Download Public Speaking Apprehension and Motivations and more Study Guides, Projects, Research Public Sociology in PDF only on Docsity!
ISSUES IN ACCOUNTING EDUCATION
Vol. 24, No. 3 August 2009 pp. 265–
Public Speaking Apprehension (PSA),
Motivation, and Affect among Accounting
Majors: A Proof-of-Concept Intervention
Tim C. Miller and Dan N. Stone
ABSTRACT: The importance of public speaking (PS) skills to professional accounting success motivates improving students’ self-perceptions of these skills. In addition, ev- idence of higher levels of public speaking apprehension (PSA) among accounting ma- jors makes understanding and working with students’ affective (emotive) reactions to PS critical to their future success. In three studies, we design and implement an inter- vention based on principles from self-determination theory (SDT) and motivational in- terviewing (MI). Its purpose is to improve students’ PSA and PS motivation; it includes substantive PS instruction, dialogues, nonjudgmental feedback, and interpersonal sup- port. The results of the three ‘‘proof-of-concept’’ interventions (Study 1, n 23; Study 2, n 14; Study 3, n 36) suggest improvements in students’ perceptions of their PS cognition, motivation, and affect. Despite the limitations of self-reported measures and exclusively graduate participants, the results suggest that (1) the interventions, described in appendices, may merit replication and extension, and (2) SDT- and MI- based interventions may prove useful in addressing aspects of accountancy pedagogy that induce student apprehension.
INTRODUCTION
P
ublic speaking (PS) is an important determinant of professional accounting success.
For example, practice analyses and surveys indicate that oral communication is a core
function of professional accounting work (Albrecht and Sack 2000; National Center
for O*NET Development 2007; Siegel and Sorensen 1999). Accordingly, building com-
munication skills is of essential import to accounting students and the accounting curricu-
lum. In this paper, we focus on a subset of the communication skills needed for professional
success in accounting. Specifically, we investigate whether an intervention based on moti-
vational interviewing (MI) and self-determination theory (SDT) increases PS motivation
and reduces PSA (public speaking apprehension).
Oral communication apprehension (OCA) includes four related fears: group discussion,
meetings, interpersonal communication, and PS (McCroskey 1982; Aly and Islam 2003,
2005; Gardner et al. 2005). The clinically diagnosable fear of PS, called glossophobia, is
the most common adult phobia (irrational fear). For example, Stein et al. (1996) surveyed
Tim C. Miller is a Ph.D. candidate, and Dan N. Stone is a Professor, both at the University
of Kentucky.
The authors gratefully acknowledge the comments of Linda Rachula, John Hill, Red Medley, Candace Witherspoon, and two anonymous reviewers at the 2007 Accounting, Behavior, and Organizations (ABO) Conference; the com- ments of the editor, Kent St. Pierre, and two anonymous reviewers; and financial support of the Gatton College of Business and Economics and the Von Allmen School of Accountancy. Data are available from the first author.
266 Miller and Stone
499 Canadian residents and found evidence of glossophobia among one-third of respon-
dents. Evidence suggests that PS fear is more common than is the fear of death. For
example, large sample survey data indicates that about 42 percent of respondents had PSA
while only 19 percent feared death (Wallechinsky et al. 1977, 469–470). These data suggest
that Jerry Seinfeld’s quip that ‘‘the average person at a funeral would rather be in the casket
than doing the eulogy’’ may not be hyperbole (Wikipedia 2007). PS fear is not unique to
the untrained and uneducated; it is also common among senior managers (Anonymous
2007; Huber 2005).
Learning is often complicated by students’ cognitive (e.g., distracting thoughts) and
affective (e.g., fear) anticipations of and responses to learning content and educational
environments. Reducing glossophobia and improving PS skills is difficult because of the
complex interaction of cognitive, affective, and physiological anticipations and responses
to PS. Human and mammalian anticipations of and reactions to stressful events are often
referred to as the ‘‘fight or flight’’ response (Marmot and Wilkinson 2006). Glossophobia
is a dysfunctional reaction to the fight or flight response; it is characterized by negative
cognitions (e.g., ‘‘I’m going to fail.’’) and negative affect (e.g., feelings of fear and incom-
petence). These mental changes are preceded, or triggered, by physiological and biological
stress responses that include an increased heart rate and the release of cortisol into the
blood stream, which increases blood pressure, blood sugar levels, and suppresses autoim-
mune and immune system responses (al’Absi et al. 1997; Buchanan et al. 1999; Beatty and
Behnke 1991).^1 These complex mental and physiological changes increase PSA and reduce
PS motivation, i.e., one’s willingness to seek or accept opportunities to speak in public.
SDT provides a psychological theory, and MI a set of methods from counseling practice
that are, to our knowledge, unexplored but potentially efficacious in addressing the psy-
chological and physiological impediments to reducing PSA and improving PS effectiveness.
SDT and MI, when combined with concurrent PS instruction, may be useful in addressing
the complex nexus of affective, cognitive, motivational, and substantive learning impedi-
ments to PS success. These methods are potential alternatives, or supplements, to existing
methods such as systematic desensitization (e.g., McCroskey et al. 1970; McCroskey et al.
1983) that are efficacious in reducing PSA.
Herein, we report the results of three intervention variations based on SDT principles
and MI practices. The semester-long interventions are designed to develop PS skills and
reduce PS anxiety. Although subject to a set of important limitations, our results suggest
that the interventions increased student confidence in, and reduced their anxiety about, PS.
The results may justify replication and extension to other accounting curriculum applica-
tions where affect, e.g., fears, impedes learning. For example, affective responses may
impede the learning of ‘‘social’’ or ‘‘emotional’’ intelligence (Goleman 1995, 2000; Stone
et al. 2000) or impede preparation for the CPA examination.
Five sections follow this introduction: (1) ‘‘Motivation and Literature Review,’’ which
explores evidence related to the importance of PS skills and PSA on professional accounting
success, (2) ‘‘Theory and Hypotheses,’’ which discusses the theory that underlies our in-
terventions, hypotheses, and metrics, (3) ‘‘Research Method,’’ which discusses the inter-
vention methods and differences, (4) ‘‘Results,’’ including benchmarking against our pre-
vious results, and (5) ‘‘Limitations and Conclusions.’’
(^1) See also Weick (1983) for a general discussion of the importance, and deleterious effects of, stress in accounting
work.
268 Miller and Stone
among students from disadvantaged socioeconomic backgrounds that was designed to im-
prove four aspects of students’ professionalism: career knowledge, job search process, team
learning, and communication skills. The PS intervention consisted of students presenting
with self- and peer-evaluations of the presentations. Comparison of pre- and post-measures
of self-evaluations of presentation and oral communication skills indicated improvement
consistent with intervention success. Smith and King (2004) examined student reactions to
the wording of critiques of their PS performance. Results indicated that more respectful,
less pejorative critiques improved PS performance more than did less respectful, more
pejorative critiques.
Alternatively, however, McCroskey and colleagues (for a summary, see McCroskey et
al. 1983) provide evidence from 15 years of research that while systematic desensitization
techniques reduce OCA and PSA, communication courses that do not include elements that
specifically address the affective (emotional) components of PS are ineffective at reducing
OCA and PSA. Consistent with McCroskey et al.’s (1983) findings, evidence suggests that
accounting education, in the absence of specific curriculum interventions, has no effect on
OCA. For example, Aly and Islam (2003) administered the Personal Report of Communi-
cations Apprehension (PRCA-24) instrument, which assesses OCA and PSA, to 151 first-
year, 125 final-year, and 58 graduate students in accountancy. Results indicated no differ-
ences in OCA among the three samples. Similar results are reported among non-U.S.
samples of accounting students (Hassall et al. 2000; Gardner et al. 2005; Aly and Islam
Based on a review of previous literature, SL (1990) recommended three intervention
strategies that they argued would be useful in reducing OCA: assertiveness training, sys-
tematic desensitization, and cognitive restructuring. Assertiveness training and cognitive
restructuring are both cognitive approaches to reducing OCA. Assertiveness training ex-
plicitly focuses on skill development, such as improving ‘‘eye contact, distance between
communicators, facial expression, gestures, and postures and body orientation.’’ Cognitive
restructuring consists of identifying ‘‘negative self-statements [that] represent irrational ov-
ergeneralizations’’ (SL 1990, 190) and recasting these as positive statements. Students prac-
tice rethinking and refuting negative self-statements in thought experiments and role-playing
exercises. Systematic desensitization implementations attempt to reduce OCA through, pri-
marily, inducing changes in affective responses. Interventions involve progressively imag-
ining more stressful PS situations while maintaining a state of deep relaxation. This method
associates the stressor event with a relaxed state, instead of the previous high-arousal,
negative affective state.
We are aware of one accounting curriculum intervention that successfully reduced OCA
among accounting students. RH (1994) designed and implemented an intervention to reduce
OCA that included assertiveness training, trust-building, and social support (see RH 1994,
288–289) across four exercises: cold classroom calls, meetings and discussions with visiting
professionals, office visits and interviews of professionals, and oral presentations. Interven-
tion participants were students in an advanced managerial accounting class; control group
participants were Beta Alpha Psi members not enrolled in the class. Pre- to post-intervention
comparisons supported the intervention’s success in reducing OCA and PSA.
We next describe an alternative intervention based in differing core principles and
methods from that of RH (1994) that is designed to improve PS skills and motivation,
and reduce PSA, among accountancy students. Our focus is on PSA, not OCA. To increase
internal validity, e.g., to test for halo effects from the intervention, we also collect and
report measures of OCA.
Public Speaking Apprehension (PSA), Motivation, and Affect among Accounting Majors 269
THEORY AND HYPOTHESES
A Self-Determination Theory (SDT) and Motivational Interviewing (MI)-Based
Intervention to Reduce PSA
SDT is rooted in a set of explicit assumptions about human nature and motivation (e.g.,
Deci and Ryan 1985, 2008; Ryan and Deci 2000). Humans are inherently motivated to
grow and achieve, and will fully commit to and engage in even uninteresting tasks when
their meaning and value is understood. According to SDT, humans have three core psy-
chological needs: competence, relatedness, and autonomy. Competence concerns the belief
that one has the ability to influence important outcomes. Relatedness concerns the need
to have satisfying and supportive social relationships. Finally, autonomy does not refer to
independence, but rather to the necessity of volitional choice of inter- or independence.
Satisfying human needs for autonomy, competence, and relatedness creates sustainable,
enduring motivation and reduces negative, performance-related affect. Increasing self-
perceived autonomy, competence, and relatedness increases productivity, creativity, and
happiness (Deci and Ryan 1985; Ryan and Deci 2000). For example, within education, an
SDT-based intervention increased students’ interest and engagement in learning activities
(Reeve et al. 2004) and improved student learning in medical school (Williams and Deci
1998). Evidence also suggests that students who perceive their instructors as more suppor-
tive become more autonomous in their own learning, which also increases self-perceived
competence (Williams and Deci 1996).
Self-determination theory (SDT) is closely aligned with a set of clinical psychology
practice methods called ‘‘motivational interviewing’’ (MI) (Vansteenkiste and Sheldon 2006;
Markland et al. 2005). MI is a client-centered counseling style that assists clients in ad-
dressing problematic behaviors that impede their success and happiness (Moyers 1998;
Rollnick and Miller 1995). While the original application of MI was to alcoholism, MI has
found increasing application in counseling and education (Miller and Rollnick 2002). Ac-
cording to Miller et al. (1992), the core principles underlying MI are:
1. Express empathy. Teachers and counselors work to see the world through the students’
or clients’ eyes and to understand the students’ or clients’ feelings and experiences.
2. Support self-efficacy. Teachers and counselors support students’ and clients’ realistic
beliefs that meaningful change is possible and achievable.
3. Roll with resistance. Expressions of skepticism and doubt are never challenged or
disputed, but are reacted to with empathy and encouragement.
4. Develop discrepancy. Meaningful change occurs when students or clients perceive dis-
crepancy between current and desired behaviors. Teachers and counselors make salient
the discrepancy between students’ and clients’ current and desired behaviors.
Some argue that SDT lacks corresponding practical (i.e., clinical and pedagogical)
methods while MI lacks an underlying theory. Because of the close linkages between SDT
and MI, we applied and adapted constructs and methods from both sources. To our knowl-
edge, SDT and MI have not been applied to improving self-perceived PS skills and reducing
PSA. However, previous applications of SDT and MI to creating learning environments
that provide interpersonal and emotional support, along with substantive instruction (Black
and Deci 2000; Williams and Deci 1996, 1998) appear uniquely suited to addressing PSA.
Both SDT and MI emphasize the creation of an environment that acknowledges and sup-
ports individual feelings of autonomy, competence, and relatedness. We designed and im-
plemented three variations of an SDT- and MI-based intervention as a ‘‘proof-of-concept’’
Public Speaking Apprehension (PSA), Motivation, and Affect among Accounting Majors 271
TABLE 1
Comparison of Proposed or Implemented PSA Interventions in Five Studies
(a)Conceptual Intervention(Independent) Variables
(b) TheorySource
(c)
Target (Dependent)
Construct
(d)
Stanga andLadd (1990)
(e)
Ruchala andHill (1994)
(f) Study 1
(g) Study 2
(h) Study 3
PS Instruction (Lecture and / orReadings)
POC
PS knowledge and skill
Supporting Choice andAutonomy / AssertivenessTraining
SDT/ MI,Adler (1977)
Affect and motivation
Instructor Dialogues withOpen Questions, Reflection,Affirmation, and Summary(OARS)
MI
PS knowledge and skill,affect and motivation
email dialogue
Acknowledging andSupporting Feelings Aboutand Development of PSCompetence
SDT/ MI
Affect and motivation
Factual, ObservationalFeedback without Judgment
SDT/ MI
PS knowledge and skill,affect and motivation
taping
Systematic Desensitization
Wenrich et al.(1976)
Affect and motivation
Cognitive Restructuring /Awareness of Self-Statements
Beck et al.(2005)
PS knowledge and skill,affect and motivation
Key:POC
principles of oral communication;
SDT
self-determination theory;
MI
motivational interviewing; and
PS
public speaking.
272 Miller and Stone
Appendix A, Panel 1, presents these measures; we predict that the SDT- and MI-based
intervention will:
H1: Increase participants’ PS motivation.
H2: Increase participants’ positive PS SS.
H3: Decrease participants’ negative PS SS.
Measures and Hypotheses in Studies 2 and 3: OCA, PSA, Affect, and PS Sub-Domain
Skill
In Studies 2 and 3, we add three sets of measures to reduce mono-measurement bias
and enhance comparability with previous research: an assessment of OCA and PSA (PRCA-
24) (Appendix A, Panel 3), a state (not trait) PS-related assessment of positive and negative
affect (PANAS) (Appendix A, Panel 4) (Watson et al. 1988), and an assessment of PS self-
perceived performance in six sub-domains (Appendix A, Panel 2).
OCA and PSA: Hypotheses 4 through 7
Adding the PRCA-24 increases the comparability of our intervention to existing com-
munication and accounting research (e.g., Simons et al. 1995; Fordham and Gabbin 1996;
Hassall et al. 2000; Gardner et al. 2005; Arquero et al. 2007; RH 2004; SL 1990). In
addition, the PRCA-24 assesses both OCA and PSA. Since our interventions are targeted
at PSA, collecting the PRCA-24 allows us to: (1) assess whether there are halo, i.e., cross-
over improvements, from our intervention on three other oral communication dimensions:
group discussion, meetings, and interpersonal communication, and (2) control for the pos-
sibility that improvements in the treatment groups, if any, may be explained by effects that
are unrelated to intervention, e.g., group maturation or history (Shadish et al. 2002).
Hypotheses 4 through 6 predict that the intervention, which is designed to decrease
PSA, will not impact the three dimensions of OCA that are unrelated to PSA: group dis-
cussion, meetings, and interpersonal communication. Hypothesis 7 predicts that the inter-
vention will decrease PSA. Specifically, we predict that the SDT- and MI-based intervention
will:
H4: Result in no pre- to post-intervention changes in OCA in the domain of meetings.
H5: Result in no pre- to post-intervention changes in OCA in the domain of groups.
H6: Result in no pre- to post-intervention changes in OCA in the domain of interper-
sonal communication.
H7: Result in a decrease from pre- to post-intervention in the domain of PSA.
Affect: Hypotheses 8 and 9
The PANAS assesses positive and negative affect (emotion); it has been validated and
extensively applied (Watson et al. 1988) including to PS (Mano 1991, 1992). As applied
in our study, the PANAS directly measures PS-related affect (cf., Mano 1991, 1992). In it,
participants rate feeling states along two dimensions, using positive (dimension 1) and
negative (dimension 2) emotion-laden adjectives (see Appendix A, Panel 4, for instrument).
Participants high in positive affect have high energy, strong concentration, and pleasurable
engagement, whereas low positive affect is characterized by sadness and lethargy. Alter-
natively, high negative affect indicates anger, contempt, disgust, or fear, whereas low neg-
ative affect is characterized by calmness and serenity (Watson et al. 1988). We predict that
the SDT- and MI-based intervention will:
274 Miller and Stone
Week 2 emphasized students’ choice (i.e., autonomy) to work on presentation skills or
not, and solicited their thoughts and feelings about this work (Miller and Rollnick 2002;
Reeve 2002; Reeve et al. 2004). During this week, students completed an online instrument
that asked whether they were committed to improving their presentation skills and wanted
to read material related to improving their presentation skills. Participants who commit-
ted to work on these skills (n 56 across all studies) had higher pre-intervention positive
SS (p .000) and motivation (p .009) than participants who did not (n 15).^5 The
instrument also asked students whether points found in the readings were consistent with
their own experience and whether and how the readings were useful in improving their PS
skills.
Week 3. Students received a personalized email from the instructor that summarized
the students’ thoughts and feelings about PS that were expressed in the Week 2 activity.
This exercise is aimed at facilitating dialogue, acknowledging and supporting feelings, and
increasing awareness of PS-related SS (Miller and Rollnick 2002).
Weeks 4 and 5. Students presented during Week 4. The intervention focused on sup-
porting feelings of competence (Reeve 2002; Reeve et al. 2004) related to these presenta-
tions. During Week 5, participants who committed to improving their presentation skills
received individualized feedback (by email) on 25 dimensions of PS skill (see Appen-
dix D).
Week 8. Participants met with the instructor, in groups, regarding the semester projects.
For students who committed to improving their presentation skills, these meetings included
a discussion of oral presentations and a brief, spontaneous presentation by each student
participant. Following each presentation, the instructor offered observations and supportive
comments based on MI principles (Miller and Rollnick 2002).
Week 16. Participants presented their semester projects. Post-intervention measures were
collected after the presentation but prior to receiving a presentation grade and summary
evaluation.
Study 2 Intervention
We modified two aspects of the Study 1 intervention for Study 2 (see Appendix E for
additional descriptions).
Week 3. Writing and distributing the Week 3 email required 20 minutes per student.
To reduce the instructor time required by the intervention and to increase its feasibility in
larger-section classes, this aspect of the intervention was dropped.
Week 4. Consistent with suggestions in RH, the instructor obtained and used video
recording technology for the Week 4 student presentations. The instructor reviewed selected
portions of the videotape with students in Week 5, using nonjudgmental feedback principles
from MI training (Miller and Rollnick 2002). This intervention required less instructor time
than did the personalized email feedback completed in Study 1. Hence, the net instructor
effort required by the intervention in Study 2 was lower than that of Study 1.
Study 3 Intervention
Study 3 replicated Study 2, but eliminated the Week 4 videotaping. Hence, Study 3
required the least instructor effort of the interventions.
(^5) Similar results obtain for students who choose to read PS-related material.
Public Speaking Apprehension (PSA), Motivation, and Affect among Accounting Majors 275
Measure Timing and Reliability
Measures were assessed pre- and post-intervention. The three common measures (see
Appendix A, Panel 1) and the PRCA-24 (see Appendix A, Panel 3) were assessed, at pre-
intervention, during the first week. Because they measure PS affect and PS-related
self-perceptions, the ‘‘pre-intervention’’ PS PANAS (see Appendix A, Panel 4) and PS self-
perception sub-domain (SPSD: see Appendix A, Panel 2) measures were assessed in Week
4 immediately following the participants’ first presentations. Hence, our chances of finding
effects on the PANAS and SPSD measures are lessened since the ‘‘pre-test’’ assessments
occur in Week 4 of the intervention. Post-intervention measures were assessed after the
final presentation but before students received a presentation grade and evaluation.
Appendix C presents construct reliability assessments for the Studies 1, 2, and 3, and
the aggregated data set. Reliability was adequate to good, with the exception of four con-
struct measures in specific studies; three Study 2 measures: (1) negative self-statements, (2)
PRCA-24 measure of meetings, and (3) content domain self-assessment, and one Study 3
SPSD measure, visual aids. Given the smaller sample size and corresponding higher Beta
error likelihood in Study 2, it is unsurprising that construct validity is lower in Study 2
than the other studies.
Ability
As an experimental control, we obtained data from the university registrar on partici-
pants’ ability as measured by overall undergraduate grade point average (GPA), undergrad-
uate accounting GPA, verbal/quantitative GMAT score, and GMAT analytical writing score
(see Appendix A, Panel 5). Sample sizes for the ability measures were as follows: overall
undergraduate GPA (n 70), undergraduate accounting GPA (n 34), GMAT scores (n
70), and analytical writing (n 28). An ANOVA to test for between-study differences
indicates no difference in undergraduate accounting GPA, verbal/quantitative GMAT score,
or GMAT analytical writing score. However, we do find a difference in overall undergrad-
uate GPA. Post hoc analyses (Bonferroni correction) indicates that the overall undergraduate
GPA is lower among Study 1, compared with Study 3, participants (Study 1 mean 3.41,
Study 3 mean 3.66; p 0.003).
Study 3: PS Sub-Domain (SD) Skill Scale Difference
PS SD skills were measured on a 1–5 scale (1 poor, 5 good) in Study 2, and at
post-intervention in Study 3 (see Appendix A, Panel 2). However, due to a programmer
error, the PS SD skills were measured on a 1–7 scale (1 very poor, 7 very good) at
pre-intervention in Study 3. We followed suggestions in the scale development literature
(Dawes 2002, 2008) to convert the seven-point scale used pre-intervention in Study 3 to
the five-point scale used in the other cases. Specifically, we converted the seven- to a five-
point scale with the following equation: ((2/3) *((seven-point scale response) 4) 3).^6
RESULTS
We present (1) correlational analysis, (2) tests of hypotheses and related analyses, and
(3) benchmark comparisons against prior studies. Tests of hypotheses were assessed using
repeated-measure (i.e., intervention, which compares pre- with post-intervention) ANOVA
(^6) With this formula, the mapping of seven- to five-point scale values are [7, 5; 6, 4.3; 5, 3.7; 4, 3; 3, 2.3; 2, 1.7;
1, 1].
Public Speaking Apprehension (PSA), Motivation, and Affect among Accounting Majors 277
TABLE 2
Pre- to Post-intervention Correlations
Panel A: PS Positive and Negative Self-Statements and Motivation (n
71 to 73)
PS MotivationPre-Intervention
PS Motivation Post-Intervention
Positive Self- Statements Pre-
Intervention
Positive Self- Statements Post-
Intervention
Negative Self-Statements Pre-Intervention
Negative Self- Statements Post-
Intervention
PS Motivation Pre-
Intervention
PS Motivation Post-
Intervention
Positive Self-Statements
Pre-Intervention
Positive Self-Statements
Post-Intervention
Negative Self-
Statements Pre-Intervention
Negative Self-
Statements Post-Intervention
*, **, *** Significant at p
.10, p
.05, and p
.01, respectively (two-tailed).
Pearson correlations below diagonal, Spearman above diagonal.
( continued on next page
278 Miller and Stone
TABLE 2 (continued)
Panel B: PRCA-24 Correlations (n
Group Pre-Intervention
Group Post-Intervention
Meeting Pre-Intervention
Meeting Post-Intervention
Interpersonal
Pre-
Intervention
Interpersonal
Post- Intervention
PS Pre- Intervention
PS Post- Intervention
Group Pre-Intervention
Group Post-intervention
Meeting Pre-
Intervention
Meeting Post-
Intervention
Interpersonal Pre-
Intervention
Interpersonal Post-
Intervention
PS Pre-Intervention
PS Post-Intervention
*, **, *** p
.10,
.05, and
.01, respectively, Pearson correlations, two-tailed test significance.
Pearson correlations below diagonal, Spearman above diagonal. Panel C: Correlations of PRCA-24 with PS Positive and Negative SS and Motivation (n
Pre-Intervention
Post-Intervention
PS
Motivation
Positive Self- Statements
Negative Self-
Statements
PS
Motivation
Positive Self- Statements
Negative Self-
Statements
Group Pre-Intervention
Meeting Pre-Intervention
Interpersonal Pre-Intervention
PS Pre-Intervention
Group Post-Intervention
Meeting Post-Intervention
Interpersonal Post-Intervention
PS Post-Intervention
*, **, *** p
.10,
.05, and
.01, respectively, Pearson correlations, two-tailed test significance.
( continued on next page
280 Miller and Stone
TABLE 2 (continued)
Panel E: Correlations among PS SPSD and Other Variables—Initial Presentation versus Post-Intervention or Final Presentation Dependent
Variables (n
Initial Presentation
Post-Intervention orFinal Presentation
Appearance
Audience
Structure
Content
Visual Aids
Slides
PositiveAffect
Negative
Affect
PRCA—Group
PRCA—Meetings
PRCA—Interpersonal
PRCA—PS
PS Motivation
Positive SS
Negative SS
Appearance
Audience
Structure
Content
Visual Aids
Slides
Positive Affect
Negative Affect
Pearson correlations; *, **, *** significant at p
.10, p
.05, and p
.01, respectively (two-tailed).
Public Speaking Apprehension (PSA), Motivation, and Affect among Accounting Majors 281
TABLE 3
Pre- to Post-Intervention Differencesa
Panel A: PS Motivation and Positive and Negative Self-Statements (n 73)
Measure Timing^ Mean^ Std. Dev.
Repeated Measures ANOVA (F, p values) Time F(2,69)
Study F(2,69)
Time * Study F(2,69)
PS Motivation (H1) Pre-test 11.808 3.554 17.854 0.070 0. Post-test 13.822 2.756 (0.000) (0.933) (0.815)
Positive Self- Pre-test 16.164 3.598 8.274 1.016 0. Statements (H2) Post-test 17.452 3.395 (0.005) (0.367) (0.686)
Negative Self- Pre-test 5.712 4.427 5.155 0.442 0. Statements (H3) Post-test .297 4.001 (0.026) (0.644) (0.675)
Panel B: PRCA-24 Measures (n 50)
Measure Timing^ Mean^ Std. Dev.
Repeated Measures ANOVA (F, p values) Time F(1,48)
Study F(1,48)
Time * Study F(1,48)
Meeting (H4) Pre-test 16.08 4.831 1.638 0.728 0. Post-test 15.16 4.152 (0.207) (0.398) (0.607)
Group (H5) Pre-test 14.36 3.853 1.044 0.315 0. Post-test 13.68 3.793 (0.312) (0.577) (0.969)
Interpersonal Pre-test 14.76 3.868 2.564 0.054 0. Communication (H6) Post-test 13.66 4.064 (0.116) (0.818) (0.911)
PS (H7) Pre-test 19.84 4.648 12.718 5.224 1. Post-test 17.68 4.714 (0.001) (0.027) (0.271)
Panel C: PANAS Measures (n 50)
Measure Timingb^ Mean^ Std. Dev.
Repeated Measures ANOVA (F, p values) Time F(1,48)
Study F(1,48)
Time * Study F(1,48)
Positive Affect (H8) IP 28.54 6.961 6.761 0.000 0. FP 31.66 8.211 (0.012) (0.984) (0.593)
Negative Affect (H9) IP 18.94 7.427 0.001 21.855 2. FP 18.24 5.854 (0.979) (0.000) (0.153)
Panel D: PS Self-Perception Sub-Domain (SPSD) Measures (n 50)
Measure Timingb^ Mean^ Std. Dev.
Repeated Measures ANOVA (F, p valuesb) Time F(1,48)
Study F(1,48)
Time * Study F(1,48)
Appearance (H10) IP 24.287 5.877 18.158 4.795 5. FP 29.62 5.558 (0.000) (0.033) (0.023)
Audience (H11) IP 23.913 4.790 25.516 0.743 1. FP 30.04 5.700 (0.000) 0.393 0.
( continued on next page )
Public Speaking Apprehension (PSA), Motivation, and Affect among Accounting Majors 283
TABLE 4
Hofmann and DiBartolo (HD) Benchmarking Results for Negative Self Statements
Panel A: HD Benchmark Negative and Positive Self Statement Means (n 301)
Mean Std. Dev.
Negative SS 7.366 5.
Positive SS 15.667 4.
Panel B: Results of our Studies 1, 2, and 3 versus HD 1 and 2a
Study (n) Mean Std. Dev. df t-tests p
Positive Self-Statements 1 (n 23) 15.739 3.671 322 0.072 0. Pre-Intervention 2 (n 14) 17.214 3.423 313 1.216 0. 3 (n 36) 16.028 3.637 335 0.445 0. All (n 73) 16.164 3.598 372 0.845 0.
Positive Self-Statements 1 (n 23) 17.565 3.287 322 1.901 0. Post-Intervention 2 (n 14) 18.429 2.738 313 2.180 0. 3 (n 36) 17.000 3.680 335 1.642 0. All (n 73) 17.452 3.395 372 3.057 0.
Negative Self-Statements 1 (n 23) 6.043 4.269 322 1.176 0. Pre-Intervention 2 (n 14) 5.000 3.211 313 1.665 0. 3 (n 36) 5.778 4.975 335 1.719 0. All (n 73) 5.712 4.427 372 2.478 0.
Negative Self-Statements 1 (n 23) 3.957 4.772 322 3.011 0. Post-Intervention 2 (n 14) 3.643 2.977 313 2.623 0. 3 (n 36) 4.889 3.838 335 2.735 0. All (n 73) 4.356 4.001 372 2.735 0.
a (^) Significant results shown in bold.
samples of undergraduate college students; we compare weighted average composites
(weighted by sample sizes) of the HD Study 1 and 2 data to our results.
Table 4, Panel A, presents means and standard deviations of the HD participants’ levels
of negative and positive SS. Table 4, Panel B, presents the means and standard deviations
for our three studies individually and aggregated; it also reports t-test results comparing
our and HD’s SS data.
Pre-intervention, our participants’ levels of positive SS, individually and in the aggre-
gate, do not differ from HD’s (see Table 4, Panel B). At post-intervention, our participants’
levels of positive SS are higher in Study 2 and in aggregate, and are marginally higher in
Study 1, than are HD’s (see Table 4, Panel B). Individually by study and combined, our
participants’ levels of negative SS are lower at post-intervention than are HD’s. In addition,
our combined participants’ levels of negative SS are lower at pre-intervention than are HD’s;
this result does not hold when comparing levels of negative SS for our individual studies
with HD’s. These results provide some support for the assertion that the intervention reduces
graduate accounting majors’ negative SS and improves graduate accounting majors’ positive
SS in relation to the HD data.
PRCA-24 Comparisons: National Norms and RH
We also compared our PRCA-24 results against two data sets: national norms for U.S.
undergraduate students (McCroskey 1982) and the pre- and post-intervention data from
284 Miller and Stone
RH’s (1994) study of OCA among accountancy students. We expect to find no difference
between national norms and our pre-intervention data; we expect lower levels of PSA in
our post-intervention data than in the national undergraduate norms, but no difference
in PSA in our and RH’s post-intervention data.
The results are generally consistent with these expectations. Our pre-intervention scores
are statistically equivalent to national norms and RH’s participants at pre-intervention (re-
sults not shown; p 0.133). In addition, our participants’ post-intervention PSA scores are
lower than national U.S. undergraduate norms (McCroskey 1982) (t(12,466) 3.072, p
0.002). However, contrary to our expectations, our post-intervention PSA scores are
higher than are RH’s post-intervention scores (t(70) 2.25, p 0.027).
To understand why our post-intervention PSA scores are higher than are RH’s
post-intervention scores, we separately compared our Study 2 and Study 3 pre- and post-
intervention PSA scores with those of RH. The pre- (t(34) 0.207, p 0.796) and
post-(t(34) 0.199, p 0.844) intervention PSA scores of our Study 2 and RH’s partic-
ipants do not differ. However, our Study 3 participants have marginally higher pre-
intervention PSA scores (t(56) 1.815, p 0.075) and higher post-intervention PSA scores
(t(56) 2.837, p 0.006) than do RH’s participants. Hence, differences in our and RH’s
results occur in Study 3 but not Study 2. The post-intervention differences in our Study 3
and RH’s study may obtain because of either the: (1) initially marginally higher PSA levels
among our Study 3 participants, or (2) because the reduced intervention activity in Study
3 may have reduced intervention effectiveness.
PRCA-24 Comparisons: High OCA Base Rates
McCroskey suggests that students who are one standard deviation or above national
norms on total PRCA-24 score are ‘‘high’’ OCA. Using this benchmark, RH and SL report
approximately 7.5 percent and 19 percent high OCA participants, respectively. In our sam-
ples, six students (6/50 or 12 percent) are high OCA at pre-intervention, while three stu-
dents (3/50 or 6 percent) have high OCA at post-intervention. The SL sample is of intro-
ductory accounting classes, while the RH and our sample are of upper division and graduate
accounting students, respectively.
Do Changes in PS Motivation, and Positive and Negative SS, Correlate with Ability?
RH found that students with higher standardized test scores, i.e., students of higher
ability, benefited less from their intervention than did students with lower standardized test
scores. This suggests that the effectiveness of the RH intervention was greater for students
of lower rather than higher ability. We tested whether changes in the PS measures (PS
motivation; positive and negative SS) that are common in Studies 1, 2, and 3 correlated
with ability (see Table 5). Specifically, we correlated changes in these measures with four
measures of ability: (1) overall undergraduate GPA, (2) undergraduate accounting GPA, (3)
GMAT total (verbal and quantitative) score, and (4) GMAT analytical writing score.
Only one of the 12 correlations is significant: a negative correlation of GMAT total
score with the change in negative SS; this result means that students of higher ability (with
higher GMAT scores) evidence greater intervention benefits in negative SS compared
with students of lower ability. This result is the opposite of RH’s. We offer three speculative
explanations for these differing results: (1) sampling (random) error in one or all of our
studies or RH’s study, (2) the differing interventions in the present versus RH studies lead
to an opposite relationship between intervention efficacy and student ability, and (3) sys-
tematic, though unidentified, sample differences in our and RH’s studies. These speculations
await future investigation.
