Communication Methods and Measures

ISSN: 1931-2458 (Print) 1931-2466 (Online) Journal homepage: http://www.tandfonline.com/loi/hcms20

Development and Validation of the Student

Interest and Engagement Scales

Joseph P. Mazer

To cite this article: Joseph P. Mazer (2012) Development and Validation of the Student
Interest and Engagement Scales, Communication Methods and Measures, 6:2, 99-125, DOI:
10.1080/19312458.2012.679244

To link to this article: http://dx.doi.org/10.1080/19312458.2012.679244

Published online: 08 Jun 2012.

Submit your article to this journal

Article views: 273

View related articles

Citing articles: 12 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=hcms20

Download by: [University of Pennsylvania] Date: 13 April 2016, At: 11:28

 Communication Methods and Measures, 6:99–125, 2012
Copyright © Taylor & Francis Group, LLC
ISSN: 1931-2458 print/1931-2466 online
DOI: 10.1080/19312458.2012.679244

Development and Validation of the Student Interest

and Engagement Scales
Joseph P. Mazer
Clemson University
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

A series of studies report the development of empirically derived instruments that measure student
interest and engagement. The first study inductively develops an initial item pool through open-ended
questionnaire data. A second study subjects the measures to exploratory factor analysis to ascertain an
underlying factor structure. The third study deductively tests the measures through confirmatory fac-
tor analysis and examines associations among teacher communication behaviors, student emotional
and cognitive interest, and engagement. A fourth study offers discriminant validity evidence, suggest-
ing that the new measures are distinct from scales that assess similar yet divergent constructs. Guided
by prior theory (Mottet, Frymier, & Beebe, 2006) and research (Harp & Mayer, 1997), the instruments
developed here possess heuristic potential for instructional communication research. Implications and
areas for future research are discussed.

Data reported by the National Center for Educational Statistics (2010) indicate that only 73%
of high school freshmen graduate within four years. For those who continue their education at
the university level, only 57% attain a bachelor’s degree and just over 18% leave postsecondary
education altogether. Although multiple factors undoubtedly contribute to students’ academic
risk, negative emotions associated with learning—such as a lack of interest and engagement in
their courses—could be a substantial reason for students’ disengagement, withdrawal, and failure
in school (Skinner, Furrer, Marchland, & Kindermann, 2008).
Classroom communication research suggests that students approach rather than avoid imme-
diate teachers because those immediacy behaviors facilitate a sense of liking (Mehrabian, 1981).
That is, as students feel compelled to approach a teacher, they do so out of a state of height-
ened interest. Titsworth (2001b) argued that teachers can use interest cues (e.g., immediacy)
to increase students’ attention and prompt them to be inquisitive and emotionally connected to
learning situations. Similarly, research indicates that clear teaching behaviors can lead students
to experience greater interest in a course because students possess a greater understanding of
the content (Titsworth, 2001b). Weber (2003, 2004) equated interest with learner empowerment,
a motivation-based construct (Frymier, Shulman, & Houser, 1996), whereas Harp and Meyer
(1997) argued that students experience emotional and cognitive forms of interest in learning

This article reports a portion of the author’s dissertation, which was directed by Elizabeth E. Graham and Scott
Titsworth. The author graciously thanks Professors Graham and Titsworth for their assistance.
Correspondence should be addressed to Joseph P. Mazer, Assistant Professor, Department of Communication Studies,
Clemson University, 404 Strode Tower, Clemson, SC 29634. E-mail: jmazer@clemson.edu
 100 J. P. MAZER

situations. Unfortunately absent from the corpus of instructional communication research is a

theoretically driven mechanism to assess students’ experiences of emotional and cognitive forms
of interest in engaged learning environments.
Student engagement can be fostered by effective teacher communication behaviors and stimu-
lated by student interest. Research has noted that engagement encompasses students’ willingness
to participate in their learning (Bomia et al., 1997) and includes the intensity and emotional qual-
ity of their involvement in learning activities (Connell & Wellborn, 1991; Skinner, 1991). Frymier
and Houser’s (1999) seven-item scale assesses indicators and student-perceived estimates of cog-
nitive learning; however, this instrument does not fully account for the range of oral (e.g., asking
questions, participating in discussions) and silent (e.g., taking notes, listening) behaviors indica-
tive of engagement inside (e.g., giving the teacher your full attention) and outside (e.g., preparing
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

for class, talking about the course material with others) the classroom. Although Frymier and
Houser’s Learning Indicators Scale broadly assesses student thought through one scale item, a
new scale can more thoroughly assess specific cognitive behaviors on the part of students such
as thinking about how a course might lead to benefits in the future, how the content relates to
a student’s life, and how material can be put into practice. Indeed, scholars argue that opera-
tionalizing engagement and learning through student-perceived estimates is subject to individual
student bias (McCroskey, Sallinen, Fayer, Richmond, & Barraclough, 1996; Richmond, Gorham,
& McCroskey, 1987; Richmond, McCroskey, Kearney, & Plax, 1987). As an alternative to the
Learning Indicators Scale (Frymier & Houser), which contains an item that measures perceived
learning, a new measure might avoid scale items that tap student-perceived learning estimates and
focus more exclusively upon actual student behaviors that are indicative of silent and oral engage-
ment activities that occur inside and outside of the classroom (Connell & Wellborn; Greenwood,
1991; Skinner, 1991).
The overarching goal of the series of studies reported herein is to extend instructional com-
munication theory and research by developing reliable and empirically derived instruments to
measure student interest and engagement. Adhering to common social scientific research pro-
cedures (Creswell, 2002), these studies first derive a theoretical model of student interest and
engagement through inductive observation and then test this model through deductive quanti-
tative analysis. Specifically, Study One utilizes open-ended survey data to inductively derive
a foundation for scale item generation. Study Two presents the results of exploratory factor
analysis and advances the Student Interest Scale and Student Engagement Scale. Study Three
further refines the instruments by submitting the scales to confirmatory factor analysis (CFA), a
technique that holistically and deductively tests data against a theoretical factor structure spec-
ified a priori by the researcher. Study Three reports the CFA results and examines associations
among teacher communication behaviors, student interest, and engagement. Study Four compares
the measures to similar yet divergent scales and offers discriminant validity evidence. Before
turning to Study One, literature connecting communication and emotional responses provides a
theoretical foundation for the studies reported herein.

THEORETICAL BACKGROUND

Emotion and communication are inherently intertwined as communicators symbolically expe-

rience, construct, and express feelings toward others and their environment (Lupton, 1994).
 INTEREST AND ENGAGEMENT 101

Instructional communication scholars have explored specific types of emotion, including com-
munication apprehension (Bourhis & Allen, 1992; Frymier, 1993; O’Mara, Allen, Long, & Judd,
1996) and more general affective reactions to learning (Andersen, 1979; Rodríguez, Plax, &
Kearney, 1996; Titsworth, 2001b). Arguing for a more focused emphasis on emotion in classroom
communication, Titsworth, Quinlan, and Mazer (2010) found that teacher behaviors influence the
emotional support, emotion work, and overall emotional valence experienced by students. Each
of these research streams implicitly or explicitly acknowledge the role of emotion in explaining
and predicting student learning. Mottet, Frymier, and Beebe (2006) proposed emotional response
theory as a holistic way of synthesizing instructional communication research linking classroom
communication, emotion, and learning. Their theory posits that implicit messages from teachers
(e.g., nonverbal immediacy, affinity-seeking, and behavioral alteration messages) elicit emotional
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

responses from students. In turn, those responses modify students’ orientations to either approach
or avoid learning. Responding to Mottet et al.’s call to explore “specific instructional communi-
cation behaviors or conditions [that] lead to enhanced student emotional responses” (p. 264),
subsequent sections of this review develop reasons for considering the relationship between
teacher communication behaviors, students’ emotional responses, and engagement.

Interest and Emotional Response

Scholarship in the area of interest and learning has been compiling for nearly a century. Various
scholars, some writing in the early 1800s, have conceptualized interest in diverse ways. Herbart
(1806/1965), Dewey (1913, 1916), and Schiefele (1986) argued that interest is embodied in the
person-object relation and that special relations with an object (e.g., a topic or subject matter
area) lead to interest. Other scholars, influenced by more cognitive-developmental theorists (e.g.,
Baldwin, 1906; Piaget, 1981), explored interest in relation to a person’s activity within a larger
socio-cultural environment (Deci, 1992; Renninger, 1992). Other scholars build on writing by
Dewey (1913), Berlyne (1960), and Thorndike (1935) and view specific features of the environ-
ment as able to create interest and explore how and why those particular features might generate
interest (Hidi & Anderson, 1992).
Generally, interested students perceive a content area to be important, are active and involved
in the subject, and feel knowledgeable in the subject (Schiefele, 1991; Tobias, 1994). Interest is
often triggered in the moment by certain conditions (e.g., textual material or teacher behavior)
in the environment and can be characterized from the perspective of the cause—the conditions
that induce interest—or from the standpoint of the person who is interested (Krapp, Hidi, &
Renninger, 1992). Research exploring the conditions eliciting interest has emphasized textual
material of general interest to people (Hidi & Baird, 1988). From this perspective, interest is
not unique to the individual; rather, interest tends to be common across individuals and can be
regulated through individuals’ anticipations and self-evaluations (Bandura, 1977).
Although interest has received significant attention in the educational literature for decades
(Renninger, Hidi, & Krapp, 1992), the connections between interest and communication in the
classroom have received little attention. Titsworth (2001b) argued that teachers can use interest
cues (e.g., immediacy) to increase students’ attention levels and prompt them to be inquisitive
and emotionally connected to learning situations, whereas Weber (2003, 2004) equated interest
with learner empowerment, a multidimensional construct that considers students’ perceptions
of their competence and of a course’s meaningfulness and impact. Absent from these lines of
 102 J. P. MAZER

investigation is a theoretically driven mechanism to assess students’ experiences of emotional and

cognitive forms of interest in learning situations. Mottet et al.’s (2006) emotional response theory
offers a framework to understand the relationship between student interest and engagement.
Emotional interest builds when the addition of interesting but irrelevant material to a les-
son energizes students so that they learn more. Emotional interest cues are entertaining details
in textual material or in a lecture (e.g., engaging words, illustrations, or behaviors) that poten-
tially excite and emotionally engage students. Exploring the role of interest in scientific text
and illustrations, Harp and Mayer (1997) found that emotional interest cues, such as seductive or
entertaining details, enhanced a reader’s enjoyment of the text. This increase in emotional arousal
heightens the reader’s attention, making it easier to encode more material from the text. Titsworth
(2001b) argued that such details often can be related to content material (e.g., examples or stories)
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

and include relational details such as teacher immediacy, leading to heightened student emotional
interest.
Cognitive interest builds when clarity indicators such as explanative summaries influence stu-
dents’ cognition by promoting their structural understanding of content. Harp and Mayer (1997)
argued that those indicators help the reader focus attention on relevant pieces of information and
build internal connections among content. This, in turn, enhances in the reader emotional arousal
related to the subject. In essence, the reader’s achievement of structural understanding leads to a
positive emotional response. That is, course material will become more interesting for students
when they understand it (Harp & Mayer, 1997). Teacher clarity behaviors such as preview state-
ments, transitions, and visual presentation materials can increase cognitive interest because they
make information clearer for students. Ultimately, enhanced emotional and cognitive interest can
lead students to be more engaged in the learning process.

Engagement and Emotional Response

Cognitive learning emphasizes students’ ability to make sense of course concepts and master
course content through the retention of information, by analyzing and synthesizing information,
and through critical evaluation (Bloom, 1956). However, a vexing problem confronting scholars
concerns the operationalization of cognitive learning. Objective exams and tests of recall offer
important consistency and standardization across participants, but by their very nature they are
typically applicable only to a single course or subject area (Chesebro, 2003; Titsworth, 2001a).
Course grades, uniform measures of success in a class, are potentially influenced by teacher
bias, student attendance, and participation (Andersen & Andersen, 1982), whereas self-reported
estimates of perceived learning are subject to individual student bias (McCroskey et al., 1996;
Richmond et al., 1987). Collectively, these cognitive learning measures might function as limited
indicators of learning, as most tend to emphasize the product rather than the process of learning.
Students’ engagement behaviors highlight important attributes that occur as part of the learn-
ing process, as academic engagement time is considered one of the best predictors of learning
(Frymier & Houser, 1999; Woolfolk & McCune-Nicolich, 1984).
Prior research has revealed a positive association between student learning and participation
inside (Myers, 2010) and outside of the classroom (Myers, Martin, & Knapp, 2005). Students
often have the opportunity to listen attentively, orally participate, take notes, and ask questions of
instructors (Bomia et al., 1997). Students might prepare for class by reading assigned material,
reviewing notes, studying for a test, completing homework, and talking about class content with
 INTEREST AND ENGAGEMENT 103

friends (Frymier & Houser, 1999). They might think about how the course material relates to their
lives, how they can utilize their new knowledge and skills, and how the class content will benefit
their future careers (Connell & Wellborn, 1991; Skinner, 1991). Research suggests that students
who spend the most time engaged in attending, working, or interacting with others (on matters
related to the content) experience the highest levels of academic achievement (Greenwood, 1991;
McGarity & Butts, 1984; Rosenshine, 1979).
Early engagement research often utilized time-based measures (e.g., time-on-task) to assess
student engagement rates (Brophy, 1983). More recently, however, scholars have noted that
engagement encompasses the intensity and emotional quality of students’ involvement in ini-
tiating and executing learning activities (Connell & Wellborn, 1991; Skinner, 1991). Emotional
response theory proposes that student emotional responses (e.g., emotional and cognitive inter-
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

est) will increase students’ time on task and ultimately heighten their engagement in the learning
process. Therefore, guided by prior theory and research (Mottet et al., 2006), students who
experience heightened levels of emotional and cognitive interest might be more engaged in the
classroom and approach learning in more meaningful ways. Consequently, measures of student
interest and student engagement can further explain the relationship among classroom commu-
nication, emotion, and learning and advance the propositions set forth in emotional response
theory.

STUDY ONE: ITEM GENERATION

Participants and Procedures

Participants were 374 undergraduate students enrolled in an introduction to human commu-

nication course at a large midwestern university (141 first-year students, 134 sophomores,
60 juniors, 39 seniors). The sample consisted of 157 males and 217 females, with an average
age of 19.89 years (ranging from 18 to 34 years, SD = 1.77). The racial/ethnic distribution was
91.4% Caucasians, 4.5% African Americans, 1.3% Hispanics, 1.1% Asian Pacific Islanders, .4%
American Indian/Alaskan Native, and 1.3% Other.
All procedures were approved through the university’s Institutional Review Board (IRB).
Scale development procedures included an online open-ended questionnaire designed to solicit
participants’ reactions to issues associated with interest and engagement. Participants were asked
to reflect on their experience as a student in college courses and answer the following open-
ended items: (1) You are likely to be very interested in some classes and not interested in other
classes. What contributes to your interest in a particular class? (2) What keeps you alert, atten-
tive, engaged, and involved during class? This could be behaviors of your teacher, classmates,
or things about the subject area. (3) Based upon your answers to prior questions, what things do
you do inside and outside of class that are influenced by your interest?

Data Analysis

Open-ended survey data were analyzed to provide an inductively derived base for item construc-
tion. Following similar analytic techniques popular in communication research (Ledbetter, 2009;
Thompson & Mazer, 2009), Owen’s (1984) criteria to obtain cohesive patterns occurring in a
 104 J. P. MAZER

set of discourse was utilized. Patterns were identified through recurrence—“when at least two
parts of a report had the same thread of meaning, even though different wording indicated such
a meaning”—and through repetition “of key words, phrases or sentences” (p. 275). Owen’s third
criterion—forcefulness of nonverbal cue usage—was irrelevant given the textual nature of the
data. This analysis revealed overarching themes that characterize emotional interest, cognitive
interest, and engagement activities relevant to student behaviors that occur inside and outside the
classroom (e.g., participating in class, actively listening to the teacher, asking questions of the
teacher, talking about the content with others). A subsequent examination of extant literature also
confirmed the importance of these patterns. The open-ended survey data and information gleaned
from a review of relevant literature provided a base for item construction. Communication schol-
ars, in particular, have similarly utilized qualitative data to explore a phenomenon and then
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

develop a measure for later use (e.g., Kroman Myers & Oetzel, 2003; Thompson & Mazer;
Treadwell & Harrison, 1994; Vangelisti, Crumley, & Baker, 1999). The information gleaned
from the qualitative data provided a base of material to “identify or narrow the focus of
the possible variables” to be measured in the scales developed herein (Creswell et al., 2003,
p. 228).

STUDY ONE RESULTS

The open-ended survey data served as a foundation for scale item construction. First, the author
read through the data several times to develop a holistic sense of the data. To employ a derived etic
approach, no predetermined categories were used. This approach revealed “what units make sense
within the world of the messages” instead of using prescribed categories that may not encompass
the range of actual student experiences (Neuendorf, 2002, p. 72). Analysis revealed that most of
the respondents broadly referred to teacher and student behaviors and content area characteristics
that influenced their interest (e.g., enthusiasm about being in class, understanding the flow of
ideas). Sixteen items were drafted to tap outcomes of cognitive interest and emotional interest.
Each item began with the stem: “I am interested in this class because . . .” Likert response options
were drafted ranging from strongly disagree to strongly agree. See Table 1 for the Student Interest
Scale item pool.
Face validity has been regarded as a component of content validity (Baxter & Babbie, 2004;
DeVellis, 2003). To enhance the face validity of the new interest measure and ensure differ-
entiation between emotional interest and cognitive interest, three coders, blind to the purpose
of the project, assisted with scale development. The coders were given a randomized list of
drafted interest items and a sheet containing a description of cognitive interest and emotional
interest. Cognitive interest was described as a cognitive response in students who are interested
in the material/topics because they are able to understand, recall, and remember course material.
Emotional interest was described as an affective response in students who are enthused, engaged,
and excited by course content and the class experience. The coders were instructed to place each
item in a single category (cognitive interest or emotional interest) and indicate the intensity with
which each particular item tapped the interest category. The intensity rating allowed the coders
to indicate the “strength” to which each item reflected the particular type of student interest (i.e.,
emotional interest or cognitive interest). Coders utilized a 1–10 scale for the intensity rating, with
1 indicating very weak and 10 indicating very strong. Coding results were inspected for evidence
of face validity. Visual inspection of the results indicated that all coders clearly placed the items
 INTEREST AND ENGAGEMENT 105

TABLE 1
Student Interest Scale Initial Item Pool

I am interested in this class because . . .

1. . . . I understand the course material (CI)

2. . . . The class makes me feel excited (EI)
3. . . . I can remember the course material (CI)
4. . . . Being in the class is enjoyable (EI)
5. . . . I can understand the flow of ideas (CI)
6. . . . The topics covered in the course fascinate me (EI)
7. . . . The information in the course is useful (CI)
8. . . . I feel enthused about being in class (EI)
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

9. . . . I realize what is expected of me (CI)

10. . . . The material fascinates me (EI)
11. . . . I feel like I am learning topics covered in the course (CI)
12. . . . The class causes me to feel energized (EI)
13. . . . The information covered in the course is making me more knowledgeable (CI)
14. . . . The class experience makes me feel good (EI)
15. . . . The class experience feels very positive (EI)
16. . . . I like the things we cover in class (EI)

Note. EI = Emotional Interest; CI = Cognitive Interest.

in the appropriate categories and provided strong intensity ratings (ranging from 7 to 10) for each
item, indicating that each item was worded to fittingly tap the particular type of student inter-
est. Given that all coders achieved superior accuracy as they placed items in their appropriate
categories and provided exceptional strength scores, intercoder reliability statistics were not cal-
culated. A visual inspection of these items and coders’ analysis suggests that requirements for
face validity—a standard necessary for content validity (Baxter & Babbie; DeVellis)—appear to
be met.
The construction of the interest scale item pool was guided by prior theory and research that
specified two types of interest (i.e., emotional interest and cognitive interest). Unfortunately, the
engagement literature does not clearly feature theories that might suggest an underlying factor
structure of an engagement scale. In light of this, 20 engagement items were drafted based upon
participants’ responses in the open-ended survey data. Those items will then be submitted to
exploratory factor analysis in Study Two. Submitting these items to exploratory factor analysis
will allow the statistical procedure to dictate the clusters of items that might emerge in specific
categories to represent subsets of the larger engagement construct. The drafted engagement items
were related to engagement activities that might occur inside and/or outside of the classroom and
permitted students to specify the frequency with which they engage in behaviors associated with
the learning process (e.g., participating in class, actively listening to the teacher, asking ques-
tions of the teacher, talking about the content with others). A seven-point semantic differential
response option (never/very often) was crafted to provide a greater degree of variability and give
participants more freedom to indicate the frequency of their behaviors as measured by the indi-
vidual scale items. See Table 2 for the Student Engagement Scale item pool. Study One findings
provided an initial item pool for the interest and engagement scales. A second study offered an
initial test of the structure of the scales through exploratory factor analysis to further establish
content validity (Baxter & Babbie, 2004).
 106 J. P. MAZER

TABLE 2
Student Engagement Scale Initial Item Pool

1. Orally (verbally) participated during class discussions

2. Participated during class discussions by sharing your thoughts/opinions
3. Took notes during class
4. Listened attentively to the instructor during class
5. Listened attentively to your classmates’ contributions during class discussions
6. Attended class
7. Asked questions of your teacher during class if something was unclear
8. Gave your teacher your full attention during class
9. Studied for a test or quiz
10. Prepared for class by reading the assigned readings
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

11. Reviewed your notes outside of class

12. Talked about the course material with others outside of class
13. Took it upon yourself to read additional material in the course topic area
14. Completed the assigned homework
15. Completed the assigned readings
16. Contacted your teacher (in person or via email) about your progress in the course
17. Applied/Used the course material in my interactions with others outside of class
18. Thought about how the course material related to your life
19. Thought about how you can utilize the course material in your everyday life
20. Thought about how the course material will benefit you in your future career

STUDY TWO: EXPLORATORY FACTOR ANALYSIS

Participants and Procedures

Participants were 313 undergraduate students enrolled in an introduction to human commu-

nication course at a large midwestern university (145 first-year students, 121 sophomores,
33 juniors, 14 seniors). The sample consisted of 130 males and 183 females, with an average
age of 19.13 years (ranging from 18 to 30 years, SD = 1.28). The racial/ethnic distribution was
87.8% Caucasians, 7.7% African Americans, 1.3% Hispanics, 1.3% Asian Pacific Islanders, .3%
American Indian/Alaskan Native, and 1.6% Other.
All procedures were approved through the university’s IRB. Following a similar approach
common in instructional communication research (see Plax, Kearney, McCroskey, & Richmond,
1986), students were asked to identify the first class they attend in a particular week and to use that
class as the reference point for all questions on the survey. The size of students’ target classes was
moderate, with an average of 43 students enrolled. Students reported on slightly more male target
class teachers (n = 159, 50.8%) than females (n = 152, 48.6%). Two participants did not report
target teacher sex. Students described more classes as “discussion oriented” (n = 161, 51.4%);
47.9% of the courses (n = 150) were described as “mostly lecture oriented,” and two students
(.6%) did not answer the question. Almost half of the students (n = 140, 44.7%) reported they
were enrolled in their target class because of their major, whereas a slightly smaller number (n =
117, 37.4%) indicated that they were enrolled because of a general education requirement. A total
of 22 students (7%) indicated they were taking the class as part of a second major or minor, and
 INTEREST AND ENGAGEMENT 107

34 students (10.9%) did not provide a reason for taking the target class. Participants were asked to
reflect on their target class and complete the interest and engagement scales developed in Study
One. After the participants completed the scales, the measures were subjected to exploratory
factor analysis and reliability procedures.

STUDY TWO RESULTS

Baxter and Babbie (2004) argue that content validity evidence can be compiled through fac-
tor analysis procedures to analyze the multidimensional structure of measures. An iterative data
reduction process was employed during the exploratory factor analysis of the interest and engage-
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

ment scales. Specifically, principal axis factoring using promax rotation was used to identify
factors within the scales and eliminate survey items that did not adequately load onto a factor.
Factors with eigenvalues greater than 1.00 and depicted above the bend or elbow in a visual
inspection of the scree plot were retained. Furthermore, the rotated factor matrix was scrutinized
to determine which survey items met a factor loading criteria of .60 or greater on the primary
factor and a .40 or lower loading on the secondary factor (McCroskey & Young, 1979).

Exploratory Factor Analysis of Student Interest Scale

Participants reported their levels of emotional and cognitive interest using a five-point Likert scale
with options ranging from strongly disagree to strongly agree. The Bartlett Test of Sphericity
suggested that the data met assumptions necessary for factor analysis, χ 2 (120) = 4711.99, p <
.05. The factor analysis produced a two-factor solution that met this criterion. Examination of the
items loading on these factors revealed one factor that addresses emotional interest and one factor
that addresses cognitive interest (see Table 3).
All interest scale items began with the stem—“I am interested in this class because . . .”
Based on the rotated factor matrix, the first factor contained nine items related to emotional
interest and included items such as: “. . . The class causes me to feel energized” and “. . .
The topics covered in the course fascinate me.” The first factor was subsequently labeled emo-
tional interest. Although item nine’s secondary loading slightly exceeded .40, it was retained
in light of its strong primary loading. The second factor contained seven items related to cog-
nitive interest and included items such as: “. . . I can understand the flow of ideas” and “. . . I
can remember the course material.” The second factor was labeled cognitive interest. Although
the primary loading of item 16 did not meet the .60 criterion, it was retained because its sec-
ondary loading was small. The two factors collectively accounted for 73.60% of variance in the
scale.
After summing scores for the two factors, items comprising each factor were analyzed for
reliability. Using Cronbach’s alpha, reliability estimates were calculated for the two factors: emo-
tional interest α = .97 and cognitive interest α = .91. All items on the scale were analyzed to
determine an overall reliability estimate for the scale (α = .96).
Finally, a Pearson correlation was calculated for the pairwise combination of factors (r =
.84, p < .01). Because the primary factor loadings were robust and the reliability estimates
for each factor were high, the 16-item interest scale is best described by a two-factor solution
that assesses two dimensions of interest: emotional interest and cognitive interest. Furthermore,
 108 J. P. MAZER

TABLE 3
Factor Loadings for Student Interest Scale

I am interested in this class because . . . Emotional Interest Cognitive Interest

1. . . . I feel enthused about being in class. .83 .37

2. . . . The class makes me feel excited. .82 .38
3. . . . The class causes me to feel energized. .79 .28
4. . . . The topics covered in the course fascinate me. .78 .36
5. . . . Being in the class is enjoyable. .77 .34
6. . . . The class experience makes me feel good. .77 .38
7. . . . The material fascinates me. .74 .35
8. . . . I like the things we cover in class. .72 .39
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

9. . . . The class experience feels very positive. .64 .41

10. . . . I can remember the course material. .32 .76
11. . . . I feel like I am learning topics covered in the course. .38 .74
12. . . . I can understand the flow of ideas. .37 .73
13. . . . I understand the course material. .33 .69
14. . . . The information covered in the course is making me .34 .66
more knowledgeable.
15. . . . The information in the course is useful. .39 .64
16. . . . I realize what is expected of me. .19 .56
Eigenvalue 10.49 1.28
% of Variance 65.59 8.01
Cronbach’s Alpha .97 .91

Note. Underlined factor coefficients indicate the primary factor loading.

Harp and Mayer’s (1997) work related to emotional interest and cognitive interest highlights the
presence of two forms of student interest and suggests that these characteristics can separately
influence student engagement and learning.

Exploratory Factor Analysis of Student Engagement Scale

Participants reported how frequently they took part in each of the engagement activities using
a seven-point semantic differential scale with a bipolar response option (never/very often). The
Bartlett Test of Sphericity suggested that the data met assumptions necessary for factor analysis,
χ 2 (91) = 3044.46, p < .05. Factor analysis produced a four-factor solution that met this criterion.
Examination of the items loading on these factors revealed factors that address (1) silent in class
behaviors, (2) oral in class behaviors, (3) thinking about course content, and (4) out of class
behaviors (see Table 4).
Based on the rotated factor matrix, the first factor contained four items included items such
as: “Listened attentively to the instructor during class” and “Gave your teacher your full attention
during class.” The first factor was subsequently labeled silent in class behaviors. The second fac-
tor contained two items related to verbal in class behaviors: “Participated during class discussions
by sharing your thoughts/opinions” and “Orally (verbally) participated during class discussions.”
The second factor was labeled oral in class behaviors. The third factor contained three items that
addressed how students thought about the course content (e.g., “Thought about how you can
 INTEREST AND ENGAGEMENT 109

TABLE 4
Factor Loadings for Student Engagement Scale

Silent in class Oral in class Thinking about Out of class

behaviors behaviors course content behaviors

1. Listened attentively to the instructor during class. .84 .08 .20 .20
2. Gave your teacher your full attention during class. .74 .15 .21 .25
3. Listened attentively to your classmates’ contributions .67 .30 .16 .22
during class discussions.
4. Attended class. .62 .18 .10 .05
5. Participated during class discussions by sharing your .16 .95 .14 .07
thoughts/opinions.
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

6. Orally (verbally) participated during class discussions. .20 .92 .11 .06
7. Thought about how you can utilize the course material .18 .15 .87 .28
in your everyday life.
8. Thought about how the course material related to your .22 .19 .79 .29
life.
9. Thought about how the course material will benefit .26 .12 .72 .27
you in your future career.
10. Reviewed your notes outside of class. .23 .00 .20 .88
11. Studied for a test or quiz. .31 .02 .19 .61
12. Talked about the course material with others outside .20 .26 .30 .59
of class.
13. Took it upon yourself to read additional material in the .02 .26 .28 .58
course topic area.
Eigenvalue 6.35 1.88 1.49 1.05
% of Variance 45.38 13.42 10.62 7.47
Cronbach’s Alpha .86 .96 .92 .82

Note. Underlined factor coefficients indicate the primary factor loading.

utilize the course material in your everyday life” and “Thought about how the course material
related to your life”). The third factor was appropriately labeled thinking about course content.
The fourth factor contained four items, including: “Studied for a test or quiz” and “Talked about
the course material with others outside of class.” The fourth factor was subsequently labeled out
of class behaviors. Although two items on this factor did not exceed the .60 primary loading,
they were retained because their secondary loadings were relatively small (.30 or less). The four
factors collectively accounted for 76.89% of variance in the scale.
After summing scores for the four factors, items comprising each factor were analyzed for reli-
ability. Using Cronbach’s alpha, reliability estimates were calculated for the four factors: silent
in class behaviors α = .86; oral in class behaviors α = .96; thinking about course content α =
.92; and out of class behaviors α = .82. All 13 items on the scale were analyzed to determine an
overall reliability estimate for the scale (α = .90).
Emotional response theory specifies that students’ responses to all components of the learning
experience rather than particular teacher behaviors or perceptions leads to increases in learn-
ing. In essence, the theory’s student-centered rather than teacher-centered approach guided the
efforts of the present study (Mottet et al., 2006). Frymier and Houser (1999) utilized faculty
descriptions of students’ engagement behaviors to form the basis of the Learning Indicators Scale.
 110 J. P. MAZER

The Student Engagement Scale was developed using students’ reports of common engagement
behaviors. Although Frymier and Houser’s Learning Indicators Scale broadly assesses student
cognition through one scale item, the Student Engagement Scale developed here addresses sev-
eral specific cognitive behaviors that students reported (in Study One) were indicative of their
engagement in the learning process. In light of claims that students’ self-reported estimates of
their perceived learning are subject to bias (McCroskey et al., 1996; Richmond et al., 1987), the
Student Engagement Scale avoids scale items that assess students’ perceptions of actual learning,
unlike items contained in the Learning Indicators Scale (Frymier & Houser). Guided by emotional
response theory, the Student Engagement Scale assesses specific student behaviors that comprise
their engagement in the learning process. As further specified by emotional response theory,
teacher communication behaviors can lead students to experience positive emotional responses
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

that lead them to approach rather than avoid engaged learning situations.
The Study Two findings establish the dimensionality of the interest and engagement scales
through exploratory factor analysis. The factors that emerged here offer a robust explanation
of interest and engagement and offer content validity evidence for the Student Interest Scale and
Student Engagement Scale. A third study further tested the scales’ structure through confirmatory
factor analysis and examined associations between teacher communication behaviors, interest,
and engagement to establish additional validity evidence.

STUDY THREE: CONFIRMATORY FACTOR ANALYSIS AND RELATIONSHIPS

AMONG TEACHER COMMUNICATION, STUDENT INTEREST, AND ENGAGEMENT

Teacher Immediacy

Teacher immediacy is enacted through verbal and nonverbal behaviors that generate perceptions
of psychological closeness between the teacher and students (Andersen, 1979). Teacher nonver-
bal immediacy behaviors have been operationalized to include the use of eye contact, movement,
facial expressions, and vocal variety, among others (Andersen, 1979). Studies exploring the rela-
tionship between teacher nonverbal immediacy and student learning (Richmond et al., 1987) have
shown that students are drawn to highly immediate teachers because those behaviors facilitate a
sense of liking and compel a person to approach, rather than avoid, the source of the immedi-
ate behavior (Mehrabian, 1981). This line of reasoning explicitly assumes that, as students feel
compelled to enact approach behaviors, they do so out of a state of heightened positive emo-
tion. Consistent with the predictions of emotional response theory, research suggests that teacher
immediacy can stimulate emotional arousal in students which can ultimately lead to greater
emotional interest and learning (Christophel, 1990; Frymier, 1993, 1994).
In fact, the relationship between teacher immediacy and broad indicators of emotion has been
documented with noteworthy consistency. In studies of other service professions, immediacy was
identified as a strategy for responding to emotional needs of others (Miller, 2007). Jones and
Wirtz (2006) also observed the positive effects of nonverbal immediacy on people’s perceived
emotional improvement in controlled laboratory situations. Results of experimental classroom
studies suggest that, when teachers use nonverbal immediacy, students report higher levels of
perceived effect (Chesebro, 2003; Comstock, Rowell, & Bowers, 1995; Titsworth, 2001b; Witt &
Wheeless, 2001); numerous correlational studies have observed a similar relationship (Andersen,
 INTEREST AND ENGAGEMENT 111

1979; Christophel, 1990; Plax et al., 1986; Rodríguez et al., 1996). In fact, a meta-analysis of
55 studies, examining the relationship between teacher nonverbal immediacy and students’ per-
ceived affective learning, found an average correlation of .49 (Witt, Wheeless, & Allen, 2004).
Therefore, it stands to reason that immediate teachers can energize students and influence their
emotional connection to the course and its content. This heightened emotional interest can, in
turn, lead students to be more engaged in the learning process. Therefore, a positive relationship
likely exists between students’ perceptions of teacher immediacy and their emotional interest.
Students’ feelings of emotional interest will be positively related to their engagement.

Teacher Clarity
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

Teachers have at their disposal a range of clarity behaviors, including examples, descriptions,
transitions, and explanations (Bush, Kennedy, & Cruickshank, 1977). As noted by Titsworth,
Novak, Hunt, and Meyer (2004), those behaviors can occur verbally, as teachers talk about course
material, and nonverbally, through teachers’ use of PowerPoint displays, handouts, and notes on
the board. Although many studies relied on self- and other-report measures assessing teachers’
use of clarity behaviors, scholars have also noted that clarity is a communicative process that
emerges through the give-and-take of classroom communication (Simonds, 1997).
Much of the teacher clarity literature has explicitly explored relationships between clarity and
cognitive learning outcomes. However, Titsworth and Mazer’s (2010) review of clarity research
noted that several studies have found significant and positive correlations between teachers’ use
of clarity behaviors and students’ affect/motivation toward a class. Likewise, Glaser-Zikuda and
Fuss (2008) observed a strong positive correlation between students’ reports of their perceived
well-being and their teachers’ clarity behaviors; teacher clarity was negatively related to stu-
dents’ perceptions of anxiety. Schrodt and his colleagues (2009) reported that teacher clarity was
positively associated with learner empowerment and cognitive learning indicators.
Consistent with the predictions of emotional response theory, research has consistently found
that verbal and written clarity leads students to remember content more (i.e., approach learning)
than students not exposed to clear teaching behaviors. Titsworth (2001b) argued that teachers
can utilize explanatory summaries to highlight important relationships among lecture content,
use clear transitions to help students follow the lesson content, and implement visual materials to
make abstract and un-engaging material concrete and stimulating—building students’ cognitive
interest. In essence, clear teaching behaviors can lead students to be more cognitively interested in
the course material because they possess a greater understanding of the content and its structural
connections. Students’ cognitive interest will likely be positively related to their engagement in
the learning process.

Participants and Procedures

Participants were 252 undergraduate students enrolled in classes at a large midwestern university
(67 first-year students, 67 sophomores, 53 juniors, 65 seniors). The sample consisted of 83 males
and 168 females, with an average age of 20.36 years (ranging from 18 to 33 years, SD = 1.90).
One participant failed to report sex. The racial/ethnic distribution was 94.4% Caucasians, 2%
Hispanics, 1.2% African Americans, .8% Asian Pacific Islanders, .8% American Indian/Alaskan
Native, and .8% Other.
 112 J. P. MAZER

All procedures were approved through the university’s IRB. Using the same method described
in Study Two (Plax et al., 1986), students were asked to identify the first class they attend in a
particular week and then use that class as the reference point for all questions on the survey. Of the
teachers, 133 were male and 119 were female. In terms of class format, participants reported
that 137 classes were mostly lecture oriented, and 113 were mostly discussion oriented. Two
participants did not report class format. The average class size was 31.58 students. Most of the
students (n = 143, 56.7%) reported they were enrolled in their target class because of their major,
whereas a smaller number (n = 62, 24.6%) indicated that they were taking the class because of a
general education requirement. A total of 23 students (9.1%) indicated they were taking the class
as part of a second major or minor, and 24 students (9.5%) did not provide a reason for taking the
target class.
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

Measurement

Teacher immediacy

Teacher immediacy was operationalized using Richmond, McCroskey, and Johnson’s (2003)
26-item measure of nonverbal immediacy, the most recent version of the scale that has received
widespread use. The teacher immediacy scale assesses the extent to which students perceive a
teacher to use nonverbally immediate behaviors in the classroom. Participants responded using a
five-point Likert-type scale with items never, rarely, occasionally, often, and very often. Versions
of the scale have enjoyed decades of widespread use with high alpha reliability estimates often
ranging from .83 (Frymier, 1994) to .94 (Titsworth, 2004). In this study, the reliability was α =
.81 (M = 48.98, SD = 8.05).

Teacher clarity

Despite substantial scholarly interest in teacher clarity, no single method for assessing clarity
has emerged. Whereas a variety of options exist (see Titsworth & Mazer, 2010), the decision
to use any particular scale generally is determined based on the level of specificity required for
a particular study. This study utilized the 12-item Clarity Behaviors Inventory (CBI) developed
by Titsworth et al. (2004) because that scale assesses a range of clarity behaviors across distinct
channels. The scale has enjoyed use in prior research (Titsworth et al., 2010). Using a five-point
Likert scale, the CBI operationalizes students’ perceptions of teachers’ oral (e.g., “The teacher
verbally stresses important issues presented in the lecture”) and written (e.g., The teacher provides
us with a written description of the most important things in the lecture”) clarity behaviors. The
reliability estimates were strong, with values of .88 (M = 23.64, SD = 4.65) and .86 (M = 21.80,
SD = 5.47) for verbal and written clarity, respectively.

Student interest

Interest was operationalized using the Student Interest Scale developed in Studies One and
Two. Participants responded using a five point Likert scale with response options ranging from
strongly disagree to strongly agree. In this study, the measure yielded high reliability coefficients:
 INTEREST AND ENGAGEMENT 113

α = .96 (M = 29.99, SD = 9.81) for emotional interest; α = .89 (M = 27.49, SD = 5.18) for
cognitive interest.

Student engagement

Engagement was measured using the 13-item Student Engagement Scale developed in Studies
One and Two. Participants reported how frequently they took part in each of the engage-
ment activities using a seven-point semantic differential scale with a bipolar response option
(never/very often). The results of this study revealed that the engagement scale was reliable:
silent in class behaviors α = .77 (M = 23.57, SD = 4.19); oral in class behaviors α = .93 (M =
8.85, SD = 3.85); thinking about course content α = .91 (M = 13.87, SD = 5.60); and out of
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

class behaviors α = .70 (M = 16.63, SD = 5.55).

Confirmatory factor analysis

The overall structures of the interest and engagement scales were tested via confirmatory
factor analysis. All CFA procedures were conducted using LISREL 8.80, and four popular
indices assessed model fit: (a) model chi-square, (b) the root mean square error of approxima-
tion (RMSEA), (c) the nonnormed fit index (NNFI), and (d) the comparative fit index (CFI).
Model fit is generally considered acceptable if CFI and NNFI values are above .90 (and prefer-
ably above .95) and the RMSEA statistic does not exceed .08 (and preferably .05) (Kline, 2005;
MacCallum, Browne, & Sugawara, 1996).

STUDY THREE RESULTS

One goal of Study Three was to test the structure of the interest and engagement scales. The
interest items were submitted to confirmatory factor analysis with each manifest indicator of
emotional interest and cognitive interest loading onto its respective latent construct. Analysis
revealed that all items loaded onto their respective latent construct. Considering the standards for
model fit reported previously, the confirmatory model for the interest scale indicated good model
fit, χ 2 (8) = 20.94, p < .05, RMSEA = .056[90% CI: .027-.086] , NNFI = .99, CFI = .99. A one-factor
solution combining both latent constructs (i.e., emotional interest and cognitive interest) produced
a significant decline in model fit, χ 2 (1) = 287.73, p < .01, suggesting that the two-factor
solution was most appropriate. The engagement items were submitted to confirmatory factor
analysis with each manifest indicator of engagement loading onto its respective latent construct
(i.e., silent in class behaviors, oral in class behaviors, thinking about course content, and out
of class behaviors). The confirmatory model for the engagement scale indicated good model fit,
χ 2 (59) = 207.27, p < .05, RMSEA = .067[90% CI: .057-.078] , NNFI = .96, CFI = .97. A three-factor
solution was tested by constraining to 1.0 the covariance estimate between the two factors that
produced the highest intercorrelation (i.e., silent in class behaviors and out of class behaviors).
This analysis produced a significant decline in model fit, χ 2 (1) = 158.43, p < .01, suggesting
that the four-factor solution was most appropriate. Thus, the series of confirmatory factor analyses
on the interest and engagement scales confirmed that the factor structures of the instruments
 114 J. P. MAZER

TABLE 5
Descriptive Statistics and Pearson Product-Moment Correlations for all Variables

M SD 1 2 3 4 5 6 7 8 9

1. Teacher Immediacy 48.98 8.05 –

2. Teacher Verbal Clarity 23.64 4.65 .61∗∗ –
3. Teacher Written Clarity 21.80 5.47 .37∗∗ .59∗∗ –
4. Student Emotional Interest 29.99 9.81 .53∗∗ .55∗∗ .29∗∗ –
5. Student Cognitive Interest 27.49 5.18 .43∗∗ .58∗∗ .36∗∗ .76∗∗ –
6. Silent In Class Behaviors 23.57 4.19 .35∗∗ .26∗∗ .18∗∗ .32∗∗ .26∗∗ –
7. Oral In Class Behaviors 8.85 3.85 .27∗∗ .13∗ .00 .32∗∗ .22∗∗ .27∗∗ –
8. Thinking About Course Content 13.87 5.60 .25∗∗ .23∗∗ .09 .53∗∗ .51∗∗ .33∗∗ .33∗∗ –
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

9. Out of Class Behaviors 16.63 5.55 .18∗∗ .20∗∗ .31∗∗ .26∗∗ .15∗ .36∗∗ .22∗∗ .33∗∗ –

Note. ∗ Correlations are significant at p < .05. ∗∗ Correlations are significant at p < .01.

were viable and compiled important evidence for construct validity by showing that the data
theoretically fit expectations for the measures (deVaus, 2001; DeVellis, 2003).
The second goal of Study Three was to examine the associations between teacher immediacy
and student emotional interest, emotional interest and engagement, teacher clarity and student
cognitive interest, and cognitive interest and engagement. Descriptive statistics, including means,
standard deviations, and Pearson product-moment correlations for all variables, are reported in
Table 5. A series of hierarchical regression models, using course type (nonmajor vs. major) and
course format (lecture vs. discussion) as control variables were estimated.
The first model examining the relationship between teacher immediacy and clarity and stu-
dents’ emotional interest was significant, R = .63, R2 = .40, F(5, 246) = 32.10, p < .001,
accounting for 40% of the shared variance in emotional interest. At step 1, R2 = .07,
F = 9.56, p < .001, course format was a significant predictor of emotional interest, β = .24,
t = 3.92, p < .01. After controlling for course format and type, at step 2, R2 = .38, F = 43.84,
p < .001, teacher immediacy, β = .28, t = 4.38, p < .01, and verbal clarity, β = .35, t = 4.86,
p < .01, were significant predictors of students’ emotional interest.
The second model exploring the relationship between teacher immediacy and clarity and stu-
dents’ cognitive interest was significant, R = .61, R2 = .37, F(5, 246) = 29.23, p < .001, and
accounted for 37% of the variance in cognitive interest. At step 1, R2 = .05, F = 6.67, p <
.01, course format, β = .16, t = 2.55, p < .01, and course type, β = .15, t = 2.41, p < .01,
were significant predictors of cognitive interest. After controlling for course format and type, at
step 2, R2 = .32, F = 42.08, p < .001, verbal clarity was a significant predictor of students’
cognitive interest, β = .47, t = 6.36, p < .01.
To explore the four forms of student engagement, a series of hierarchical regression models
were estimated. Course type and format were entered as control variables, teacher immediacy
and clarity were entered as predictor variables at step 2, and student emotional and cognitive
interest were entered at step 3. With silent in class behaviors as the criterion variable, the overall
model was significant, R = .39, R2 = .15, F(7, 244) = 6.07, p < .001, accounting for 15% of the
shared variance in silent in class engagement behaviors. At step 1, R2 = .01, F = .96, p >
.05, course format and type were not significant predictors of engagement. After controlling for
 INTEREST AND ENGAGEMENT 115

course format and type, at step 2, R2 = .12, F = 11.20, p < .001, teacher immediacy was a
significant predictor of students’ silent in class behaviors, β = .30, t = 3.91, p < .01. At step 3,
R2 = .02, F = 3.08, p < .05, teacher immediacy remained a significant predictor of students’
engagement, β = .25, t = 3.15, p < .01.
With oral in class behaviors as the criterion variable, the overall model was significant, R =
.45, R2 = .20, F(7, 244) = 8.83, p < .001, accounting for 20% of the variance in students’ oral
in class engagement behaviors. At step 1, R2 = .13, F = 18.15, p < .001, course format was
a significant predictor of engagement, β = .35, t = 5.97, p < .01. After controlling for course
format and type, at step 2, R2 = .05, F = 4.51, p < .01, teacher immediacy was a significant
predictor of students’ oral in class behaviors, β = .24, t = 3.19, p < .01. At step 3, R2 = .03,
F = 4.49, p < .05, teacher immediacy, β = .18, t = 2.29, p < .05, and student emotional
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

interest, β = .22, t = 2.33, p < .05, significantly predicted students’ engagement.

With thinking about the course content as the criterion variable, the overall model was sig-
nificant, R = .58, R2 = .34, F(7, 244) = 18.01, p < .001, accounting for 34% of the variance
in students’ engagement behaviors. At step 1, R2 = .06, F = 8.35, p < .001, course format,
β = .17, t = 2.83, p < .01, and course type, β = .17, t = 2.71, p < .01, were significant predic-
tors of engagement. After controlling for course format and type, at step 2, R2 = .05, F =
4.88, p < .01, teacher immediacy and clarity did not emerge as significant predictors of students’
engagement behaviors. At step 3, R2 = .23, F = 41.68, p < .001, student emotional interest,
β = .35, t = 4.06, p < .001, and cognitive interest, β = .32, t = 3.78, p < .001, were significant
predictors of students’ engagement.
Using out of class engagement behaviors as the criterion variable, the overall model was sig-
nificant, R = .38, R2 = .15, F(7, 244) = 5.91, p < .001, accounting for 15% of the variance in
students’ out of class engagement behaviors. At step 1, R2 = .00, F = .07, p > .05, course
format and type were not significant predictors of engagement. After controlling for course for-
mat and type, at step 2, R2 = .10, F = 9.29, p < .01, teacher written clarity emerged as a
significant predictor of students’ engagement behaviors, β = .31, t = 3.94, p < .001. At step 3,
R2 = .04, F = 6.09, p < .01, teacher written clarity, β = .32, t = 4.13, p < .001, student
emotional interest, β = .34, t = 3.48, p < .01, and cognitive interest, β = .19, t = 1.98, p < .05,
significantly predicted students’ engagement.

STUDY FOUR: DISCRIMINANT VALIDITY

Although strong correlations may indicate construct similarity, different measures must have
divergent factor structures if they are indeed measuring similar but distinct constructs (Campbell
& Fiske, 1959; Goodboy, Martin, & Bolkan, 2009). To add further credibility to the findings
observed in Study Three, and to maximize the utility of the new measures for other scholars,
a fourth study was conducted to confirm the dimensionality of the interest and engagement
measures and to gather additional validity evidence. Discriminant validity for the scales was
assessed by calculating CFAs specifying that the two factors on the Student Interest Scale and
the four factors on the Student Engagement Scale were distinct from students’ affective learn-
ing, learner empowerment, and learning indicators. Although the Student Interest Scale may
share conceptual commonalities with other scales, discriminant validity must be provided to
 116 J. P. MAZER

further distinguish the interest scale from measures of affective learning (McCroskey, 1994) and
learner empowerment (Weber, Martin, & Cayanus, 2005). Furthermore, discriminant validity evi-
dence must also distinguish the Student Engagement Scale from common learning indicators
(Frymier & Houser, 1999). If the measurement models demonstrate adequate fit, evidence of
discriminant validity will exist for the interest and engagement scales. If the Student Interest
Scale and measures of affective learning and learner empowerment are unique and there is sup-
port for discriminant validity, the data should fit a six-factor solution—two factors of student
interest with affective learning and the three dimensions of learner empowerment each loading
onto a separate factor. If the six-factor measurement model demonstrates adequate fit, evidence
of discriminant validity will exist for the interest scale.
If the Student Engagement Scale and Learning Indicators Scale are unique and there is evi-
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

dence for discriminant validity, the data should fit a five-factor solution—four factors of student
engagement with learning indicators loading onto a separate factor. If the five-factor measure-
ment model demonstrates adequate fit, support for discriminant validity will exist for the Student
Engagement Scale.

Participants and Procedures

Participants were 183 undergraduate students enrolled in classes at a large midwestern univer-
sity (15 first-year students, 33 sophomores, 54 juniors, 78 seniors, 3 no reports). The sample
consisted of 86 males and 96 females, with an average age of 21.84 years (ranging from 18 to
38 years, SD = 4.89). One participant failed to report sex. The racial/ethnic distribution was
78.7% Caucasians, 9.8% African Americans, 4.4% Hispanics, 2.2% Asian Pacific Islanders, 1.1%
American Indian/Alaskan Native, and 3.8% Other.
All procedures were approved through the university’s IRB. Using the same method described
in Studies Two and Three (Plax et al., 1986), students were asked to identify the first class they
attend in a particular week and then use that class as the reference point for all questions on the
survey. Of the teachers, 118 were male and 65 were female. In terms of class format, participants
reported that 131 classes were mostly lecture oriented and 51 were mostly discussion oriented.
One participant did not report class format. The average class size was 17.89 students. Most of the
students (n = 117, 63.9%) reported they were enrolled in their target class because of their major,
whereas a smaller number (n = 36, 19.7%) indicated that they were taking the class because of
a general education requirement. A total of 30 students (16.4%) indicated they were taking the
class as part of a second major or minor.

Measurement

Student interest

Interest was operationalized using the Student Interest Scale developed in Studies One, Two,
and Three. Participants responded using a five point Likert scale with response options ranging
from strongly disagree to strongly agree. In this study, the measure yielded high reliability coef-
ficients: α = .95 (M = 31.38, SD = 7.94) for emotional interest; α = .88 (M = 28.25, SD =
4.77) for cognitive interest.
 INTEREST AND ENGAGEMENT 117

Student engagement

Engagement was measured using the 13-item Student Engagement Scale developed in Studies
One, Two, and Three. Participants reported how frequently they took part in each of the engage-
ment activities using a seven-point semantic differential scale with a bipolar response option
(never/very often). The results of this study revealed that the engagement scale was reliable:
silent in class behaviors α = .77 (M = 24.90, SD = 3.45); oral in class behaviors α = .91 (M =
8.87, SD = 3.60); thinking about course content α = .92 (M = 15.74, SD = 4.48); and out of
class behaviors α = .71 (M = 19.97, SD = 5.12).

Affective learning
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

Affective learning was operationalized using McCroskey’s (1994) measure. Participants com-
pleted four seven-point, four-item bipolar scales reflecting thoughts on the course content, the
instructor, and the likelihood of taking future courses in this content area with the instructor.
Alphas for the affective learning subscales were attitude about instructor, .96 (M = 23.75, SD =
5.95); attitude about content, .90 (M = 23.50, SD = 5.04); likelihood of enrolling in a similar
course, .96 (M = 21.47, SD = 7.26); likelihood of enrolling in another course with the instructor,
.95 (M = 21.40, SD = 7.55).

Learner empowerment

Weber et al.’ (2005) 18-item measure of learner empowerment was used. The instrument con-
sists of three dimensions: meaningfulness, impact, and competence that are measured with a
Likert-type scale anchored by 1 (never) and 5 (always). The measure was reliable: meaningfulness
α = .89 (M = 23.17, SD = 5.10); impact α = .79 (M = 20.57, SD = 4.64); and competence α =
.90 (M = 24.87, SD = 4.78).

Learning indicators

Frymier and Houser’s (1999) seven-item Revised Learning Indicators scale was used. The
scale reflects learning activities that students may engage in when involved in the learning
process, such as “I think about the course content outside of class” and “I see connections
between the course content and my career goals.” Participants indicated how frequently they
performed each of the behaviors using a five-point Likert-type scale (never to very often). The
alpha reliability was .79 (M = 25.37, SD = 4.72).

STUDY FOUR RESULTS

This study posited that items on the Student Interest Scale were distinct from items assess-
ing students’ affective learning and learner empowerment. Considering the standards for model
fit reported previously, the six-factor confirmatory model indicated good model fit, χ 2 (134) =
259.40, p < .05, RMSEA = .071[90% CI: .058-.084] , NNFI = .98, CFI = .99. The results indicate that
 118 J. P. MAZER

TABLE 6
Descriptive Statistics and Pearson Product-Moment Correlations for Confirmatory Factor Analyses

M SD 1 2 3 4 5 6 7 8 9 10 11 12 13

1. Emotional Interesta 31.38 7.94 –

2. Cognitive Interesta 28.25 4.77 .62 –
3. Affect Toward Contentb 23.50 5.04 .64 .69 –
4. Affect Future Coursesb 21.47 7.26 .57 .38 .50 –
5. Affect Toward 23.75 5.95 .68 .66 .78 .45 –
Instructorb
6. Affect Future Teacherb 21.40 7.55 .66 .59 .65 .52 .81 –
7. Meaningfulnessc 23.17 5.10 .63 .65 .76 .56 .62 .59 –
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

8. Impactc 20.57 4.64 .65 .52 .58 .37 .59 .51 .61 –
9. Competencec 24.87 4.78 .55 .63 .57 .34 .51 .41 .49 .50 –
10. Oral In Class 8.87 3.60 .34 .22 .22 .27 .20 .24 .21 .51 .27 –
Behaviorsd
11. Silent In Class 24.90 3.45 .45 .38 .52 .19 .51 .38 .49 .45 .36 .34 –
Behaviorsd
12. Out of Class Behaviorsd 19.97 5.12 .30 .30 .24 .23 .26 .23 .21 .23 .27 .21 .25 –
13. Thinking About Course 15.74 4.48 .52 .44 .51 .41 .32 .30 .60 .45 .24 .23 .30 .21 –
Contentd
14. Learning Indicators 25.37 4.72 .57 .52 .56 .41 .49 .45 .70 .57 .37 .24 .43 .41 .56

Note. All correlations are significant at p < .01. a Student Interest Scale. b Affective Learning Scale. c Learner
Empowerment Scale. d Student Engagement Scale.

items on the Student Interest Scale are distinct from measures of affective learning and learner
empowerment. This study also posited that items on the Student Engagement Scale were distinct
from items assessing students’ learning indicators. The five-factor confirmatory model indicated
good model fit, χ 2 (66) = 119.57, p < .05, RMSEA = .062[90% CI: .041-.081] , NNFI = .96, CFI =
.97. The findings reveal that items on the Student Engagement Scale are distinct from the measure
of learning indicators.
Correlation coefficients reported in Table 6 show statistical relationships between factors on
the Student Interest Scale and Student Engagement Scale and each of the affective learning,
learner empowerment, and learning indicator variables. Notably, all coefficients were significant.
Emotional interest and cognitive interest were positively associated with affective learning and
learner empowerment. Moreover, the factors on the Student Engagement Scale were positively
associated with learning indicators. Both the consistency and strength of these correlations—
coupled with the CFA results—suggest that the Student Interest Scale and Student Engagement
Scale are reliable and valid indicators of students’ experiences in learning contexts.

DISCUSSION

The primary purpose of the studies reported here was to develop and validate measures of student
interest and engagement. The interest and engagement instruments, which were created using
guiding theoretical frameworks and students’ open-ended responses, achieved high reliability
 INTEREST AND ENGAGEMENT 119

estimates and performed well in exploratory and confirmatory factor analyses. The results of
exploratory factor analysis in Study Two offered content validity evidence by uncovering the
dimensionality of the measures; confirmatory factor analysis in Study Three further validated this
structure. Study Four offered discriminant validity evidence, suggesting that the new measures are
distinct from scales that assess similar, yet divergent constructs.
The results of Study Three, in particular, suggest that teacher communication behaviors, such
as immediacy and clarity, can arouse students’ interest. Immediacy behaviors such as smiling,
moving close to and making eye contact with students, and using warm vocal cues and per-
sonalized examples can energize students, stimulate emotional interest, and engage students
so that they pay more attention to course content and learn more. The findings suggest that
teacher immediacy is strongly associated with student emotional interest (r = .53), a find-
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

ing consistent with prior research related to interest (Harp & Mayer, 1997). The regression
analyses illustrated that when controlling for course format (lecture vs. discussion) and type
(nonmajor vs. major), teacher immediacy and verbal clarity significantly predicted students’
emotional interest. Written clarity did not contribute significantly to students’ emotional inter-
est. This finding is understandable given that the richness of teachers’ verbal and nonverbal
behaviors might produce a more energizing effect on students than static written materials.
Although the correlational findings indicate a positive relationship between teacher imme-
diacy and student cognitive interest (r = .43), future research might further explore these
associations to clarify how teacher immediacy might influence particular forms of student
interest.
Teachers can use clarity behaviors by previewing and reviewing main points of a lesson,
defining major concepts, providing relevant examples, and creating appropriate linkages among
concepts and examples. In this manner, teachers can heighten cognitive interest in students by
focusing their selective attention on relevant information and assisting them in building internal
connections among content. Results indicate that teacher clarity appears to be strongly related to
student cognitive interest, a finding consistent with prior studies (Harp & Mayer, 1997). In fact,
verbal forms of clarity (r = .58) were more strongly associated with students’ cognitive interest
than written types of clarity (r = .36). Furthermore, the findings also suggest positive associa-
tions between verbal (r = .55) and written clarity (r = .29) and student emotional interest. The
regression analyses illustrated that when controlling for course format (lecture vs. discussion)
and type (nonmajor vs. major), teacher verbal clarity significantly predicted students’ cogni-
tive interest. Teacher immediacy and written clarity did not contribute significantly to students’
cognitive interest. This finding underscores the importance of teacher verbal clarity in leading
students to experience heightened levels of cognitive interest. Although correlational evidence
specified relationships between teacher clarity and student cognitive interest, the regression find-
ings highlight the importance of course format and type as factors affecting the degree to which
students experience cognitive interest in a particular course. Future research might examine
potential differences between lecture-based courses and classes that rely more on discussion-
based approaches to learning. While also considering differences between required courses and
elective classes, research in this vein can clarify how student interest and engagement might differ
in classes with these unique characteristics. Additional research might consider how teacher clar-
ity might impact specific types of student interest to clarify these important relationships. Overall
though, this process can lead students to experience heightened interest and, subsequently, higher
levels of engagement and learning.
 120 J. P. MAZER

Future research might examine the interaction of teacher immediacy and clarity and its impact
on student interest. Compelling evidence from prior research points in favor of the additivity
hypothesis (Chesebro, 2003; Comadena, Hunt, & Simonds, 2007), a proposition that states that
both clarity and immediacy are positive teaching behaviors and that students will benefit most
when both are present. To date, virtually no studies that directly test interactions between immedi-
acy and clarity have found that students’ achievement suffers in high immediacy conditions. Can
teachers suppress the negative implications of unclear teaching by using immediacy to heighten
students’ cognitive interest? Or, can non-immediate teachers heighten emotional interest through
clear teaching?
Student emotional interest and cognitive interest are both positive experiences that can lead
to important benefits for students. Much like teacher immediacy and clarity, results suggest that
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

students’ emotional and cognitive interest are associated with their engagement, including silent
(emotional interest: r = .32; cognitive interest: r = .26) and oral (emotional interest: r = .32; cog-
nitive interest: r = .22) in class behaviors and out of class activities (emotional interest: r = .26;
cognitive interest: r = .15). Notably, the findings of the present study indicate a moderate rela-
tionship between students’ emotional (r = .53) and cognitive interest (r = .51) and their tendency
to think about the course content. The regression analyses indicated that teacher immediacy sig-
nificantly predicted students’ silent in class engagement behaviors, including listening attentively
to the instructor and classmates. This finding is not surprising given the tendency for immedi-
acy behaviors to heighten student attention in the classroom (Christophel, 1990; Frymier, 1993,
1994). Notably, students’ emotional interest significantly predicted their oral in class engagement
behaviors. In essence, students who are energized, passionate, and emotionally connected to a
specific course are more likely to be verbally engaged in the class. Student emotional and cogni-
tive interest significantly predicted students’ tendency to think more deeply about the course and
its content, including how they might utilize the material in everyday life and how it might benefit
them in the future. The findings underscore the importance of emotional and cognitive interest as
key contributors to students’ tendency to engage in greater cognitive thought about the course.
Furthermore, teacher written clarity and students’ emotional and cognitive interest significantly
predicted increases in students’ out of class engagement behaviors. In other words, written clarity
behaviors and students’ feelings of emotional and cognitive interest led them to study more, talk
about the course material with others, and explore the subject matter in greater detail.
Although scholars have examined the direct effects of emotional and cognitive interest on
student learning (Harp & Mayer, 1997; Titsworth, 2001b), prior research has not explored how
the two variables work together to influence student engagement. Are students most engaged
when they experience increased degrees of emotional and cognitive interest? Is emotional inter-
est a requirement for engaged student learning? In addition, what is the relationship between
students’ cognitive interest and their perceptions of their self-efficacy? Future research can seek
to build evidence in this area. It may be that teachers can, through their communication behav-
ior, “pull” students toward the course content and lead them to frequently think about how they
can use the material in their everyday lives and in the future. Through this process, students
can develop meaningful connections to the particular discipline (Bruner, 1960/1977). The Study
Three findings support emotional response theory by explaining how teacher communication
behaviors (e.g., immediacy and clarity) can stimulate positive emotional responses (e.g., emo-
tional and cognitive interest) that lead students to approach rather than avoid engaged learning
situations. Moreover, the associations uncovered in Study Three offer construct validity evidence
 INTEREST AND ENGAGEMENT 121

for the instruments, as interest and engagement appear to be positively associated with teacher
immediacy and clarity.
Mottet et al.’s (2006) original explanation of emotional response theory highlights teach-
ers’ communication behaviors as the initiating step in classroom emotional processes. Initiating
behaviors include teacher immediacy, affinity-seeking, and behavioral alteration techniques as
actions that introduce emotional meaning into teacher-student interactions. The present study fur-
ther develops emotional response theory by explaining how clear teaching behaviors can stimulate
in students heightened emotional responses (e.g., cognitive interest). Furthermore, responding to
Mottet et al.’s call to explore “specific instructional communication behaviors or conditions [that]
lead to enhanced student emotional responses” (p. 264), the present study explained how specific
student emotional responses—emotional interest and cognitive interest—can lead students to be
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

more engaged in the learning process. Future research using emotional response theory might
explore how specific emotional responses are developed over time within a relationship or within
a particular class. If interest is regarded as an emotion that yields long-term benefits for students
(Bruner, 1960/1977), future studies in this area can capitalize on opportunities to longitudinally
examine how emotional and cognitive interest develops over the course of an academic term. Such
studies could potentially explore how certain teacher behaviors and student emotional responses
can lead to various beneficial (e.g., greater engagement) or detrimental (e.g., stress and burnout)
outcomes.
Any study must be interpreted within the limitations imposed by the research design. The
use of self-report methods warrants caution, as does the nonexperimental design of the research.
Statements of causality based on the results of statistical techniques useful for detecting rela-
tionships between constructs must be treated with caution given the correlational data analyzed.
Future research might address these limitations through structural equation modeling to better
understand the processes at play in the classroom. Full experiments with random samples of
students across different educational levels can examine the impact that teacher communication
behavior might have on student interest and engagement. Future experimental studies can explore
possible reciprocal effects between teacher and student engagement. That is, do engaged teach-
ers produce engaged students? Or, do engaged students produce engaged teachers? Although the
present studies did not account for whether or not class attendance was required and graded as
part of course requirements, future research might address how students’ attendance can influence
their engagement. This factor is especially important in light of specific in class behaviors (e.g.,
listening attentively to the instructor during class, attended class) that emerged as stable items on
the Student Engagement Scale and reflect students’ engagement in the learning process.
Even though the interest and engagement instruments fared well in exploratory and
confirmatory factor analyses and demonstrated robust scale reliabilities, the instruments could be
further improved and scrutinized. Individual items might require revision to correct for redundan-
cies in wording. Subjecting the instruments to Horn’s (1965) Parallel Analysis, which compares
random data to the data collected in order to identify which factors emerge as having higher
eigenvalues than the random data, would help to further build evidence supporting the factor
structure of the scales. Moreover, subjecting the interest and engagement scales to a multitrait,
multimethod analysis would be a useful step in expanding the construct validity evidence for
the instruments (Campbell & Fiske, 1959). Furthermore, future research might administer the
scales to less traditional groups of students to further verify the factor structures of the mea-
sures. As validity evidence for the interest and engagement scales is compiled through future
 122 J. P. MAZER

research, educators might consider teacher clarity instruments and the Student Interest Scale as
diagnostic measures of instructor quality. Teachers might consider implementing these measures
as part of their midterm and end-of-term evaluations to assess students’ perceptions of the teach-
ing and learning process. While the interest and engagement scales might offer practical benefits
for teachers and administrators who evaluate classroom teaching, the instruments can serve to
advance scholarly conversations related to the role of communication in teaching and learning.
Teachers are, according to Bruner (1960/1977), communicators, immediately personal sym-
bols of the educational process, and figures that can ignite in students an interest for a particular
subject area. Teachers have in their arsenals a repertoire of behaviors that can positively impact
students in ways that have immediate and long lasting effects on emotion and learning. Among
them, immediacy and clarity are two such behaviors that can have substantive effects on students
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

and promote a classroom environment where emotional and cognitive interest and engagement
can flourish. In the end, heightened student interest and engagement can lead to significant
improvements in their learning, satisfaction, and success in school.

REFERENCES

Andersen, J. F. (1979). Teacher immediacy as a predictor of teaching effectiveness. In D. Nimmo (Ed.), Communication
yearbook 3 (pp. 543–559). New Brunswick, NJ: Transaction Books.
Andersen, P. A., & Andersen, J. F. (1982). Nonverbal immediacy in instruction. In L. Barker (Ed.), Communication in the
classroom (pp. 98–120). Engelwood Cliffs, NJ: Prentice Hall.
Baldwin, J. M. (1906). Thought and things: A study of the development and meaning of thought (Vol. I). New York, NY:
Macmillan.
Bandura, A. (1977). Social learning theory. Upper Saddle River, NJ: Prentice Hall.
Baxter, L. A., & Babbie, E. (2004). The basics of communication research. Belmont, CA: Wadsworth.
Berlyne, D. E. (1960). Conflict, arousal and curiosity. New York, NY: Grove Press.
Bloom, B. S. (Ed.). (1956). Taxonomy of educational objectives: Handbook 2: The cognitive domain. New York, NY:
Longmans, Green.
Bomia, L., Beluzo, L., Demeester, D., Elander, K., Johnson, M., & Sheldon, B. (1997). The impact of teaching strategies
on intrinsic motivation. (ERIC Document Reproduction Service No. ED 418925).
Bourhis, J., & Allen, M. (1992). Meta-analysis of the relationship between communication apprehension and cognitive
performance. Communication Education, 41, 68–76.
Brophy, J. (1983). Conceptualizing student motivation. Educational Psychologist, 18, 200–215.
Bruner, J. S. (1960/1977). The process of education. Cambridge, MA: Harvard University Press.
Bush, A. J., Kennedy, J. J., & Cruickshank, D. R. (1977). An empirical investigation of teacher clarity. Journal of Teacher
Education, 28, 53–58.
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix.
Psychological Bulletin, 56, 81–105.
Chesebro, J. L. (2003). Effects of teacher clarity and nonverbal immediacy on student learning, receiver apprehension,
and affect. Communication Education, 52, 135–147.
Christophel, D. (1990). The relationship among teacher immediacy behaviors, student motivation, and learning.
Communication Education, 39, 323–340.
Comadena, M. E., Hunt, S. K., & Simonds, C. J. (2007). The effects of teacher clarity, nonverbal immediacy, and caring
on student motivation, affective and cognitive learning. Communication Research Reports, 24, 241–248.
Comstock, J., Rowell, E., & Bowers, J. W. (1995). Food for thought: Teacher nonverbal immediacy, student learning, and
curvilinearity. Communication Education, 44, 251–266.
Connell, J. P., & Wellborn, J. G. (1991). Competence, autonomy, and relatedness: A motivational analysis of self-esteem
processes. In M. R. Gunnar & L. A. Sroufe (Eds.), Self processes in development: Minnesota symposium on child
psychology (pp. 43–77). Hillsdale, NJ: Erlbaum.
 INTEREST AND ENGAGEMENT 123

Creswell, J. W. (2002). Research design: Qualitative, quantitative, and mixed method approaches (2nd ed.). Thousand
Oaks, CA: Sage.
Creswell, J. W., Plano Clark, V. L., Gutmann, M. L., & Hanson, W. E. (2003). Advanced mixed method research
designs. In A. Taskakkori & C. Teddlie (Eds.), Handbook of mixed methods research in the behavioral social sciences
(pp. 209–240). Thousand Oaks, CA: Sage.
Deci, E. L. (1992). The relation of interest to the motivation of behavior: A self-determination theory perspective. In
K. A. Renninger, S. Hidi, & A. Krapp (Eds.), The role of interest in learning and development (pp. 43–70). Hillsdale,
NJ: Erlbaum.
de Vaus, D. (2001). Research design in social research. London, England: Sage.
DeVellis, R. F. (2003). Scale development: Theory and applications (2nd ed.). Thousand Oaks, CA: Sage.
Dewey, J. (1913). Interest and effort in education. Boston, MA: Riverside Press.
Dewey, J. (1916). Democracy and education: An introduction to the philosophy of education. New York, NY: The Free
Press.
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

Frymier, A. B. (1993). The relationships among communication apprehension, immediacy and motivation to study.
Communication Reports, 6, 8–17.
Frymier, A. B. (1994). A model of immediacy in the classroom. Communication Quarterly, 42, 133–144.
Frymier, A. B., & Houser, M. L. (1999). The revised learning indicators scale. Communication Studies, 50, 1–12.
Frymier, A. B., Shulman, G. M., & Houser, M. L. (1996). The development of a learner empowerment measure.
Communication Education, 45, 181–199.
Glaser-Zikuda, M., & Fuss, S. (2008). Impact of teacher competencies on student emotions: A multi-method approach.
International Journal of Educational Research, 4, 136–147.
Goodboy, A. K., Martin, M. M., & Bolkan, S. (2009). The development and validation of the student communication
satisfaction scale. Communication Education, 58, 372–396.
Greenwood, C. R. (1991). Longitudinal analysis of time, engagement, and achievement in at-risk versus non-risk students.
Exceptional Children, 57, 521–534.
Harp, S. F., & Mayer, R. E. (1997). The role of interest in learning from scientific text and illustration: On the distinction
between emotional interest and cognitive interest. Journal of Educational Psychology, 89, 92–102.
Herbart, J. F. (1965). Allgemeine Pädagogik, aus dem Zweck der Erziehung abgeleitet. In J. F. Herbart (Ed.),
Pädagogische Schriften (Vol. 2). Dusseldorf, Germany: Kupper. (Original work published 1806)
Hidi, S., & Anderson, V. (1992). Situational interest and its impact on reading and expository writing. In K. A.
Renninger, S. Hidi, & A. Krapp (Eds.), The role of interest in learning and development (pp. 215–238). Hillsdale, NJ:
Erlbaum.
Hidi, S., & Baird, W. (1988). Strategies for increasing text-based interest and students’ recall of expository texts. Reading
Research Quarterly, 23, 465–483.
Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179–185.
Jones, S., & Wirtz, J. (2006). How does the comforting process work? An empirical test of an appraisal-based model of
comforting. Human Communication Research, 32, 217–243.
Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd ed.). New York, NY: Guilford.
Krapp, A., Hidi, S., & Renninger, K. A. (1992). Interest, learning, and development. In K. A. Renninger, S. Hidi, &
A. Krapp (Eds.), The role of interest in learning and development (pp. 3–25). Hillsdale, NJ: Erlbaum.
Kroman Myers, K., & Oetzel, J. G. (2003). Exploring the dimensions of organizational assimilation: Creating and
validating a measure. Communication Quarterly, 51, 438–457.
Ledbetter, A. M. (2009). Measuring online communication attitude: Instrument development and validation.
Communication Monographs, 76, 463–486.
Lupton, D. (1994). Medicine as culture: Illness, disease and the body in Western society. London, UK: Sage.
MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size for
covariance structure modeling. Psychological Methods, 1, 130–149.
McCroskey, J. C. (1994). Assessment of affect toward communication and affect toward instruction in communication. In
S. Morreale & M. Brooks (Eds.), 1994 SCA summer conference proceedings and prepared remarks: Assessing college
student competence in speech communication. Annandale, VA: Speech Communication Association.
McCroskey, J. C., Sallinen, A., Fayer, J. M., Richmond, V. P., & Barraclough, R. A. (1996). Nonverbal immediacy and
cognitive learning: A cross-cultural investigation. Communication Education, 45, 200–211.
McCroskey, J. C., & Young, T. J. (1979). The use and abuse of factor analysis in communication research. Human
Communication Research, 5, 375–382.
 124 J. P. MAZER

McGarity, J. R., & Butts, D. P. (1984). The relationship among teacher classroom management behavior, student engage-
ment, and student achievement of middle and high school science students of varying aptitude. Journal of Research
in Science Teaching, 21, 55–61.
Mehrabian, A. (1981). Silent messages: Implicit communication of emotions and attitudes. Belmont, CA: Wadsworth.
Miller, K. I. (2007). Compassionate communication in the workplace: Exploring processes of noticing, connecting, and
responding. Journal of Applied Communication Research, 35, 223–245.
Mottet, T. P., Frymier, A. B., & Beebe, S. A. (2006). Theorizing about instructional communication. In T. P. Mottet,
V. P. Richmond, & J. C. McCroskey (Eds.), Handbook of instructional communication (pp. 255–282). Boston, MA:
Pearson.
Myers, S. A. (2010). Using the Perry Scheme to explore college student classroom participation. Communication
Research Reports, 27, 123–130.
Myers, S. A., Martin, M. M., & Knapp, J. L. (2005). Perceived instructor in-class communicative behaviors as a predictor
of student participation in out of class communication. Communication Quarterly, 53, 437–450.
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

National Center for Educational Statistics. (2010). The condition of education. Retrieved from http://nces.ed.gov/
programs/coe
Neuendorf, K. A. (2002). The content analysis guidebook. Thousand Oaks, CA: Sage.
O’Mara, J., Allen, J. L., Long, K. M., & Judd, B. (1996). Communication apprehension, nonverbal immediacy, and
negative expectations for learning. Communication Research Reports, 13, 109–128.
Owen, W. F. (1984). Interpretive themes in relational communication. Quarterly Journal of Speech, 70, 274–287.
Piaget, J. (1981). Intelligence and affectivity: Their relationship during child development. In T. A. Brown & C. E. Kaegi
(Eds.), Annual review monographs. Palo Alto, CA: Annual Reviews.
Plax, T. G., Kearney, P., McCroskey, J. C., & Richmond, V. P. (1986). Power in the classroom VI: Verbal control strategies,
nonverbal immediacy, and affective learning. Communication Education, 35, 43–55.
Renninger, K. A. (1992). Individual interest and development: Implications for theory and practice. In K. A. Renninger,
S. Hidi, & A. Krapp (Eds.), The role of interest in learning and development (pp. 361–395). Hillsdale, NJ: Erlbaum.
Renninger, K. A., Hidi, S., & Krapp, A. (Eds.). (1992). The role of interest in learning and development. Hillsdale, NJ:
Erlbaum.
Richmond, V. P., Gorham, J. S., & McCroskey, J. C. (1987). The relationship between selected immediacy behaviors and
cognitive learning. In M. McLaughlin (Ed.), Communication Yearbook 10 (pp. 574–590). Beverly Hills, CA: Sage.
Richmond, V. P., McCroskey, J. C., & Johnson, A. E. (2003). Development of the Nonverbal Immediacy Scale (NIS):
Measures of self- and other-perceived nonverbal immediacy. Communication Quarterly, 51, 502–515.
Richmond, V. P., McCroskey, J. C., Kearney, P., & Plax, T. G. (1987). Power in the classroom VII: Linking behavior
alteration techniques to cognitive learning. Communication Education, 36, 1–12.
Rodríguez, J. I., Plax, T. G., & Kearney, P. (1996). Clarifying the relationship between teacher nonverbal immediacy and
student cognitive learning: Affective learning as the central causal mediator. Communication Education, 45, 293–305.
Rosenshine, B. V. (1979). Content, time and direct instruction. In P. Peterson & H. Walberg (Eds.), Research on teaching:
Concepts, findings, and implications. Berkeley, CA: McCutchan.
Schiefele, H. (1986). Interest: New answers to an old problem. Zeitschrift fär Pädagogik, 32, 153–162.
Schiefele, U. (1991). Interest, learning, and motivation. Educational Psychologist, 26, 299–323.
Schrodt, P., Witt, P. L., Turman, P. D., Myers, S. A., Barton, M. H., & Jernberg, K. A. (2009). Instructor credibility
as a mediator of instructors’ prosocial communication behaviors and students’ learning outcomes. Communication
Education, 58, 350–371.
Simonds, C. J. (1997). Classroom understanding: An expanded notion of teacher clarity. Communication Research
Reports, 14, 279–290.
Skinner, E. A. (1991). Development and perceived control: A dynamic model of action in context. In M. R. Gunnar
& L. A. Sroufe (Eds.), Self processes in development: Minnesota symposium on child psychology (pp. 167–216).
Chicago, IL: University of Chicago Press.
Skinner, E., Furrer, C., Marchland, G., & Kindermann, T. (2008). Engagement and disaffection in the classroom: Part of
a larger motivational dynamic? Journal of Educational Psychology, 100, 765–781.
Thompson, B., & Mazer, J. P. (2009). College student ratings of student academic support: Frequency, importance, and
modes of communication. Communication Education, 58, 433–458.
Thorndike, E. L. (1935). Adult interests. New York, NY: Macmillan.
Titsworth, B. S. (2001a). The effects of teacher immediacy, use of organizational cues, and students’ notetaking on
cognitive learning. Communication Education, 50, 283–297.
 INTEREST AND ENGAGEMENT 125

Titsworth, B. S. (2001b). Immediate and delayed effects of interest cues and engagement cues on students’ affective
learning. Communication Studies, 52, 169–180.
Titsworth, B. S. (2004). Students’ notetaking: The effects of teacher immediacy and clarity. Communication Education,
53, 305–320.
Titsworth, S., & Mazer, J. P. (2010). Clarity in teaching and learning: Conundrums, consequences, and opportunities. In
D. L. Fassett & J. T. Warren (Eds.), The Sage handbook of communication and instruction (pp. 241–262). Thousand
Oaks, CA: Sage.
Titsworth, S., Novak, D. R., Hunt, S. K., & Meyer, K. R. (2004, May). The effects of teacher clarity on affective and
cognitive learning: A causal model of clear teaching behaviors. Paper presented at the meeting of the International
Communication Association, New Orleans, LA.
Titsworth, S., Quinlan, M. M., & Mazer, J. P. (2010). Emotion in teaching and learning: Development and validation of
the classroom emotions scale. Communication Education, 59, 431–452.
Tobias, S. (1994). Interest, prior knowledge, and learning. Review of Educational Research, 64, 37–54.
Downloaded by [University of Pennsylvania] at 11:28 13 April 2016

Treadwell, D. F., & Harrison, T. M. (1994). Conceptualizing an assessing organizational image, commitment, and
communication. Communication Monographs, 61, 63–84.
Vangelisti, A. L., Crumley, L. P., & Baker, J. L. (1999). Family portraits: Stories as standards for family relationships.
Journal or Social and Personal Relationships, 16, 335–368.
Weber, K. (2003). The relationship of interest to internal and external motivation. Communication Research Reports, 20,
376–383.
Weber, K. (2004). The relationship between student interest and teacher’s use of behavior alteration techniques.
Communication Research Reports, 21, 428–436.
Weber, K., Martin, M. M., & Cayanus, J. L. (2005). Student interest: A two-study re-examination of the concept.
Communication Quarterly, 53, 71–86.
Witt, P. L., & Wheeless, L. R. (2001). An experimental study of teachers’ verbal and nonverbal immediacy and students’
affective and cognitive learning. Communication Education, 50, 327–342.
Witt, P. L., Wheeless, L. R., & Allen, M. (2004). A meta-analytical review of the relationship between teacher immediacy
and student learning. Communication Education, 71, 184–207.
Woolfolk, A. E., & McCune-Nicolich, L. (1984). Educational psychology for teachers (2nd ed.). Englewood Cliffs, NJ:
Prentice-Hall.