Myint Swe Khine · Tine Nielsen

Editors

Academic Self-efficacy
in Education
Nature, Assessment, and Research
Chapter 2
Assessing Academic Self-efficacy

Maria K. DiBenedetto and Dale H. Schunk

Abstract Academic self-efficacy is a dynamic motivational belief that influences the

goals we set, how hard we persist, and the amount of effort we employ. There is a great
deal of research supporting the link between self-efficacy and student achievement
yet educators grapple with finding ways to increase students’ capability beliefs to
enhance motivation for learning and performance. Traditionally, self-efficacy has
been assessed using surveys administered as pre and posttests to a learning event.
While these measures provide valuable information, there are several concerns about
measuring self-efficacy using surveys such as that learners are reporting on future
or past events, they may not fully understand or anticipate task demands at the time
of pretest or over and under estimate their performance during the posttest. More
recently, research has demonstrated there are other methods of assessing self-efficacy
during a learning event referred to as real time such as the microanalysis, think-aloud,
diaries, and trace measures. This chapter will focus on these novel approaches to
assess self-efficacy and make recommendations on the ways these methodologies
can be used among educators.

Keywords Academic self-efficacy · Motivational beliefs · Student achievement ·

Real time · Metacognition

Introduction

Self-efficacy, which is grounded in Bandura’s (1986, 1997) social cognitive theory,

refers to one’s perceived capabilities to learn or perform actions at designated levels.
Self-efficacy is predicted to influence motivation, learning, achievement, and self-
regulation (Schunk & DiBenedetto, 2020; Usher & Schunk, 2018). Although there are
multiple motivational variables that are linked to achievement (e.g., intrinsic interest,

M. K. DiBenedetto (B) · D. H. Schunk

University of North Carolina at Greensboro, Greensboro, USA
e-mail: [email protected]
D. H. Schunk
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 11
M. S. Khine and T. Nielsen (eds.), Academic Self-efficacy in Education,
https://doi.org/10.1007/978-981-16-8240-7_2
12 M. K. DiBenedetto and D. H. Schunk

goal orientation, and outcome expectations), in this chapter, our focus is on self-
efficacy. Self-efficacy has seen wide application in diverse fields such as education,
business, athletics, and health. Because of its predictive power and widespread use,
it is important to find valid and reliable ways to assess self-efficacy across different
types of contexts and activities.
The purpose of this chapter is to discuss various self-efficacy assessment method-
ologies. We begin with a description of the theoretical framework of social cognitive
theory including the role of self-efficacy in achievement. This is followed by a discus-
sion of self-efficacy assessments, educational implications of these assessments, and
recommendations for future research. Our hope is that this chapter will help to expand
research on assessing self-efficacy.

Theoretical Framework

Triadic Reciprocity

Social cognitive theory is grounded in a model of triadic reciprocality comprising

three sets of interacting processes: personal; behavioral; and environmental (Bandura,
1986). Personal processes include cognitions, beliefs, perceptions, and emotions.
They help to instigate and sustain motivational outcomes. Goals, values, outcome
expectations, attributions, emotions, cognitions, and self-efficacy are examples of
personal processes that influence behavior and the environment and are influenced
by them.
Behavioral processes are actions and verbalizations; environmental processes
include influences from the physical and social environments. Importantly, social
cognitive theory stresses the idea that people use vicarious, symbolic, and self-
regulatory processes to strive for a sense of agency, or the belief that they can exert
a large degree of control over important events in their lives.
The reciprocal nature of these influences can be illustrated with self-efficacy—
a personal process. With respect to the interaction of self-efficacy and behavior,
research shows that self-efficacy instigates achievement behaviors such as task
choice, effort, persistence, and use of effective strategies (Usher & Schunk, 2018).
These behaviors affect self-efficacy. As students work on tasks and observe their
progress, their self-efficacy for continued learning is enhanced.
The link between personal and environmental processes can be shown with
students with learning disabilities, many of whom hold low self-efficacy for learning
(Schunk & DiBenedetto, 2014). People in their environments may react to them
based on their common attributes (e.g., low competencies) rather than on their actual
capabilities. Environmental feedback can influence self-efficacy, as when teachers
provide encouragement.
The interaction of behavioral and environmental processes can be seen in instruc-
tion when teachers announce for students to direct their attention to a display. They
2 Assessing Academic Self-efficacy 13

may do it without much conscious attention. The influence of behavior on environ-

ment is evident when learners fail to grasp important concepts, after teachers reteach
content rather than moving on.

Dimensions and Sources of Self-efficacy

Social cognitive theory postulates that self-efficacy varies on several dimensions that
have important implications for understanding how it operates during student learning
(Bandura, 1997). Self-efficacy differs in level—the nature of the task demands, in
generality—a wide range of activities or only a specific activity, and in strength—
the degree to which one feels self-efficacious to perform an activity successfully. For
example, Tabatha may feel self-efficacious about preforming well in a jazz recital
but terrified about performing well at her ballet recital. The level in this situation
is the capability to perform certain movements at an expert level in jazz versus
ballet. Tabatha’s feelings are not general, they are specific in that she feels capable
of dancing jazz and not ballet. The strength of her self-efficacy is high for jazz and
low for ballet. In designing an instrument to assess learners’ motivation, researchers
must have an understanding of what it takes to succeed at the task.
Self-efficacy does not emerge from nowhere but rather is a cognitive process
where learners use information sources to create their self-efficacy beliefs (Schunk &
DiBenedetto, 2020). These sources can be vicarious experiences, forms of social
persuasion, physiological and emotional indexes, and mastery experiences (Usher &
Pajares, 2008). Social cognitive theory postulates that by observing a successful
model, one’s self-efficacy can be raised, just as it can be lowered by observing
someone fail. Forms of social persuasion include verbal statements and feedback
from others (e.g., a coach telling a student she can catch the ball during a game of
baseball). Physiological and emotional indexes can also affect self-efficacy. Feeling
the thrill of going downhill on skis without falling can enhance a beginner’s self-
efficacy beliefs for repeating the activity successfully.
While these three sources of self-efficacy influence a learner’s capability beliefs,
the most enduring source is what one can accomplish (Usher & Pajares, 2008;
Zimmerman & DiBenedetto, 2008). Learners’ appraisal of their self-efficacy beliefs
based on their past achievements and failures influences their self-efficacy for future
similar activities (Bandura, 1986). Students who experience success at completing
a complicated science experiment, for example, are likely to feel self-efficacious in
performing well on similar future laboratory experiments. Self-efficacy is dynamic in
that it develops and changes as students become more capable and achieve at higher
levels (DiBenedetto & Schunk, 2018).
Usher and Pajares (2008) examined the sources of self-efficacy across quantitative
and qualitative school-based studies and found that while mastery experiences were
the most influential source of self-efficacy, other contextual factors must be taken into
consideration. In the following section, we expand on the various measures that have
been used to assess self-efficacy and recommend employing real-time measures.
14 M. K. DiBenedetto and D. H. Schunk

Self-efficacy Assessment Methods

Since Bandura’s (1986) seminal book discussing self-efficacy, developing reliable

and valid assessments of self-efficacy has been an important issue. In this section,
we discuss several methodologies for assessing self-efficacy and have grouped them
within two categories. In the first category, we have organized traditional instruments
for assessing self-efficacy. We are calling traditional assessments those instruments
that require learners to reflect on a learning event and recall from memory. These
assessments are usually administered after a learning event has taken place and
typically are outside of the learning context; for example, administering tenth graders
a survey asking questions about how self-efficacious they felt about computing long
division problems after they completed a test on long division.
Traditional assessments may also refer to instruments that require learners to
respond to questions that are asked in anticipation of an activity. These are also based
on prior self-efficacy experiences and beliefs and assessed outside of the learning
event; for example, administering a survey to fifth graders about their self-efficacy
to earn 100% on an upcoming World War II history exam. Regardless of whether
students are rating their motivation based on a past event or on the expectation of
performance on a future event, self-efficacy beliefs are assessed using questionnaires
that are completed outside of the realm of the learning context.
In the second category, we refer to real-time assessments, methodologies used
during a learning event. These assessments often occur within the context of an
authentic learning situation, in other words, while the learning is taking place. Real-
time assessments do not require the learner to recall a previous learning event or to
anticipate a future learning event; for example, asking students during a gymnastics
exercise how self-efficacious they feel about being able to walk across the high beam
without falling off while they are walking on the high beam.
We begin each category with a description of self-efficacy assessment methodolo-
gies we deem to fall within that category. Following the description, we provide an
example, discuss the reliability and validity, and then, the strengths and limitations of
that approach. While we describe several approaches, we make a case for using real-
time assessments of self-efficacy, a recommendation also made by Bandura (1997)
Table 2.1.

Traditional Assessments of Self-efficacy

Surveys

A common method for assessing self-efficacy is the use of surveys, inventories, or

questionnaires where students respond to questions using some form of a Likert
Table 2.1 Summary of Self-Efficacy and Self-Regulated Learning Assessment Methodologies
Self-Efficacy and Traditional versus real time Characteristics Strengths Limitations
self-regulated learning
measurement methods
Surveys Traditional • Questions typically on a • Cost and time effective • Questions may not capture
Likert scale • Ease in administration and level, strength, and specificity
• May include other variables analysis of context
related to self-efficacy and all • Flexibility in delivery • Questions may not reflect
three phases of self-regulated • Larger sample size permits understanding of task
learning greater generalizability demands
• Self-report based on recall • Typically easy to establish • No opportunities provided for
reliability and validity elaboration
2 Assessing Academic Self-efficacy

• Dependent on students’
memories and ability to recall
• Retrospective assessment
questions
• Reliability frequently
established; questionable
construct validity due to the
nature of the survey
development and the domain
specificity
(continued)
15
Table 2.1 (continued)
16

Self-Efficacy and Traditional versus real time Characteristics Strengths Limitations

self-regulated learning
measurement methods
Diaries Traditional • Open-ended responses as • Easy to use and familiar to • Younger students may have
participants write about their most participants difficulty articulating
thoughts, feelings, and • Provide data rich in thoughts, feelings, and
behaviors qualitative information behavior
• Typically focused on the • Flexibility in administration • Data analysis can be
forethought and • Students can elaborate in complicated and time
self-reflection phase their entries consuming
• Self-report based on recall • Reliability and validity can be • Diaries administered over
established and strengthened extended periods of time may
through triangulation affect student motivation to
participate
• Sample sizes tend to be small
• Performance phase measures
not taken
• Dependent on students’
memories and ability to recall
• Retrospective results
• Issues of reliability and
validity require triangulation
adding addition time, effort,
and cost
(continued)
M. K. DiBenedetto and D. H. Schunk
Table 2.1 (continued)
Self-Efficacy and Traditional versus real time Characteristics Strengths Limitations
self-regulated learning
measurement methods
Case Studies Both • Involve closely studying a • Provide in-depth • Do not have a standard
participant or participants understanding about protocol for administration
• Include a holistic approach participants resulting in concerns of
that uses other assessment • Allow for a combination of sloppiness, bias, and inability
instruments such as those real-time and traditional to replicate
described in this table measures • Require long periods of time
• Attempt to answer the how • Can assess and make and large amounts of data
and why of behavior inferences on the dynamic causing issues of time, cost,
2 Assessing Academic Self-efficacy

changing nature of analysis, and labor intensity

self-efficacy over time • Samples tend to be small
• Can be done to include all leading to challenges in
three phases of self-regulated generalizing of results
learning and multiple • Difficulty estimating
iterations of the cycle during reliability and validity on its
learning events own since data from case
• Reliability and validity can be studies frequently come from
established and strengthened multiple sources; ecological
through triangulation validity
(continued)
17
Table 2.1 (continued)
18

Self-Efficacy and Traditional versus real time Characteristics Strengths Limitations

self-regulated learning
measurement methods
Traces Real Time • Involve technology • Assessment conducted during • May be expensive due to the
recordings of students’ real-time learning event cost of technology and
behavior during a learning • Can be used to “teach” or program to implement
event coach students toward more • Generate large amounts of
• Consist of an unobtrusive positive outcomes data that need to be
way to assess learning • Can be adapted to learner’s streamlined for interpretation
• Inferences can be made about capability to problem-solve and analysis
students’ cognitions and • Lack of interference in • Self-efficacy is difficult to
motivational beliefs based on protocol assures participant is assess in traces
tracking not distracted or influenced • Generalizability to other
by questioning learning situations and
• Provide rich data learners is questionable
• Reliability and validity
frequently estimated through
triangulation
Think-Alouds Real time • Participant thinks aloud while • Include a standard protocol • Concern of reactivity effect to
engaged in a learning task which is repeatable learning prompts
• Occurs in a one-on-one • Assessment occurs during • Best when used with other
setting real-time learning event measures of self-efficacy and
• Vocalizations, movements, • Researcher can provide other covert processes
and eye behavior are recorded practice simulations so that • Typically a small sample size
simultaneously while learning learner understands the • Artificial setting such as a
• Able to assess several of the protocol prior to taking place laboratory
processes in the three phases • Provide diagnostic value for • Triangulation recommended
of self-regulated learning intervention to establish reliability and
validity
(continued)
M. K. DiBenedetto and D. H. Schunk
Table 2.1 (continued)
Self-Efficacy and Traditional versus real time Characteristics Strengths Limitations
self-regulated learning
measurement methods
Microanalyses Real time • Participants are asked fine • Include a standard protocol • Limited studies in academic
grained questions targeted at with questions that assess contexts have been conducted
each of the processes in the covert and overt behavior using this methodology
three-phase cycle during a • Provides both quantitative • Most studies conducted have
learning event and qualitative data been with high school or
• Students respond to closed • Provide diagnostic value for university students—need
and open-ended questions intervention research on younger
• May be conducted • May be used with small- or populations
2 Assessing Academic Self-efficacy

one-on-one or in small groups medium-sized samples • Interrupting students during

• Can be conducted in a • Easy to implement and learning events may influence
classroom or laboratory provides a moderate amount outcomes
to data • Limited information on the
• Little financial cost is reliability and validity since
associated with this this assessment methodology
assessment is in its infancy
• Assessment occurs during
real-time learning event
• Have demonstrated reliability
and validity as an assessment
instrument
19
20 M. K. DiBenedetto and D. H. Schunk

scale (Wolters et al., 2011; Wolters & Won, 2018). Surveys are self-report assess-
ments whereby the students respond to prompts that are designed to elicit informa-
tion regarding their thoughts, beliefs, perceptions, attitudes, feelings, and knowledge
(Wolters & Won, 2018).
Prior to the construction, preliminary information on the task demands should be
identified to develop reliable and valid scales (Bandura, 2000). This information can
be obtained by focus groups, piloting questionnaires, and conducting open-ended
interviews. This step is critical in survey construction because the surveys are asking
participants to judge their capability to complete a task and researchers may not be
experts on the subtle gradations of difficulties for task completion. It also ensures the
survey is domain and task specific and not generalized across content. For example,
asking an 11th grader to rate his self-efficacy beliefs for earning a 90 on an upcoming
French vocabulary test is different than asking him to rate his self-efficacy for earning
a 90 in French at the end of the school year.
Example. DiBenedetto and Bembenutty (2013) examined the dynamic nature of
self-efficacy over the course of a semester in 113 undergraduate students enrolled in
intermediate-level science courses. Surveys assessed students’ self-efficacy beliefs
for learning and performing well in their science courses and were administered at
the beginning of the semester and then again at the end. Students were also asked
questions about their socialization experiences in childhood and adolescence. Find-
ings revealed that students’ self-efficacy beliefs declined at the end of the semester
and that these beliefs more positively predicted final science grades than earlier
appraisals of self-efficacy.
Reliability and Validity. In survey construction, researchers seeking to estimate reli-
ability typically use test–retest, alternate-form, or measures of internal consistency
(Litwin, 1995). Test-retest typically involves having the same group of students take
the same survey at two different points in time to determine how stable the responses.
Correlation coefficients are then obtained to determine how consistent the responses
are and should be 0.70 or higher. Alternate form involves creating two surveys with
different questions to measure the same attribute or using a simpler approach such as
changing the order of the questions and the responses. These two forms are admin-
istered to the same group of students but at different points of time. Analysis of
this approach involves using the split-halves method whereby one compares the
random selection of one half of one survey to a random selection of one half of the
alternate-form survey.
The third approach to ensuring reliability is obtaining a measure of internal consis-
tency. Internal consistency involves determining how well different questions assess
the same issue (Litwin, 1995). Internal consistence is typically measured using Cron-
bach’s coefficient alpha. Typically, a score of 0.70 or higher is an indication that
the different questions are measuring the same variable. For example, in the study
cited earlier by DiBenedetto & Bembenutty (2013), Cronbach’s coefficient alphas
for self-efficacy were 0.79 and 0.72 for pre- and post-assessments suggesting each
self-efficacy questionnaire was a reliable instrument.
2 Assessing Academic Self-efficacy 21

Face, content, criterion, and construct (both, convergent and discriminant) validity
are ways in which one may determine if the survey created is assessing what it is
supposed to be (Litwin, 1995). Face validity refers to a more casual approach, for
example, developing a survey on self-efficacy for learning algebra and then asking
a friend to review the survey to determine if it appears to be assessing self-efficacy
for learning algebra.
Content validity is also a subjective approach to assessing validity (Litwin, 1995).
In this approach, a researcher may ask colleagues or other experts in the field to
review the survey to determine if it appears to be assessing the content at-hand.
While this approach can be helpful in providing revisions to the questions, it also not
considered scientific. The third approach, criterion validity, involves determining
how the survey compares to other instruments assessing the same construct and
obtaining a correlation coefficient. If the correlation is high, it suggests the survey
has high criterion validity. Lastly, construct validity refers to five types of evidence
that are based on the content, response processes, internal structure, relations to
other variables, and consequences of testing (Litwin, 1995; Wolters & Won, 2018).
Construct validity provides information about how meaningful the survey is when
put into practice. Discriminant validity refers to evidence that two similar constructs
are actually different from one another.
Lent et al. (1997) examined the discriminant and predictive validity of academic
self-efficacy along with mathematics-specific self-efficacy and academic self-
concept among 205 university students who completed a survey. Findings supported
the previous research indicating that self-efficacy was a strong predictor of math-
related choice and achievement when it is specifically matched to a task and that it
is different from academic self-concept.
Studies have also been conducted to estimate the reliability and validity of self-
efficacy scales for learning across disciplines as well as the sources of self-efficacy
(Usher & Parajes, 2008).
Strengths. There are several advantages to using surveys that explains their wide
use. They are easy to develop and administer, require little effort on the part of the
participants, and are time and cost effective. Students may complete the surveys at
any physical location (such as the classroom, auditorium, at home), and they may be
administered through paper and pencil instruments, or online using electronic forms
(Wolters & Won, 2018). Another significant strength of surveys is that they allow
researchers to reach a larger sample of students than would be possible using other
methodologies. For example, a principal wanting to know whether students in her
school feel self-efficacious about using a particular computer program to complete
school assignments may administer a survey to all students in her school at one time
in her gymnasium; or a researcher interested in learning about student motivation
for online learning may administer a survey to all students enrolled in undergraduate
psychology courses in the country.
Surveys also lend themselves to allowing researchers to obtain information on
variables that may influence motivation. Many surveys have questions that ask for
demographic information such as age, gender, race, ethnicity, parental education, and
extracurricular activities that allow researchers to study the association between these
22 M. K. DiBenedetto and D. H. Schunk

variables with student motivation. These additional variables, while self-reported, are
much easier to obtain using a survey than through going through student records in
a school. For several of these reasons, surveys are among the most widely used
methodology for obtaining information on student motivation and learning.
Limitations. Surveys have important limitations. Bandura (2000) emphasized the
importance of not using a “one measure fits all” approach when developing scales to
assess self-efficacy. Because surveys tend to be easy to develop and administer, they
are often crafted in a way that has limited connection to specific context. Without
conducting focus groups, piloting questions, or open-ended interviews, researchers
may have an ambiguous understanding of the task demands, which affects whether
the questions are accurately capturing the level, strength, and specificity dimensions
of self-efficacy. An additional disadvantage is that surveys tend to rely on students’
memories and the ability to recall past experiences in an unbiased, honest way
(Bandura, 1997; Winnie & Perry, 2000). Bandura (1997) emphasized the validity of
surveys in their demonstrated success in predicting the triadic factors of social cogni-
tive theory. Unfortunately, self-efficacy scales are administered, results are reported,
and wide array of assumptions is often made based on the findings including a lack
of support for the predictability of self-efficacy for future learning and performance.
An additional limitation stems from the nature of the format of surveys. Because
surveys are structured, they do not provide opportunities for students to elaborate,
explain, or qualify their responses (Wolters & Won, 2018). Participants are often not
able to ask for clarification, and responses are based entirely on students’ interpre-
tations of the questions and memories over time and context. Because surveys are
so widely used, the necessity of estimating the reliability and validity and piloting
the survey prior to administration falls on the survey developer and may often be
overlooked or unreported.

Diaries

Diaries, often referred to as journaling, have been a useful tool for assessing self-
efficacy and in highlighting gradual within person changes in self-efficacy beliefs
(Schmitz et al., 2011; Wallen & Adawi, 2017). Because diaries are kept over a period
of time, they have been found to help increase students’ self-efficacy beliefs because
they can reflect small daily increments in skill (Schmitz et al., 2011). Diaries may
also help researchers examine differences across participants, their trajectories, and
what processes or events may underlie these changes (Iida et al., 2012). Diaries may
be unstructured, allowing the students the freedom to record their thoughts or feelings
without any prompts or guidance from the instructor (Fritson, 2008). Students may
also receive prompts of what to record in their diaries. For example, learners can
be instructed to use diaries to record learning processes and reflections on learning
outcomes. Baleghizadeh and Mortazavi (2014), for example, used journaling to see
if the impact of teacher or peer feedback influenced students’ self-efficacy beliefs
for learning English as a second language and found that both types of journaling
2 Assessing Academic Self-efficacy 23

improved students’ self-efficacy as compared to students who did not receive any
feedback. Regardless of the style of journaling, the primary purpose is to get students
to think, reflect, and record their thoughts and feelings (Fritson, 2008). Traditionally,
diaries are associated with teenagers writing their secret feelings in a book (Schmitz
et al., 2011); however, in scientific research, diaries can be used to measure processes
that occur over a short intervention period or over an extended period of time using
paper and pencil journals or online tools.
Example. Fritson (2008) conducted a study on college-level psychology students’ use
of journaling on self-efficacy beliefs and feelings about locus of control. One class of
students received a ten-minute discussion on cognitive behavioral strategies such as
self-talk, visual imagery, and distorted thinking and was provided with a journal entry
template encouraging students to write about their thoughts and feelings. The students
in the other class did not receive the instruction or template and were simply told to
record their impressions, beliefs, and thoughts about the course content. Students in
both courses showed an improvement in self-efficacy beliefs over the semester, and
the author suggested that the practice of writing in the journals had a positive impact
on students’ self-efficacy, regardless of whether they received additional information
on cognitive behavioral strategies.
Reliability and Validity. Because diaries are often used immediately following
a learning activity, their accuracy should be higher than surveys (Schmitz et al.,
2011). The two most common approaches to estimating reliability involve estab-
lishing internal consistency and test–retest. Internal consistency can be estimated
by stopping several times during the school year, for example, to examine within
person data entries or responses to prompts. Diaries can also be tested for reliability
using the test–retest approach or split-half reliability approach to test the stability
of time series data (Iida et al., 2013; Schmitz & Skinner, 1993). In the test–retest
approach, diaries are examined at two different points in time but before instruction
or an intervention has taken place and then they correlated with one another (Iida
et al., 2013). In the split-half approach, the diaries can be broken down in the first
half and the second half of the diary collection period, or into even and odd days
to examine consistencies in feelings, thoughts, and reflections (Schmitz & Skinner,
1993).
Face validity and convergent validity are two approaches used to estimate validity
in diaries (Iidia et al., 2013). Others have examined the construct validity of diaries
by comparing them to other measures (Carp & Carp, 1981). Schmitz et al. (2011)
also indicate that diaries would be predicted to have ecological validity because when
students write in their journals, they are typically closer in time to the learning event
and thus require less retrospective than surveys.
Carp and Carp (1981) examined the reliability, validity, and generalizability of
diary data over a one-day period in which participants responded to interview ques-
tions and a one-week period in which diaries were kept. Participants were retired
residents, ages 60 and over, and interview questions targeted at ascertaining the
number of times the participants left their home and for what purposes. Findings
24 M. K. DiBenedetto and D. H. Schunk

revealed that the one-week diaries had construct validity for salient activities and
were less influenced by retrospective bias than were interview data.
Strengths. There are several advantages to using diaries. Diaries are easy to use as
most students are familiar with the use of diaries as a means of recording feelings or
record keeping. Diaries provide sequential information which makes it possible to
capture changes that occur in student motivation over time. They provide researchers
with a richness of information from the students’ perspective as they may elaborate
on their motivation, feelings, and thoughts regarding the learning events. Diaries also
provide information on students’ problem-solving strategies (Fogarty & McTighe,
1993) and may provide insight into students’ trajectories (Schmitz et al., 2011).
Diaries may also be administered using paper and pencil or Web-based method-
ologies making them accessible in various environments. Diaries are flexible and
allow for both cross-sectional and longitudinal data to be analyzed (Schmitz et al.,
2011). Finally, teachers and students may complete diaries simultaneously and use
them to develop and refine intervention techniques that may help build self-efficacy
beliefs.
Limitations. Diaries are typically used with students who are in middle school and
older because they must be able to articulate their thoughts and feelings into words
to enter them into the diaries. Younger students, therefore, may have difficulty using
diaries to describe motivational processes in relation to learning events. Diaries are
likely to involve a much smaller sample size, tend to be more time demanding on
the both the students and the researchers, and rely heavily on self-reports. Another
limitation concerns the willingness and motivation of the students in filling out the
diaries which may affect the quality of the data obtained (Schmitz et al., 2011).
A common recommendation to compensate for these limitations in addition to
promoting reliability and validity is to use other sources of information for triangu-
lation that can lead to additional cost, time, and effort for both learners and researchers
(Carp & Carp, 1981). Lastly, and most importantly, while diaries may be completed
close to the actual learning event, they are not completed during a real-time learning
event and are based on the students’ cognitions and feelings after the event has
occurred.

Case Studies

The case study methodology is used when researchers want to gain an in-depth
understanding of students within an authentic context (Butler & Cartier, 2018; Yin,
2014). Case studies provide a wide breadth of information that can include a variety
of methodologies to measure motivation and learning such as observations, inter-
views, student performance measures, diaries, surveys, think-aloud protocols, the
microanalysis, and stimulated recalls. Stimulated recalls involve video recording the
participant while learning, then while viewing the recording asking the participant
questions about the processes of self-regulated learning such as their self-efficacy
beliefs (Schunk, 2020).
2 Assessing Academic Self-efficacy 25

Because case studies attempt to provide a “holistic” view of participants, they

seek to explain the how and why questions about a contemporary set of events in a
situation in which the researcher has little or no control (Yin, 2014, p. 14). According
to Yin (2014), when conducting a case study, one must have a set of case study
questions and its propositions, identified participants or entity (such as a classroom),
an understanding of the data to be obtained, and a criterion for interpreting the
findings. Versland and Erickson (2017), for example, conducted a case study on a
principal of a high achieving but impoverished high school to examine whether his
self-efficacy for achievement influenced the collective self-efficacy of the teachers
and students. Findings revealed the principal’s high self-efficacy beliefs motivated
teachers and students and impacted their self-efficacy for success.
As in the example above, case studies can be designed to capture the individual
processes such as self-efficacy, metacognition, and strategy use (Butler & Cartier,
2018). Because self-efficacy is dynamic and changing, case studies that are conducted
over extended periods of time can capture these changes within naturalist learning
contexts as they develop and evolve. This can provide researchers with information
on subtle changes in self-efficacy beliefs, as well as the relationship between these
changes and instruction and performance. While we placed case studies under the
traditional instrument category, one could argue they capture real-time measures,
particularly if part of the case studies involves the structured interview, think-aloud,
and microanalytic methodologies where students are asked questions during learning.
Example. Scott (2011) conducted a study exploring the impact of self-efficacy and
self-regulated learning instruction on students with varying levels of achievement,
including students with disabilities. Seven students working on literacy tasks were
observed while their teachers implemented supportive instruction. These students
who were in either inclusive, supportive, or pull-out classes for literacy instruction
were part of the case study analysis along with four teachers. The four teachers
in the study were supportive of implementing self-regulated learning instruction in
their classrooms. Multiple data collection measures such as observations, running
records, interviews, and probes were obtained to ensure a more holistic view. Findings
revealed that students’ self-efficacy varied across literacy tasks and was related to
personal factors such as past reading difficulties and environmental factors such as
their perceptions about the writing task.
Reliability and Validity. Reliability of case studies is often estimated by using a
standard case study protocol and through establishing measures of evidence (Yin,
2014). Developing a standard protocol consists of asking questions that get at the
issues under investigation, are unbiased, and generate data that can be interpreted.
Data from case studies may come from multiple sources. Maintaining a strict protocol
can ensure consistency across cases. Examining multiple sources of data provides
evidence through triangulation. Patton (2002) describes four different sources of data
that can be used to estimate reliability: multiple data sources, multiple investigators
conducting cases studies on the same students, agreement of the theoretical applica-
tions by the investigators to the data, and standard methodology for conducting the
case study.
26 M. K. DiBenedetto and D. H. Schunk

Musyoka et al. (2017) conducted a case study on 40 teachers’ perceptions of

their capability to teach students who are hard of hearing with additional disabilities.
Thirty-one of the teachers had certification in teaching students with deafness, the
remaining 9 had at last one student who was hard of hearing in their classes. Teachers
were administered a survey that included comprehensive open-ended questions in
addition to closed-ended questions using a Likert scale. Reliability was established
using triangulation with other data sources and intercoder reliability. Overall findings
revealed that teachers did not feel prepared due to lack of information on various
disabilities.
Validity can be estimated through construct validity, internal validity, and external
validity (Yin, 2014). Construct validity can be examined using multiple sources of
evidence; internal validity can be found by examining patterns and rival explanations
to explain outcomes; and external validity can be estimated through the application
of theory in single-case studies or replication logic in multiple-case studies.
Ruokonen (2018) conducted a case study on tenth grade students’ self-efficacy
and social skills through an arts education program over a one-year period. Twenty
students from four participating schools participated in the arts program, while
another twenty, also from four participating schools, remained in the control group.
Data were collected using students’ evaluations before and after the school year
using Likert scale surveys with previously established validity, along with teachers’
responses to open-ended questionnaires to provide insight and explanation to
students’ responses. The focus of the arts project was for students to create their
own life stories using artistic projects such as videos, painting, dance, drama, or
music and included instruction from artists as well as several visits to museums and
cultural arts centers. Students who participated in the arts program showed no signifi-
cant differences in self-efficacy and but some significant differences on socialization
skills. The author suggests that these findings may be due to a lack of full support
by the schools’ administrations.
Strengths. Case studies have the potential to provide a comprehensive picture of
participants in their natural settings. Different types of data collection methods may be
used to help inform the understanding of student learning and motivation. In addition,
one may also include students’ records, feedback from teachers, parental information,
medical information, previous learning experiences, and any other relevant material
to help inform the data analysis. Unlike diaries that do not capture learning while it
is occurring, case studies provide the opportunity to observe multiple motivational
and learning processes such as self-efficacy and strategy use as they unfold (Butler &
Cartier, 2018).
Another advantage of using case studies is that they provide diagnostic infor-
mation about where in instruction a future intervention may be used to help build
students’ self-efficacy beliefs. An additional strength of case studies is that it may be
possible to make inferences about changes in self-efficacy over time. For example, in
observing a first grader over the school year learn to write in cursive one may witness
an increase in her self-efficacy by her mannerisms, excitement, flair when writing,
or changes in the way she holds her pen. Overall, case study designs that include
2 Assessing Academic Self-efficacy 27

real-time assessments have the potential to provide rich qualitative and quantitative
data on student self-efficacy and self-regulated learning.
Limitations. Yin (2014) indicates that unlike other assessment methodologies, case
studies typically do not have a standard systematic protocol to follow which results
in concern over sloppiness, not following systematic procedures, or the potential
for bias in analyzing the results and in drawing conclusions. Another limitation is
the confusion on what exactly a case study is and what data should be collected as
part of the study. Case studies tend to require long periods of time during which
the researcher acquires large amounts of data to sort through and analyze which can
be overwhelming, complex, and confusing. While case studies have the potential
of providing a wealth of information critical to understanding the dynamic nature
of self-efficacy, findings from case studies are difficult to interpret and generalize
to other populations because they are typically specific to the student, classroom,
or school and based on observations by the researcher which may involve issues of
subjectivity.

Real-Time Assessments of Self-efficacy

Traces

One approach to assessing student motivation is through the use of technology.

Technology can be designed in a way that captures multiple forms of inputs such as
students’ selection of menu options, dropdown bars, navigation, clicks, eye move-
ments, use of online coaches, time spent on and off task, facial expressions, physiolog-
ical changes, among others (Azevedo et al., 2018; Bernacki, 2018). When learners
use technology to complete learning events, these data sources can be traced and
recorded in a log that contains a transcript that is referred to as trace data. (Bernacki,
2018).The trace data are then interpreted and analyzed in ways to understand the
learner’s use of motivational, cognitive, and affective processes to foster learning
during a real-time event and when incorporating the use of Webcams can provide
additional rich information through video recordings. (Azevedo et al., 2018). Traces
include marks students make in texts, such as when they underline, highlight, circle
key words, or write notes in margins. These traces can also be captured through
technology.
Traces can be used to examine the amount of effort and persistence a student puts
into an assignment or toward reaching a goal (Winne & Stockley, 1998). When
a student continues to conduct research on a topic even though she has repeat-
edly not found what she is searching for on the Web, she is engaging in effort
and persistence as compared to another student who surfs the Web and after a few
minutes of unsuccessful searches gives up. Bandura (1986) indicated that learners
who are self-efficacious will apply effort and persist when pursuing goals. Traces
28 M. K. DiBenedetto and D. H. Schunk

allow researchers to examine cognitive monitoring as they observe learners apply

different tactics online to regulate their learning (Winne & Stockley, 1998).
Example. Bernacki et al. (2015) examined changes in self-efficacy among ninth
grade students studying algebra using an intelligent tutoring system that traced and
recoded their behaviors. The trace log information included attempts at problem-
solving, use of tools available related to the content, and responses to pop-up prompts
designed to assess students’ self-efficacy and problem-solving skills. Self-efficacy
prompts that consisted of questions asking students to report on their level of efficacy
for completing similar mathematics problems in the future were presented using a
Likert scale. Findings revealed that self-efficacy changed in response to learning.
In addition, as students’ math solutions became more accurate, not only did their
self-efficacy improve, but their use of help-seeking also declined.
Reliability and Validity. While traces can only be interpreted in terms of the learning
context, they tend to be reliable because they refer to objective measures. Traces can
also be estimated as reliable using triangulation. A trace study conducted by Winne
and Jamieson-Noel (2002) involved collecting trace measures of study strategies
from undergraduates while they learned about lightning. Trace data were recorded by
instructional software as students studied material. Traces recorded students’ behav-
iors such as scrolling through text and opening windows. Students also completed
a self-report measure of strategies used, and the trace data were matched as closely
as possible to the self-report items such as those assessing planning a method for
studying, creating notes, and reviewing objectives. Through triangulation of data,
the results showed that students tended to self-report overuse of study strategies,
especially for planning a method for studying, highlighting, copying text verbatim
into a note, and reviewing figures. For example, students reported having reviewed
figures 26% more than traces indicated, which suggests that students’ beliefs about
their study strategies were not calibrated with the actual behavior reported in the
traces.
Traces can also provide both face and ecological validity and an unobtrusive
approach to assessing learning processes such as self-monitoring and strategy use.
Azevedo and his colleagues (Azevedo et al., 2018) have done extensive research using
advanced learning technologies (such as intelligent tutoring systems, simulations, and
virtual reality) to assess and foster learning processes in domain-specific contexts.
His findings suggest that data obtained from trace measures may be used to scaffold,
coach, and teach explicit learning and study strategies thus ultimately building self-
efficacy and success. Winne (2020) emphasized the importance of the connection
between theory and traces and the inferences made from them when examining the
validity of trace measures. Bernacki (2018) suggested that think-alouds (discussed
in the subsequent section) can provide concurrent validity to traces because they
provide self-reports of the learning event as it is being traced.
Strengths. Traces are real-time assessments because they track students’ actions
during learning while providing rich sources of data on the processes students engage
in. In addition, the use of advanced learning technologies can enhance achievement by
2 Assessing Academic Self-efficacy 29

“teaching” or guiding students who may be struggling or faltering in their learning,

to engage in different strategies by providing immediate feedback in the form of
explanations and suggestions for alternative ways to improve learning and foster
self-efficacy. Trace measures provide researchers with the opportunity to observe
how students work on a task and study without interference, in real-time rather
than retrospectively, and can be adapted to fit the learners’ ability to problem-solve
comprehensively and contextually (Bernacki, 2018). This methodology also allows
for a larger sample size than case studies as long as access to technology is available
to the participants during the study.
Limitations. Limitations of trace measures include the cost of technology and
program design. The logs also produce large quantities of complex information,
which might be overwhelming and difficult to interpret if the knowledge sought is
not narrowed down in the program development stage and summarized coherently,
and if researchers are not provided with sufficient training to interpret the logs.
Bernacki (2018) highlighted the challenge in demonstrating adequate validity and
reliability of the inferences made from the log transcripts without triangulating the
logs with assessment measures such as surveys or think-aloud protocols. Because
trace measures capture student learning in a nonobtrusive protocol, without supple-
mental information from other sources such as surveys or interviews, researchers may
not have an understanding of students’ self-efficacy beliefs during learning or why
students are behaving as they are during a learning situation (Bernacki, 2018). Self-
efficacy is difficult to assess without directly asking a student how self-efficacious
he or she feels to learn or perform well on a task. Another limitation is the gener-
alizability of the findings to other learning situations and other learners as well as
the replicability of the study using standardized trace measure designs on learners
of various ages and across various domains.

Think-Alouds

The think-aloud methodology requires participants to verbalize what they are “think-
ing” as they engage in a learning task (Moos & Azevedo, 2008). The think-aloud
involves using a set of instructions that are provided to participants prior to begin-
ning the learning activity and specific instructions for how participants should be
prompted during the learning event to ensure consistency across participants (Greene
et al., 2018). One of the most important points for consideration during the think-
aloud is to provide as little amount of interference as possible with the learner as
he or she engages in a task to not interrupt the learning processes. Participants are
instructed to think out loud as they work on problem-solving activities typically, in
a one-on-one setting with the researcher and participant in a laboratory environment
(Ericsson & Simon, 1994). The researcher records the verbalizations, and it assumed
that the thought processes follow a sequential order as the student works through
the learning event. Researchers have used different methods to record participants’
30 M. K. DiBenedetto and D. H. Schunk

verbalizations such as taking notes, eye tracking, video recordings, and audio tapings,
which are then transcribed and analyzed (Greene et al., 2018).
Moos and Azevedo (2008) argue that the think-aloud protocol provides a powerful
way to assess self-efficacy, especially through the use of hypermedia. Hypermedia
is a computer-based environment that contains videos, text, diagrams, and audio.
Hypermedia provides the flexibility of a “rich interactive” learning environment that
can capture self-efficacy as a student engages in a learning event (Moos & Azevedo,
2008). One of the benefits of using hypermedia is that it can provide scaffolds that
can assist students when they are learning a task that is just beyond their capability
thus impacting self-efficacy and performance.
Example. Moos and Azevedo (2008) conducted a study examining the effects of
scaffolds on self-efficacy and self-regulated learning using the think-aloud method-
ology within a hypermedia learning environment. Using hypermedia, thirty-seven
college education majors were randomly assigned to two conditions: conceptual
scaffolding or no scaffolding. Participants were assessed three times using a Likert
scale on their self-efficacy beliefs to learn about the circulatory system to examine
fluctuations in motivational beliefs while learning. Students were also assessed on
their self-regulatory processes as they worked through the hypermedia and at the
end of the learning event. Both groups were asked to think-aloud. Prior to begin-
ning the experiment students were given a pretest to assess their knowledge on the
circulatory system, followed by a five-minute training session on how to use the
hypermedia, and the first self-efficacy assessment. The second self-efficacy scale
was administered during the 30-min learning event, followed by the last one imme-
diately after learning. Results indicated that both groups of students had higher levels
of self-efficacy before beginning the hypermedia which was explained by the lack
of familiarity and experience using hypermedia.
Reliability and Validity. Reliability of the think-aloud can be estimated depending
on the approach for assessing self-efficacy. For example, in Moos and Azevedo’s
(2008) study described above, because participants completed a Likert scale, the
researchers examined internal consistency using Cronbach’s alpha on each adminis-
tration of the questionnaire. In approaches that do not have questionnaires embedded,
researchers may opt to use interrater reliability. Greene et al. (2018) suggest a few
different approaches for participants’ responses recorded manually or by tape record-
ings. The first approach involves selecting a subset of data and checking for interrater
reliability on that subset. Another approach suggests that interrater reliability is esti-
mated on several subsets of data, while a third approach is to estimate reliability on
the entire data set.
Predictive validity may be estimated by examining performance and achievement
measures (Greene & Azevedo, 2009). Triangulation is also recommended to demon-
strate validity for think-aloud measures, particularly with regard to covert processes
such as academic motivation (Greene et al., 2018). Greene et al. (2018) indicated
that some studies have demonstrated discriminant validity, but few have examined
the construct validity of the coding schemes in using the think-aloud.
2 Assessing Academic Self-efficacy 31

Strengths. The think-aloud protocol has many advantages. Learners are assessed
during a learning event, thus providing real-time measures of self-efficacy.
Researchers may gain information on what students are thinking as they work
through an academic task—providing potential opportunities for specific and targeted
interventions. Bandura (1986) indicated that assessments of self-efficacy should be
targeted toward specific learning tasks or within specific learning contexts. Since
think-alouds occur while students are working on a domain-specific task, reported
self-efficacy beliefs should be predictive of learning outcomes once reliability and
validity have been established. Think-alouds can assess motivation and learning
processes with little to no interruptions, thus providing insight during an authentic
learning event (Moos & Azevedo, 2008).
Limitations. A major concern of the think-aloud protocol has been the “reactiv-
ity” effect on student performance because when students are asked or prompted
to explain their cognitions, they may engage in additional cognitive processing that
might not have taken place otherwise and that may affect performance (Ericsson &
Simon, 1994). Ericsson and Simon (1994) recommend resisting the urge to ask
participants to explain their cognitions and to provide explicit instructions and prac-
tice opportunities prior to the assessment so that students have a clear understanding
of the protocol procedures. Another challenge with administering the protocol is that
during the learning events, students may forget to think-aloud or chose not to do
so (Greene et al., 2018). Ericsson and Simon (1994) suggest that researchers use
simple, neutral language such as “keep talking” to remind learners to think-aloud.
These reminders, however, can interrupt a student’s cognitive processing or strategy
use. In addition, because the think-aloud is typically administered in a one-on-one
setting such as a laboratory to eliminate any potential distractions, sample sizes tend
to be small, and the setting is an artificial one which limits the ecological validity of
the methodology.

Microanalyes

The microanalytic methodology is an approach to assessing fine grained measures

of student motivation and self-regulated learning during authentic learning contexts
(Cleary, 2011; DiBenedetto & Zimmerman, 2010). The microanalytic protocol has
been used to assess self-efficacy and the processes of the three phases of self-regulated
learning by asking specific questions targeted at the processes while learning is taking
place. Questions are asked of students prior to learning, while learning, and after
learning during the three phases of self-regulated learning. While typically conducted
on a one-on-one basis, the microanalysis can be conducted in small groups (Cleary
et al., 2008).
The microanalysis differs from the think-aloud in several ways. In the micro-
analysis, specific questions targeting the learning processes such as self-efficacy,
goal setting, strategy use, metacognitive monitoring, and attributions are prepared
in advance of the session. Then, at key points in time during the learning session,
32 M. K. DiBenedetto and D. H. Schunk

students are asked one of the questions aimed at understanding what the student
is doing and why. Students typically respond by pointing to a Likert scale or by
providing a short open-ended response which is then recorded by hand, tape, or
video. To be able to develop these questions, one must be able to understand the
nature of the task and the task demands prior to implementing the microanalysis
(Cleary & Callan, 2018). As indicated earlier is this chapter, Bandura (1997) empha-
sized the necessity of having knowledge and understanding of the task to accurately
assess it and recommends using a microanalytic approach.
Example. DiBenedetto and Zimmerman (2010) examined self-efficacy and the use
of self-regulated learning among students who were at-risk, average achievers, and
advanced level performers in science. Fifty-one high school juniors were individu-
ally assessed using the microanalysis. Participants were presented with a three-page
document on the development and destructive power of tornadoes and were instructed
to read, study, and prepare for a test to be administered at the end of the session.
Students were provided with paper, pencils, highlighters, index cards, and a variety
of other materials they could use to study. Participants were instructed to study as
long as needed and to indicate to the researcher when they felt prepared enough to
take the test. During the protocol, the researcher used the microanalysis to assess
self-regulated learning processes including self-efficacy for learning and performing
well on the tornado knowledge test. Upon test completion, the researcher graded the
student’s exam and asked additional questions to assess students’ feelings and beliefs
and his or her performance. Findings revealed that high achievers had lower levels
of self-efficacy than low achievers and that all students’ test scores were inversely
related to their self-efficacy beliefs. Possible explanations are that low achieving
students may not be able to judge their self-efficacy accurately or they may not fully
understand the task demands whereas high achieving students may underestimate
their capability or worry the task will be too difficult for them to perform at the level
they are accustomed to performing at (DiBenedetto & Zimmerman, 2010).
Reliability and Validity. The microanalytic methodology has demonstrated relia-
bility and predictive and construct validity in various academic, music, athletic, and
medical contexts (Artino et al., 2014; DiBenedetto & Zimmerman, 2013; Kitsantas &
Zimmerman, 2002). Reliability has been established by examining the relationship
among the variables as well as through interrater reliability on open-ended questions
and has been found to be quite high (DiBenedetto & Zimmerman, 2010; Kitsantas &
Zimmerman, 2002).
Convergence of the microanalytic assessment protocol and other measures such
as teacher rating scales and self-report surveys has shown support for validity of
the microanalysis (Cleary & Callan, 2018). DiBenedetto and Zimmerman (2013)
studied the construct and predictive validity of the microanalysis and found the
microanalysis to be a stronger predictor of learning processes than another previously
established measure of learning. More research is needed that focuses specifically
on the microanalytic assessments of self-efficacy.
2 Assessing Academic Self-efficacy 33

Strengths. The microanalysis provides the opportunity to understand the cognitions

and affects of students during authentic learning activities. This real-time measure
includes both quantitative and qualitative measures of self-efficacy and self-regulated
learning and because the measures are phase-linked, they have a potential diagnostic
value for guiding instructional interventions (Cleary & Callan, 2018; DiBenedetto &
Zimmerman, 2013). For example, when assessing students’ self-efficacy, the point at
which students indicate they are not self-efficacious about their capability to complete
a task would be a time to intervene. This could prevent students from struggling or
performing poorly, thus building mastery and increasing self-efficacy.
An additional strength is that the microanalysis is easy to implement and analyze
and may be used with small- and or medium-sized samples. The microanalysis
enables researchers to examine the dynamic nature of motivational and learning
processes in a context-specific setting and provides researchers with a window in
which to view students’ thinking, feelings, and actions while learning (Cleary, 2011).
Limitations. Perhaps the most significant limitation of the microanalytic method-
ology is that it is in its infancy stage. Additional research is needed to examine
the effectiveness of measuring students’ motivation and learning in other academic
contexts and among younger students who may have difficulty articulating what they
are thinking or why they are doing what they are doing when learning. Research is
also needed to determine whether question students during a learning event influences
the outcomes. Additional research that supports the implementation of interventions
would make a significant contribution. For example, if the microanalysis was used
as a diagnostic measure, then during an intervention, and then again as a post assess-
ment, it would provide information on where and when motivation and self-regulated
learning processes shift in a positive direction for the learner (Cleary & Callan, 2018).
In the following sections, we describe the educational implications of obtaining
real-time measures and offer suggestions for future research.

Educational Implications

Each of our educational implications aligns with our recommendations for future
research described below. One of the major reasons why it is so critical for assess-
ment instruments to establish psychometric properties is for use in the classroom.
Practitioners may use assessment instruments to help them reach students who are
struggling with motivation and academically. Through professional development
opportunities, educators can be taught how to use assessment instruments, inter-
pret their findings, and develop intervention strategies that are useable in their daily
activities.
With the knowledge and understanding about the ways to assess self-efficacy and
learning, practitioners could make decisions about data collection. For example, it
would not be practical for a teacher to have her class do a think-aloud all at once;
however, she may be able to arrange for students to be assessed individually using
34 M. K. DiBenedetto and D. H. Schunk

the library or counseling office with the help of a substitute teacher covering her
class. Through professional development, educators may also use multiple sources
of information to help guide instruction and intervention such as a combination of
diaries and surveys administered at key points such as before, during, and after a new
unit.
Finally, an important educational implication lies within the sources of self-
efficacy. Through professional development opportunities, educators can learn to
assess self-efficacy beliefs and develop intervention strategies that capture the
vicarious, persuasive, physiological/emotional, and mastery experiences that foster
student motivation. Most educators offer students words of encouragement (I know
you can do this!), but for example, through the use of trace measures, a practitioner
may decide to use modeling. After administering a trace assessment of students’
studying a chapter in a novel, the teacher realizes students are highlighting almost
every line in the chapter. She interprets this as problematic for two reasons: The
students are unable to distinguish what is important from unimportant; and by under-
lining too much, the students are overwhelmed with how much to remember for the
quiz. The teacher notices that her high achieving students seem to be able to differ-
entiate what is important to highlight from what is not and asks those students to
work with others in small groups and to model the strategy of highlighting, followed
by note taking, and rehearsing. In this simple example, the teacher is using modeling
to help students succeed leading to a mastery experience, thus ultimately increasing
their self-efficacy.

Future Directions

A major concern for educational psychologists and educators has been how best
to assess student motivation and learning in reliable and valid ways so that the
results can be used to predict student behavior and performance. While Bandura
(2000) provided guidance on how to develop surveys to best measure self-efficacy,
we recommend similar guidelines for the other assessment options described above.
Ensuring the reliability and validity of the instruments will mean that when assess-
ments are conducted, regardless of which measurement instrument is used, findings
will be consistent across instruments. This research dedicated to the measurement
properties of the assessment instruments should vary across academic content areas
and grade levels.
Along the same line of ensuring that the psychometric properties of instruments
are met, we recommend more research dedicated to the administration, analysis, and
communication of the assessment methodologies. Through a standardized protocol,
replications of the studies may take place across age level and content area—thus
rather than examining the effectiveness of the instrument’s ability to assess motiva-
tion, researchers will be able to examine the differences across various populations
and domains.
2 Assessing Academic Self-efficacy 35

Additional research is needed demonstrating how the use of multiple assessment

instruments can be used to increase our knowledge and understanding about the
dynamic nature of self-efficacy. There is evidence that self-efficacy changes over
time (DiBenedetto & Bembenutty, 2013). The use of multiple methodologies could
help provide insight into these changes, whether they are subtle or significant, and
provide information on how and where in the classroom interventions may impact
self-efficacy.
Assessing self-efficacy in school settings can have powerful results. Motivation
is not a static process; rather it is dynamic and constantly changing due to personal,
behavioral, and environmental variables. Thus, we recommend that assessment of
motivation is studied in real-time contexts such as in classes while instruction is
proceeding. Studying teacher and student behaviors in these settings will require
methodologies that capture moment-to-moment changes, such as traces, think-aloud
verbal protocols, and microanalytic methods that involve frequent measurements of
key variables. Collecting data in real time will allow researchers to determine the
antecedents and consequences of changes in motivation and provide practitioners
with insights on how to help their students succeed and feel self-efficacious to do so!

References

Artino, A. R., Cleary, T. J., Done, T. Hemmer, P. A., &Durning, S. J. (2014). Exploring clinical
reasoning in novices: A self-regulated learning microanalytic assessment approach. Medical
Education, 48(3), 280–291. https://doi.org/10.1111/medu.12303f
Azevedo, R., Taub, M., & Mudrick, N. V. (2018). Understanding and reasoning about real-time
cognitive, affective, and metacognitive processes to foster self-regulation with advance learning
technologies. In D. H. Schunk & J. A. Greene (Eds.), Handbook of self-regulation of learning
and performance (2nd ed., pp. 254–270). Routledge.
Baleghizadeh, S., & Mortazavi, M. (2014). The impact of different types of journaling techniques
on EFL learners’ self-efficacy. Profile Issues in Teachers’ Professional Development, 16, 77–88.
Bandura, A. (2000). Guide for creating self-efficacy scales. In F. Pajares & t. Urdan (Eds.), Self-
efficacy beliefs of adolescents (pp. 207–338). Information Age Publishing.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Prentice-
Hall Inc.
Bandura, A. (1997). Self-efficacy: The exercise of control. Freeman.
Bernacki, M. L. (2018). Examining the cyclical, loosely sequenced, and contingent features of
self-regulated learning: Trace data and their analysis. In D. H. Schunk & J. A. Greene (Eds.),
Handbook of self-regulation of learning and performance (2nd ed., pp. 370–387). Routledge.
Bernacki, M. L., Nokes-malach, T., & Aleven, V. (2015). Examining self-efficacy during learning:
Variability and relations to behavior, performance, and learning. Metacognition and Learning,
10(1), 99–117. https://doi.org/10.1007/s11409-014-9127-x
Butler, D. L., & Cartier, S. C. (2018). Advancing research and practice about self-regulated learning:
The promise of in-depth case study methodologies. In D. H. Schunk & J. A. Greene (Eds.),
Handbook of self-regulation of learning and performance (2nd ed., pp. 352–369). Routledge.
Carp, F. M., & Carp, A. (1981). The validity, reliability, and generalizability of diary data.
Experimental Aging Research, 7(3), 281–296.
36 M. K. DiBenedetto and D. H. Schunk

Cleary, T. J. (2011). Emergence of self-regulated learning microanalysis: Historical overview, essen-

tial features, and implications for research and practice. In B. J. Zimmerman & D. H. Schunk
(Eds.), Handbook of self-regulation of learning and performance (pp. 329–345). Routledge.
Cleary, T. J., & Callan, G. L. (2018). Assessing self-regulated learning using microanalytic methods.
In D. H. Schunk & J. A. Greene (Eds.), Handbook of self-regulation of learning and performance
(2nd ed., pp. 338–351). Routledge.
Cleary, T. J., Platten, P., & Nelson, A. (2008). Effectiveness of the self-regulated learning empower-
ment program (SLEP) with urban high school youth: An initial investigation. Journal of Advanced
Academics, 20, 70–107.
DiBenedetto, M. K., & Schunk, D. H. (2018). Self-efficacy in education revisited through a socio-
cultural lens: Current and future directions. In G. A. D. Liem & D. M. McInerney (Eds.), Big
theories revisited 2 (pp. 117–140). Information Age Press.
DiBenedetto, M. K., & Bembenutty, H. (2013). Within the science pipeline: Self-regulated learning
and academic achievement among college students in science classes. Learning and Individual
Differences, 2, 218–224.
DiBenedetto, M. K., & Zimmerman, B. J. (2010). Differences in self-regulatory processes
among students studying science: A microanalytic investigation. The International Journal of
Educational and Psychological Assessment, 5(1), 2–24.
DiBenedetto, M. K., & Zimmerman, B. J. (2013). Construct and predictive validity of microanalytic
measures of students’ self-regulation of science learning. Learning and Individual Differences,
26, 30–41.
Ericsson, K. A., & Simon, H. A. (1994). Protocol analysis: Verbal reports as data (revised). The
MIT Press.
Fogarty, R., & McTighe, J. (1993). Educating teachers for higher order thinking: The three-story
intellect. Theory into Practice, 32(3), 161–169.
Fritson, K. K. (2008). Impact of journaling on students’ self-efficacy and locus of control. Insight:
A Journal of Scholarly Teaching, 3, 75–83.
Greene, J. A., & Azevedo, R. (2009). A macro-level analysis of SRL processes and their relations to
the acquisition of a sophisticated mental model of a complex system. Contemporary Educational
Psychology, 34(1), 18–29.
Greene, J. A., Robertson, J., & Costa, L. C. (2018). Assessing self-regulated learning using think-
aloud methods. In D. H. Schunk & J. A. Greene (Eds.), Handbook of self-regulation of learning
and performance (2nd ed., pp. 313–328). Routledge.
Iida, M., Shrout, P. E., Laurenceau, J., & Bolger, N. (2013). Using diaries in psychological research.
In H. Cooper (Ed.), APA handbook of research methods in psychology: Vol. I. foundations,
planning, measures, and psychometrics (pp. 277–305). APA. https://doi.org/10.1037/13619-016
Kitsantas, A., & Zimmerman, B. J. (2002). Comparing self-regulatory processes among novice, non-
expert, and expert volleyball players: A microanalytic study. Journal of Applied Sport Psychology,
14(2), 91–105.
Lent, R. W., Brown, S. D., & Gore, P. A. (1997). Discriminant and predictive validity of
academic self-concept, academic self-efficacy, and mathematics-specific self-efficacy. Journal
of Counseling Psychology, 44(3), 307–315.
Litwin, M. S. (1995). How to measure survey reliability and validity (Ser. The survey kit, 7). SAGE
Publications.
Iida, M., Shrout, P. E., Jean-Philippe, L, & Bolger, N. (2012). Using diary methods in psychological
research. In H. Cooper (Ed.). APA handbook of research methods in psychology,Vol. 1, pp. 277–
305.
Moos, D. C., & Azevedo, R. (2008). Self-regulated learning with hypermedia: The role of prior
domain knowledge. Contemporary Educational Psychology, 33(2), 270–293. https://doi.org/10.
1016/j.cedpsych.2007.03.001
Musyoka, M. M., Gentry, M. A., & Meek, D. R. (2017). Perceptions of teachers’ preparedness
to teach deaf and hard of hearing students with additional disabilities: A qualitative case study.
Journal of Developmental and Physical Disabilities, 29, 827–848.
2 Assessing Academic Self-efficacy 37

Patton, M. Q. (2002). Qualitative research and evaluation methods (3rd ed.). Sage.
Ruokonen, I. (2018). “I…from dreams to reality”: A case study of developing youngsters’ self-
efficacy and social skills through an arts educational project in schools. The International Journal
of Art & Design Education, 480–492. https://doi.org/10.1111/jade.12138
Schmitz, B., & Skinner, E. (1993). Perceived control, effort, and academic performance. Interindi-
vidual, intraindividual, and multivariate time-series analysis. Journal of Personality and Social
Psychology, 64(6), 1010–1028.
Schmitz, B., Klug, J., & Schmidt, M. (2011). Assessing self-regulated learning using diary measures
with university students. In B. J. Zimmerman & D. H. Schunk (Eds.), Handbook of self-regulation
of learning and performance (pp. 251–266). Routledge.
Schunk, D. H., & DiBenedetto, M. K. (2020). Motivation and social cognitive theory. Contemporary
Educational Psychology, 60, Article No. 101832.
Schunk, D. H. (2020). Learning theories: An educational perspective (8th ed.). Pearson Inc.
Schunk, D. H., & DiBenedetto, M. K. (2014). Academic self-efficacy. In M. J. Furlong, R. Gilman, &
E. S. Huebner (Eds.), Handbook of positive psychology in schools (2nd ed., pp. 115–130).
Routledge.
Scott, J. (2011). How instruction supportive of self-regulated learning might foster self-efficacy
for students with and without learning disabilities during literacy tasks. Unpublished Masters
Thesis. Vancouver: University of British Columbia.
Usher, E. L., & Pajares, F. (2008). Sources of self-efficacy in school: Critical review of the literature
and future directions. Review of Educational Research, 78(4), 751–796.
Usher, E. L., & Schunk, D. H. (2018). Social cognitive theoretical perspective of self-regulation. In
D. H. Schunk & J. A. Greene (Eds.), Handbook of self-regulation of learning and performance
(2nd ed., pp. 19–35). Routledge.
Versland, T. M., & Erickson, J. L. (2017). Leading by example: a case study of the influence of
principal self-efficacy on collective efficacy. Cogent Education, 4(1). https://doi.org/10.1080/233
1186X.2017.1286765
Wallen, P., & Adawi, T. (2017). The reflective diary as a method for the formative assessment of
self-regulated learning. European Journal of Engineering Education, 43(4), 507–521.
Winne, P. H. (2020). Construct and consequential validity for learning analytics based on trace data.
Computers in Human Behavior, 112. https://doi.org/10.1016/j.chb.2020.106457
Winne, P. H., & Jamieson-Noel, D. (2002). Exploring students’ calibration of self-reports about
study tactics and achievement. Contemporary Educational Psychology, 27, 551–572.
Winne, P. H., & Stockley, D. B. (1998). Computing technologies as sites for developing self-
regulated laerning. In D. H. Schunk & B. J. Zimmerman, (Eds.), Self-regulated learning: From
teaching to self-reflective practice, (pp. 106- 136). New York: The Guilford Press.
Winnie, P., & Perry, N. (2000). Measuring self-regulated learning. In M. Boekaerts, P. Pintrich, &
M. Zeidner (Eds.), Handbook of self-regulation (pp. 531–566). Academic Press.
Wolters, C. A., Benzon, M. B., & Arroyo-Giner, C. (2011). Assessing strategies for the self-
regulation of motivation. In B. J. Zimmerman & D. H. Schunk (Eds.), Handbook of self-regulation
of learning and performance (pp. 298–312). Routledge.
Wolters, C. A., & Won, S. (2018). Validity and the use of self-report questionnaires to assess
self-regulated learning. In D. H. Schunk & J. A. Greene (Eds.), Handbook of self-regulation of
learning and performance (2nd ed., pp. 307–337). Routledge.
Yin, R. K. (2014). Case study research: Design and methods. SAGE.
Zimmerman, B. J., & DiBenedetto, M. K. (2008). Mastery learning and assessment: Students and
teachers in an era of high stakes testing. Psychology in the Schools, 45(3), 206–216.