLQ)
(Pintrich et al., 1993): 1) a four-item elaboration subscale designed to assess
students’ use of elaboration strategies (e.g., paraphrasing and summarizing); and
2) a nine-item metacognitive self-regulation subscale intended to assess students’
use of metacognitive control strategies (e.g., planning, setting goals, monitoring
one’s comprehension, and regulating performance).
Results
Pearson Correlations
Table 11 presents descriptive statistics for the six variables measured in
Study 3. Additionally, Table 11 presents results from the correlation analysis. As
indicated, task value and self-efficacy were statistically significantly related to
each other (r = .34, p < .001) and to students’ negative achievement emotions
and use of cognitive and metacognitive learning strategies. As expected, the
extent to which students valued the online course was negatively related to their
boredom (r = –.50, p < .001) and frustration (r = –.47, p < .001) with the course.
Moreover, students’ task value was positively related to their use of elaboration
(r = .59, p < .001) and metacognitive strategies (r = .62, p < .001). Likewise,
students’ self-efficacy for learning online was negatively related to their boredom
(r = –.31, p < .001) and frustration (r = –.30, p < .001), and positively related
to their use of elaboration (r = .27, p < .001) and metacognitive strategies (r = .20,
p < .001).
Regression Analyses
To explore the unique variance explained by students’ motivational beliefs on
their negative achievement emotions and use of self-regulated learning strategies,
four multiple regressions were conducted. In these analyses, boredom, frustration,
ONLINE VALUE AND SELF-EFFICACY SCALE / 297
elaboration, and metacognitive self-regulation were used as the dependent vari-
ables; task value and self-efficacy served as the independent variables. Table 12
presents a summary of the regression analyses for each dependent variable. As
expected, task value and self-efficacy were both statistically significant negative
predictors (# = –.45 and –.16, respectively) of boredom, accounting for approxi-
mately 27% of its variance, F(2, 478) = 89.39, p < .001. Results from the second
analysis predicting frustration indicate that the two predictors accounted for
approximately 24% of its variance, F(2, 478) = 76.56, p < .001. Again, both task
value (# = –.42, p < .001) and self-efficacy (# = –.16, p < .001) were statistically
significant individual predictors of the outcome.
Results from the regression analyses predicting students’ use of elaboration and
metacognitive strategies were consistent, in part, with expectations. Specifically,
task value and self-efficacy accounted for approximately 36% of the variance in
elaboration, F(2, 474) = 131.41, p < .001. However, task value was the only
statistically significant individual predictor of elaboration (# = .57, p < .001);
self-efficacy only approached significance (# = .08, p = .051). Likewise, results
from the final regression analysis predicting metacognitive self-regulation indi
-
cate that task value and self-efficacy accounted for approximately 38% of its
variance, F(2, 474) = 145.52, p < .001. Again, task value (# = .62, p < .001) was the
only statistically significant individual predictor of the outcome.
Discussion
Pekrun’s (2000, 2006) social cognitive, control-value theory of achievement
emotions outlines hypothesized linkages between students’ motivational beliefs,
their achievement emotions, and, ultimately, their learning and performance.
According to Pekrun’s theory, positive achievement emotions, such as enjoyment
298 / ARTINO AND MCCOACH
Table 11. Descriptive Statistics, Cronbach’s Alphas, and Pearson
Correlations for the Measured Variables
Variable n M SD $ 123456
1. Task value
2. Self-efficacy
3. Boredom
4. Frustration
5. Elaboration
6. Metacognitive
self-regulation
481
481
481
481
477
477
4.96
5.29
3.82
3.41
4.58
4.09
1.04
1.10
1.23
1.37
1.04
1.10
.89
.92
.86
.88
.82
.92
.34
–.50
–.31
–.47
–.30
.71
.59
.27
–.37
–.30
.62
.20
–.41
–.33
.65
Note: All correlations are significant at the p < .001 level.
and hope, and negative emotions, such as boredom and frustration, are deter-
mined, in part, by students’ motivational beliefs, also known as their cognitive
appraisals. Among the many categories of motivational beliefs that may be
relevant, Pekrun (2000, 2006) has suggested that two appraisals are most impor-
tant in achievement contexts: the subjective value of achievement activities
(e.g., task value), and the perceived controllability of those activities, as indi-
cated by competence perceptions (e.g., self-efficacy). Results from Study 3 seem
to support this theoretical perspective, indicating that students’ boredom and
frustration can be explained, in part, by their task value and self-efficacy beliefs,
as measured by the OLVSES.
Social cognitive theories of self-regulation (Pintrich, 1999; Zimmerman, 2000)
stress the importance of students’ motivational beliefs in all phases of self-
regulation. For example, Pintrich and De Groot (1990) argued that “knowledge
of cognitive and metacognitive strategies is usually not enough to promote student
achievement; students also must be motivated to use the strategies as well as
regulate their cognition and effort” (p. 33). Findings from Study 3 partially
support these theoretical assumptions, indicating that students’ task value and
self-efficacy beliefs are positively related to their use of self-regulated learning
strategies in online academic settings. However, when it comes to explaining
students’ self-reported use of learning strategies in the self-paced, online context
studied here, results suggest that students’ task value beliefs may be more impor
-
tant than their self-efficacy for learning in these highly autonomous, online
learning environments.
GENERAL DISCUSSION AND CONCLUSIONS
In their previous work on expectancy-value theory, Eccles and Wigfield
(1995, 2002) found that attainment, interest, and utility value were separable in
ONLINE VALUE AND SELF-EFFICACY SCALE / 299
Table 12. Regression Summary Statistics with Boredom, Frustration,
Elaboration, and Metacognitive Self-Regulation as the Dependent Variables
Boredom Frustration Elaboration
Metacognitive
self-regulation
Variable B SE B # B SE B # B SE B # B SE B #
Task value
Self-efficacy
–.53
–.18
.05
.05
–.45*
–.16*
–.55
–.20
.06
.05
–.42*
–.16*
.56
.07
.04
.04
.57*
.08
.65
–.01
.04
.04
.62*
–.01
Model
summaries
R
2
= .27* R
2
= .24* R
2
= .36* R
2
= .38*
*p < .001
both EFA and CFA. Results from the present investigation did not support this
conclusion. Instead, these findings suggest that participants did not differentiate
between the three theoretically distinct components of task value. That being said,
results of the present study are consistent with much earlier research by Parsons
(1980), who conducted an EFA and also found no empirical distinctions between
the three positive components of task value.
Although factor analysis failed to confirm the hypothesized four-factor model,
the resulting two-factor scale appears to demonstrate evidence of adequate
construct validity, criterion-related validity, and internal consistency reliability.
Therefore, the ultimate goal of the present study was achieved; namely, to produce
a psychometrically sound survey for measuring respondents perceived task value
and self-efficacy with respect to self-paced, online learning. Using the OLVSES,
researchers interested in self-regulated learning have another tool for exploring
the relations between online learners’ motivational beliefs, as measured by task
value and self-efficacy, and other dependent variables of interest, such as students’
achievement emotions, their use of various learning strategies, and their overall
academic achievement. Understanding these relationships is important because,
as Pintrich (1999) argued in his review of the extant literature on academic
self-regulation, “the research reviewed here suggests very clearly that certain
types of motivational beliefs are adaptive and do help to promote and sustain
self-regulated learning” (p. 467). Future research in online contexts would do well
to investigate these mechanisms further and determine if the relationships found in
conventional, face-to-face classrooms generalize to online learning environments.
A significant limitation of the present study was the difference in demography
between the samples used in Study 1 and Studies 2 and 3. Participant differences
may be noteworthy if one considers recent appeals to investigate developmental
differences in learners’ motivational beliefs and self-regulatory skills (Greene &
Azevedo, 2007; Pintrich, 2003). Therefore, to improve the generalizabilty of our
results, future validation of the OLVSES should include replication of the current
study on a more diverse population. Moreover, future research should examine
other forms of validity evidence, such as convergent and discriminant validity,
which could ultimately improve the functioning of the instrument. Finally,
additional evidence of construct validity should be gathered. This could be
accomplished, for example, by assessing the extent to which the OLVSES
subscales relate to more direct, behavioral measures of academic success, such as
observations of students’ use of self-regulated learning strategies and assessments
of their actual academic performance (i.e., their course grades) in self-paced,
online situations.
REFERENCES
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84, 191-215.
300 / ARTINO AND MCCOACH
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice-Hall.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W. H. Freeman and
Company.
Bandura, A. (2006). Guide for constructing self-efficacy scales. In F. Pajares &
T. Urdan (Eds.), Adolescence and education, Vol. 4: Self-efficacy beliefs of adoles
-
cents. Greenwich, CT: Information Age Publishing.
Bernard, R. M., Abrami, P. C., Lou, Y., Borokhovski, E., Wade, A., Wozney, L., et al.
(2004). How does distance education compare with classroom instruction? A
meta-analysis of the empirical literature. Review of Educational Research, 74,
379-439.
Brown, S. W., Boyer, M. A., Mayall, H. J., Johnson, P. R., Meng, L., Butler, M. J.,
et al. (2003). The GlobalEd project: Gender differences in a problem-based learning
environment of international negotiations. Instructional Science, 31, 226-248.
Dabbagh, N., & Kitsantas, A. (2004). Supporting self-regulation in student-
centered Web-based learning environments. International Journal on E-Learning,
3(1), 40-47.
DeVellis, R. F. (2003). Scale development: Theory and applications (2nd ed.). Thousand
Oaks, CA: Sage Publications.
Eccles, J. S., & Wigfield, A. (1995). In the mind of the actor: The structure of adolescents
achievement task values and expectancy-related beliefs. Personality and Social
Psychology Bulletin, 21, 215-225.
Eccles, J. S., & Wigfield, A. (2002) Motivational beliefs, values, and goals. Annual
Review of Psychology, 53, 109-132.
Fletcher, J. D., Tobias, S., & Wisher, R. A. (2007). Learning anytime, anywhere: Advanced
distributed learning and the changing face of education. Educational Researcher,
36(2), 96-102.
Gable, R. K., & Wolfe, M. B. (1993). Instrument development in the affective domain:
Measuring attitudes and values in corporate and school settings. Boston, MA: Kluwer
Academic Publishers.
Garrison, D. R. (2003). Self-directed learning and distance education. In M. G. Moore &
W. G. Anderson (Eds.), Handbook of distance education (pp. 161-168). Mahwah, NJ:
Lawrence Erlbaum Associates.
Greene, J. A., & Azevedo, R. (2007). A theoretical review of Winne and Hadwin’s model
of self-regulated learning: New perspectives and directions. Review of Educational
Research, 77, 334-372.
Hartley, K., & Bendixen, L. D. (2001). Educational research in the Internet age:
Examining the role of individual characteristics. Educational Researcher, 30(9),
22-26.
Hayton, J. C., Allen, D. G., & Scarpello, V. (2004). Factor retention decisions in explor
-
atory factor analysis: A tutorial on parallel analysis. Organizational Research
Methods, 7, 191-205.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Structural Equation Modeling,
6, 1-55.
Larreamendy-Joerns, J., & Leinhardt, G. (2006). Going the distance with online education.
Review of Educational Research, 76, 567-605.
ONLINE VALUE AND SELF-EFFICACY SCALE / 301
Lynch, R., & Dembo, M. (2004). The relationship between self-regulation and online
learning in a blended learning context. [Electronic version]. International Review
of Research in Open and Distance Learning, 5(2).
McKenzie, J. F., Wood, M. L., Kotecki, J. E., Clark, J. K., & Brey, R. A. (1999).
Establishing content validity: Using qualitative and quantitative steps. American
Journal of Health Behavior, 23, 311-318.
Miltiadou, M., & Savenye, W. C. (2003). Applying social cognitive constructs of moti
-
vation to enhance student success in online distance education. Association for the
Advancement of Computing in Education Journal, 11(1), 78-95.
Moore, M. G. (2003). Preface. In M. G. Moore & W. G. Anderson (Eds.), Handbook of
distance education (pp. ix-xii). Mahwah, NJ: Lawrence Erlbaum Associates.
Moore, M. G., & Kearsley, G. (2005). Distance education: A systems view (2nd ed.).
Belmont, CA: Wadsworth.
Netemeyer, R. G., Bearden, W. O., & Sharma, S. (2003). Scaling procedures: Issues and
applications. Thousand Oaks, CA: Sage Publications.
Parsons, J. E. (1980). Self-perceptions, task perceptions and academic choice: Origins
and change. Unpublished final technical report to the National Institution of Education,
Washington, DC. (ERIC Document Reproduction Service No. ED186477.)
Pekrun, R. (2000). A social cognitive, control-value theory of achievement emotions. In
J. Heckhausen (Ed.), Motivational psychology of human development (pp. 143-163).
Oxford, England: Elsevier.
Pekrun, R. (2006). The control-value theory of achievement emotions: Assumptions,
corollaries, and implications for educational research and practice. Educational
Psychology Review, 18, 315-341.
Pekrun, R., Goetz, T., & Perry, R. P. (2005). Achievement emotions questionnaire (AEQ):
User’s manual. Munich, Germany: University of Munich, Department of Psychology.
Pekrun, R., Goetz, T., Titz, W., & Perry, R. P. (2002). Academic emotions in students’
self-regulated learning and achievement: A program of qualitative and quantitative
research. Educational Psychologist, 37, 99-105.
Pett, M. A., Lackey, N. R., & Sullivan, J. J. (2003). Making sense of factor analysis: The
use of factor analysis for instrument development in health care research. Thousand
Oaks, CA: Sage Publications.
Pintrich, P. R. (1999). The role of motivation in promoting and sustaining self-regulated
learning. International Journal of Educational Research, 31, 459-470.
Pintrich, P. R. (2000). The role of goal orientation in self-regulated learning. In
M. Boekaerts, P. R. Pintrich, & M Zeidner (Eds.), Handbook of self-regulation
(pp. 451-502). San Diego, CA: Academic.
Pintrich, P. R. (2003). A motivational science perspective on the role of student motivation
in learning and teaching contexts. Journal of Educational Psychology, 95, 667-686.
Pintrich, P. R., & De Groot, E. V. (1990). Motivational and self-regulated learning
component of classroom academic performance. Journal of Educational Psychology,
82, 33-40.
Pintrich, P. R., Smith, D. A. F., Garcia, T., & McKeachie, W. J. (1993). Reliability and
predictive validity of the Motivated Strategies for Learning Questionnaire (MSLQ).
Educational and Psychological Measurement, 53, 801-813.
Preacher, K. J., & MacCallum, R. C. (2003). Repairing Tom Swift’s electric factor
analysis machine. Understanding Statistics, 2, 13-43.
302 / ARTINO AND MCCOACH
Schunk, D. H., Pintrich, P. R., & Meece, J. L. (2008). Motivation in education: Theory,
research, and applications (3rd ed.). Upper Saddle River, NJ: Pearson Education.
Schunk, D. H., & Zimmerman, B. J. (Eds.). (1994). Self-regulation of learning and
performance: Issues and educational applications. Hillsdale, NJ: Lawrence Erlbaum
Associates.
Schunk, D. H., & Zimmerman, B. J. (Eds.). (1998). Self-regulated learning: From teach
-
ing to self-reflective practice. New York: The Guilford Press.
The Sloan Consortium. (2006, November). Making the grade: Online education in the
United States, 2006. Retrieved March 14, 2007, from
http://www.sloan-c.org/publications/survey/pdf/making_the_grade.pdf
Tallent-Runnels, M. K., Thomas, J. A., Lan, W. Y., Cooper, S., Ahern, T. C., Shaw, S. M,
& Liu, X. (2006). Teaching courses online: A review of the research. Review of
Educational Research, 76, 93-135.
Thompson, B. (2004). Exploratory and confirmatory factor analysis: Understanding con
-
cepts and applications. Washington, DC: American Psychological Association.
United States General Accounting Office. (2003). Military transformation: Progress and
challenges for DOD’s advanced distributed learning programs (GAO Publication
No. 03-393). Washington, DC: Author.
Zimmerman, B. J. (2000). Attaining self-regulation: A social cognitive perspective. In
M. Boekaerts, P. R. Pintrich, & M Zeidner (Eds.), Handbook of self-regulation
(pp. 13-39). San Diego, CA: Academic.
Direct reprint requests to:
Dr. Anthony R. Artino, Jr.
Department of Preventive Medicine and Biometrics
Uniformed Services University of the Health Sciences
4301 Jones Bridge Rd.
Bethesda, MD 20814
e-mail: tony_artino@yahoo.com
ONLINE VALUE AND SELF-EFFICACY SCALE / 303
... Self-efficacy was measured with the Online Learning Value and Self-Efficacy Scale (OLVSES), which was developed by Artino and McCoach (2008) for self-paced online learning and composed of 5 items (Cronbach's alpha = 0.87). All the five items were modified and translated into Chinese, and the Cronbach's alpha reached 0.881. ...
Article
Full-text available
Objective Few research efforts have substantially introduced relevant studies on Chinese students’ adaptability in relation to the ineffectiveness of blended learning mode in College English. This study is guided by social cognitive theory, and related literature has been reviewed concerning adaptability. In this study, we aim to examine the involved relationships among contextual factors, self-efficacy, motivation, and adaptability to blended learning mode among non-English majored Chinese learners in the College English course. Methods The quantitative research method was employed in this study, and 595 respondents were eventually collected to enable the data analysis. The structural equation modeling (SEM) approach was used to analyze the data. Results (1) The non-English majored learners’ adaptability to blended learning mode in College English was at a low level with a mean value of 2.26, indicating that students still suffer difficulties from class conducted in blended learning; (2) the fit indices were at the level of good fit, and it suggested the structural model had an overall good fit to the data as shown: x 2/ df = 2.496, RMESA = 0.050, GFI = 0.956, AGFI = 0.936, NFI = 0.968, RFI = 0.959, IFI = 0.980, TLI = 0.975, CFI = 0.980; (3) adaptability was positively related to contextual factors, self-efficacy, and motivation ( p < 0.001); (4) contextual factors exerted an indirect effect on the adaptability not only through the separate mediating role of self-efficacy and motivation, but through the chain mediating role of self-efficacy and motivation ( p < 0.01). Implications This study theoretically extends previous studies on adaptability by investigating the affecting factors in the framework of social cognitive theory. More practically important is that this study sheds light on the impact mechanism of positive and enjoyable environment, self-efficacy and motivation on non-English majored learners’ adaptability to blended learning mode in English course, which would provide a vital insight for administrators and College English instructors to reconsider the role of learners’ responses in the mixed mode to improve their English achievement more effectively, as well as to enhance and promote their user experience of the offered blended learning service.
... The authors stated that the scale was created by examining the scales used in previous studies (Arbaugh et al., 2008;Artino and McCoach, 2008;Barnard, Lan, To, Paton and Lai, 2009;Martin and Tutty, 2009;Miltiadou and Yu, 2000;Pintrich and DeGroot, 1990). Unlike the scales used in previous studies, the fact that all items in the scale used in this study are positive will prevent the respondents from getting confused (Netemeyer, Bearden and Sharma, 2003). ...
Article
Full-text available
Bu çalışmada, öğretmen adaylarının çevrim içi öğrenmeye yönelik öz-yeterlilik düzeyleri belli değişkenler açısından kesitsel tarama yöntemiyle araştırılmıştır. Çalışmanın örneklemi, İç Anadolu bölgesinde yer alan 4 farklı üniversitenin Eğitim Fakültesinin İlköğretim Matematik Öğretmenliğinde öğrenim gören toplam 562 öğretmen adayıdır. Çalışmada Sun ve Rogers (2020) tarafından geliştirilen çevrim içi öğrenmeye yönelik öz yeterlilik ölçeği Türkçeye uyarlanarak kullanılmıştır. Öğretmen adaylarının çevrimiçi öğrenme öz-yeterlik ölçeğinin teknoloji kullanımı, çevrimiçi öğrenme görevi, eğitici ve akran etkileşimi ve iletişimi ve öz düzenleme ve motivasyon alt boyutlarındaki puan ortalamaları iyi düzeyde bulunmuştur. Sadece teknoloji kullanımı öz-yeterlik düzeyleri açısından kız öğrenciler lehine anlamlı bir farklılığın olduğu tespit edilmiştir. Öğretmen adaylarının eğitici ve akran etkileşimi ve iletişimi öz-yeterlik düzeyleri açısından 1. ve 3. sınıf arasında 1. sınıf lehine ve 3. ve 4. sınıf arasında 4. sınıf lehine anlamlı bir farklılığın olduğu tespit edilmiştir. Diğer boyutlar açısından sınıf düzeyleri arasında anlamlı bir farklılık tespit edilememiştir.
... Motivational beliefs are defined as important psychological elements that encourage learners to be cognitively engaged in a learning task, remain effortful, invest more time, and persist in their learning and academic pursuits (Bandura 1986;Wigfield, Eccles, Fredericks, Roeser, Schiefele, & Simpkins, 2015;Metallidou and Vlachou 2010;Pintrich 2003;Wong et al. 2021;Wu et al. 2019;Zimmerman 2011). Various motivational beliefs, such as self-efficacy, expectations, and task values, have a propensity to facilitate individuals' use of self-regulated strategies during their learning process (Artino and McCoach 2008;Lee, Watson, and Watson 2020), enhance quality of individuals' task performance (Eccles 2005;Wong et al. 2021) and promote positive learning behaviours and academic performance, particularly in the context of language learning (Bandura 1986;Trautwein et al. 2012;Wigfield and Eccles 2000;Zheng, Liang, and Tsai 2017). In particular, motivational beliefs about academic selfefficacy and task values are crucial psychological factors directing students enrolling in practice-based educational programmes that require participation in theoretical learning and clinical practice experience (e.g. ...
... The belief that one has the ability to cope with the tasks necessary to achieve one's objective is particularly important tp effective learning online (Artino & McCoach, 2008;Lee, Choi, & Kim, 2013;Bradley et al., 2017). Self-efficacy at least partly derives from an ability to manage one's behavior and produce positive and acceptable results (self-regulation) as Lynch and Dembo (2004) and Hodges and Kim (2010) have indicated and the degree to which one can interact with colleagues and peers (social isolation) as Shen et al. (2013) and Artino and Jones (2012) have reported. ...
... Researchers have emphasized the importance of assessing the teaching efficacy of the teachers (e.g., Barni et al. 2019;Daniel, 2020;Di Pietro, et al. 2020;Kraft et al. 2020;Larsen & Samdal, 2012;Rice, 2003). However, other than the general teaching efficacy, most instruments were developed in the context of online teaching (e.g., Artino & McCoach, 2008;Corry & Stella, 2018;Gosselin, 2016;Horvitz et al. 2014;Lazar et al. 2020;Tschannen-Moran & Woolfolk, 2001;Zimmerman & Kulikowich, 2016). It should be noted that remote teaching, as emphasized by Hodges et al. (2020) and considering also the context of most Philippine schools, is an appropriate term to be used in emergency cases rather than online teaching. ...
Article
Full-text available
This paper describes the level of remote teaching efficacy of teachers amid the COVID-19 crisis and assesses the psychometric properties of the proposed Remote Teaching Efficacy (RTE) Scale. A 10-item scale originally used in the context of online teaching was modified to suit the context of the current massive migration to emergency remote teaching. A sample of 1,061 K to 12 employed Filipino teachers were surveyed. The items of the scale were subjected to tests of internal consistency and factor analysis. The result showed that, despite the global crisis affecting the educational system, the teachers have a high level of remote teaching efficacy. On the other hand, RTE Scale demonstrated acceptable indices of corrected item-total correlations between .73 and .82. A high overall Cronbach’s alpha at .95 was also sought, supported by excellent factor loadings between .79 and .84. Thus, this paper concludes that RTE Scale is a reliable and valid scale that can be used to measure the remote teaching efficacy of teachers.
Article
Full-text available
Many universities across the world had incorporated e-learning systems into traditional education, creating a blend education. The success of implementing e-learning systems require intensive understanding of users' acceptance. This paper focused on the investigation of students, who had adopted e-learning system at universities in Kuala Lumpur. Specifically, this study is conducted to examine on the influence of flexibility, system design and internet self-efficacy on adoption of e-learning. Self-administered questionnaire consisting of 393 sets were distributed as the research's data collection tools but only 385 sets are qualified. The conceptual framework of e-learning adoption used in this study is adopted from Technology Acceptance Model. This study also provides an indicator of learners' acceptance as well as identifying possible factors that influence the adoption process. The results will enhance the understanding of students' acceptance of e-learning and will benefit education institutions in creating qualified e-learning systems.
Article
Full-text available
The study aimed to (1) identify the differences between male and female students with hearing impairment in self-compassion, academic flow, Online learning self-efficacy, and academic self-efficacy (2) determine the relationship between the variables of these study (3) determine the direct and indirect effect of self-compassion on academic flow. The study sample included 57 students with hearing impairment at King Khalid University (M=21.86; SD=1.23). A study used a short scale of self-compassion, by Raes, Pommier, Neff & Gucht (2011), Researcher Arabization, Online Learning Self-Efficacy Scale by Zimmerman & Kulikowich (2016), Researcher Arabization, and Academic Flow Scale and Academic Self Efficiency Scale by the researcher, The results showed (1) no significant differences between the male and female students in self-compassion, academic flow, Online learning self-efficacy, and academic self-efficacy, (2) there is a significant positive correlation between the self-compassion, academic-flow, online learning self-efficacy, and academic self-efficacy, (3) the possibility of predicting with the academic-flow, Online Learning self-efficacy, and academic self-efficacy by Self-compassion, and (4) the results of the path analysis showed a significant positive direct effect of self-compassion on the academic-flow and a significant indirect effect of self-compassion on the academic flow through the mediating role of both the online learning self-efficacy and the academic self-efficacy.
Book
Knowledge Management makes the management of information and resources within a commercial organization more effective. The contributions of this book investigate the applications of Knowledge Management in the upcoming era of Semantic Web, or Web 3.0, and the opportunities for reshaping and redesigning business strategies for more effective outcomes. Explores the synergies between Knowledge Management and Web 3.0
Article
Full-text available
This chapter reviews the recent research on motivation, beliefs, values, and goals, focusing on developmental and educational psychology. The authors divide the chapter into four major sections: theories focused on expectancies for success (self-efficacy theory and control theory), theories focused on task value (theories focused on intrinsic motivation, self-determination, flow, interest, and goals), theories that integrate expectancies and values (attribution theory, the expectancy-value models of Eccles et al., Feather, and Heckhausen, and self-worth theory), and theories integrating motivation and cognition (social cognitive theories of self-regulation and motivation, the work by Winne & Marx, Borkowski et al., Pintrich et al., and theories of motivation and volition). The authors end the chapter with a discussion of how to integrate theories of self-regulation and expectancy-value models of motivation and suggest new directions for future research.
Article
Presents an integrative theoretical framework to explain and to predict psychological changes achieved by different modes of treatment. This theory states that psychological procedures, whatever their form, alter the level and strength of self-efficacy. It is hypothesized that expectations of personal efficacy determine whether coping behavior will be initiated, how much effort will be expended, and how long it will be sustained in the face of obstacles and aversive experiences. Persistence in activities that are subjectively threatening but in fact relatively safe produces, through experiences of mastery, further enhancement of self-efficacy and corresponding reductions in defensive behavior. In the proposed model, expectations of personal efficacy are derived from 4 principal sources of information: performance accomplishments, vicarious experience, verbal persuasion, and physiological states. Factors influencing the cognitive processing of efficacy information arise from enactive, vicarious, exhortative, and emotive sources. The differential power of diverse therapeutic procedures is analyzed in terms of the postulated cognitive mechanism of operation. Findings are reported from microanalyses of enactive, vicarious, and emotive modes of treatment that support the hypothesized relationship between perceived self-efficacy and behavioral changes. (21/2 p ref)