Performance visualization and behavioral disruption: a clarification.

Current data suggest that performance visualization is an effective way to reduce communication apprehension (Ayres, Hopf, & Ayres 1997), but performance visualization has not been as effective in reducing behavioral disruption as might be expected (Ayres & Hopf, 1992; Ayres et al. 1995).

One reason for this seems to be the way the rigidity and agitation (Mulac & Sherman, 1973) have been treated in these investigations. In effect, researchers exposed people to performance visualization regardless of whether they exhibited rigid or agitated behavior and then reported whether changes occurred in both of these variables. That procedure is questionable because any person who exhibits rigidity will by definition not exhibit agitation. In order to examine the impact of this confounding factor in work on performance visualization, two studies were undertaken. The first study focused on rigidity and the second on agitation. These studies suggest that performance visualization is more effective in reducing agitation and rigidity than previous research suggests.

Keywords: Communication Apprehension; Performance Visualization; Rigidity; Agitation

Introduction

Communication apprehension (CA) has been found to be a problem for many people (Daly, Caughlin, & Stafford, 1997). Accordingly a considerable amount of work has been devoted to finding ways to reduce CA (Allen, Hunter, & Donohue, 1989).

It appears that CA can be reduced by upgrading skills (Kelly, 1997), changing cognitions (Wilcox, 1997), getting people to relax (Friedrich, Goss, Cunconan, & Lane, 1997), and/or altering the way one envisions oneself as a speaker (Ayres, Hopf, & Ayres, 1997). A meta-analysis (Allen, Hunter, & Donohue, 1989) suggests that all of these approaches are of consequence in reducing fear associated with public speaking. However, our interests align with the way one envisions him/herself as a speaker (i.e., visualization) and the impact of that on certain aspects of public speaking apprehension.

Visualization (Ayres & Hopf, 1985) has become a popular way to reduce CA (Robinson, 1997) because it is effective and easy to employ. Visualization has been demonstrated to reduce communication apprehension (Ayres & Hopf, 1985, 1990) and compares favorably with other interventions (Ayres & Hopf, 1987). This form of visualization involves closing your eyes, relaxing, and listening to a script that describes a successful speaking experience. One difficulty that emerged with this form of visualization was that it reduced CA but did not seem to impact behavior (Kuruvilla, 1989).

In order to correct this difficulty, Ayres and Hopf (1992) developed what they called performance visualization. Performance visualization has been demonstrated to reduce CA and behavioral disruption (Ayres & Hopf, 1992). Performance visualization for a speaker involves watching a videotape of a proficient speaker, making a mental movie of the videotape, and replacing the image of the speaker on the tape with a vivid image of oneself as the speaker.

Even though performance visualization was developed to impact behavior, its impact on behavioral disruptions displayed by apprehensive speakers is modest. For instance, one study reports the procedure to impact agitation, but not rigidity (Ayres & Hopf, 1992). Another study found effects on rigidity but not on agitation (Ayres et al. 1995). Yet another study found effects on both variables (Ayres & Sonandre, 2003). Thus, performance visualization appears to impact behavior but not in a consistent fashion.

On close inspection, the way these authors examined the rigidity and agitation aspects of behavioral disruption may be a contributing factor to these inconsistent findings. In essence, these researchers pretested for all four aspects of behavioral disruption, manipulated performance visualization and compared changes across conditions. It appears that this approach may mask changes in agitation and rigidity because agitation refers to not standing still, using nervous, repetitive gestures and so forth, while rigidity involves being tense, not moving, not using gestures, and so on. The two things are mutually exclusive. One cannot appear to be rigid and agitated at the same time. When such behaviors are recorded, an agitated person will receive a low score on rigidity and vice versa for a rigid person. Furthermore, if a person receives a high score on rigidity (or agitation) and a low score on agitation (or rigidity) and the scores are averaged the person will appear to display a moderate level of behavioral disruption. Thus, it makes sense to examine changes in rigidity or agitation for a given individual but not both. Previous research on performance visualization reports changes in both for the same person. This approach guarantees, there will be minimal change on either agitation or rigidity (depending upon which a person displays).

Thus, the purpose of the studies reported here was to determine if performance visualization reduces agitation and rigidity respectively when one or the other (not both) of those behaviors are targeted.

Methods

The following steps were taken to identify participants for the studies reported below.

Participants

Participants were drawn from the 2214 students enrolled in a public speaking course over two semesters. These students filled out the public speaking sub-scale of the PRCA (Levine & McCroskey, 1990) during the first week of class. Several student assistants attended these classes during the first week of class and rated students on the rigidity and agitation sub-scales of Mulac and Sherman's (1973) Behavioral Assessment of Speech Anxiety during their speeches of introduction which were all delivered during the first week of class. Students who scored one standard deviation above the mean on the PRCA and who were rated as being rigid or agitated by the experimental assistant (who was unaware of the nature of this study) were eligible to participate in this study. Students with these characteristics were contacted using a random selection procedure. Data collection continued until 204 students had participated in the study.

Students were given extra-credit for participation in this study. All students in the course were able to earn the same number of extra-credit points in a variety of other ways. Of these students 81% were freshmen; 102 were male, 102 were female. Fifty percent of these students participated in Study One and the other 50% participated in Study Two. Students in the control/placebo condition were given the opportunity to participate in a speech anxiety reduction workshop at the conclusion of the study.

Study One: Rigidity

Data Gathering

Study participants, who scored one standard deviation above the mean on the public speaking sub-scale of the PRCA and who were rated as being rigid during their speeches of introduction, delivered a pretest speech to an experimental assistant, two coders, and the other participants (two to four people). The topic of the speech was 'What I expect to get out of college' or 'What I expect to do in the future'. Half of the participants spoke on one topic for the pretest and the other topic for the post-test. Following these speeches, participants completed Spielberger, Gorsuch, and Lushene's (1970) state communication apprehension measure and were rated on rigidity by the coders (blind as to the study purpose).

Immediately following the collection of the pretest data, study participants, who were randomly assigned to treatment conditions, were exposed to performance visualization (Ayres & Sonandre, 2003). This form of performance visualization featured an outstanding speech by a male student or a female student (women were shown a speech by a woman and men a speech by a man), a placebo or a control condition (the nature of these conditions is described in the subsequent section of this report).

Following treatment, participants delivered a post-test speech, completed the state CA measure and the public speaking sub-scale of the PRCA (Levine & McCroskey, 1990). They were also assessed for rigidity by two coders.

Treatments

Participants in the performance visualization treatment condition watched a videotape of a proficient male or female student speaker, made a mental movie of that speaker, and replaced the speaker on the tape with a vivid image of him or herself as the speaker. Performance visualization takes about 35 minutes to administer.

Participants in the placebo condition read material on general communication processes for a period of 35 minutes. Reading such material does not appear to affect public speaking apprehension (Ayres, Heuett, & Sonandre, 1998).

Participants in the control condition were left to their own devices for 35 minutes while the experimenter excused him or herself 'to get additional forms'.

Instruments

The PRCA (Levine & McCroskey, 1990) has been demonstrated to be reliable and valid way to measure trait CA in numerous studies (McCroskey, 1997). Alpha reliability of the public speaking sub-scale in this study was determined to be .88 and .92 in the pre and post-test applications.

The Spielberger et al. (1970) instrument has been used in a variety of studies to assess state CA (e.g., Beatty, Dobos, Balfantz, & Kuwabara, 1991). The alpha reliability of this scale was found to be .88 and .89 in the pre and post-tests in this study.

The rigidity sub-scale of the BASA (Behavioral Assessment of Speech Anxiety) developed by Mulac and Sherman (1974) was used to code rigidity in this study. The rigidity sub-scale includes three items (rigid or tense; motionless or lack of gestures; tense face muscles, grimaces, twitches) rated from 0 to 9. This instrument has been used successfully in a number of investigations (e.g., Jaremko, 1980; Trussell, 1978).

Coders/Coder Training

For this study, experimental assistants (advanced undergraduate students), blind as to the study purpose, were trained to use the rigidity sub-scale of the BASA. Coders coded sample speeches and discussed differences in their judgments. Coder training continued until inter-coder reliability averaged .80 or higher. During the study, two coders were assigned to rate each speech using the rigidity sub-scale of the BASA. These coders were excused during treatment administration to keep them blind as to the study's purpose. The coders returned when the post-test speeches were delivered. Inter-coder reliability was .93 and .94 for the pre- and post-test speeches respectively. Coders' scores were averaged for use in subsequent analyses.

Design and Analyses

A pretest/post-test control group design was employed in this study. The experimental manipulation concerned whether participants were exposed to performance visualization, placebo or control conditions. A pretest indicated NSD on gender so that variable was excluded from subsequent analyses. Scores on the public speaking sub-scale of the PRCA, state CA, and rigidity sub-scale of the BASA served as the dependent variables in this investigation. Data were analyzed using a one-way multiple analysis of covariance to determine if significant effects existed. Significant effects were pursued via ANCOVA and post hoc tests. The pretests scores served as the covariate in these analyses.

Results

The multiple analysis of covariance applied to these data proved to be significant F(3, 94) = 18.99, p < .001. Accordingly, ANCOVAs were applied to each dependent variable to determine which dependent variables were responsible for this effect. The ANCOVA applied to the state and trait CA data results proved to be statistically significant F(1, 96) = 19.02, p < .001, [[omega].sup.2] =.34; F( 1, 96) = 11.02, p <.001, [[omega].sup.2] = .28, respectively. As is to be expected, post hoc tests (Duncan's) indicated that people exposed to performance visualization reported lower state and trait CA than people in the other conditions. A main effect difference emerged for rigidity F (1, 96) = 23.94, p < .001, [[omega].sup.2] = .39. Subsequent post hoc tests (Duncan's) indicated participants exposed to performance visualization displayed less rigidity than those in the placebo and control conditions. Respondents in the placebo and control groups did not differ from one another. Table 1 presents pre and post-test means for all measures. Of particular note is that the pre/post difference scores for rigidity are 4.44 which is substantially greater than the 3.0 difference reported by Ayres and Sonandre (2003), the .50 difference reported by Ayres and Hopf (1992), or the .70 difference reported by Ayres et al. (1995).

Study Two: Agitation

The procedures, experimental design, and data analyses followed in this study were identical to the procedures followed in Study One except agitation was the focus of this study. Agitation, as operationalized by Mulac and Sherman (1974) includes three distinct behaviors (fidgeting, sways, paces, shuffles feet, lack of eye contact, extraneous eye movements). Inter-coder reliability for the pretest was .87 and .94 for the post-test. Similarly, alpha reliability for trait CA was .93 and .94 for the pre and post-tests respectively. Reliability of the state CA measure was .91 and .93 for the pre and post-tests respectively.

Results

The multiple analysis of covariance applied to these data proved to be significant F(3,94) = 24.23, p < .001. Accordingly, ANCOVAs were applied to each dependent variable to determine which dependent variables were responsible for this effect. The ANCOVA applied to the state and trait CA data results proved to be statistically significant F(1, 91) = 21.12, p < .001, [[omega].sup.2] =.33; F(1, 96) = 22.36, p<.001, [[omega].sup.2] = .35, respectively. As is to be expected, post hoc tests (Duncan's) indicated that people exposed to performance visualization reported lower state and trait CA than people in the other conditions. A main effect difference also emerged for agitation F(1, 96) = 19.99, p < .01, [[omega].sup.2] = .31. Subsequent post hoc tests (Duncan's) indicated participants exposed to performance visualization displayed less agitation than those in the placebo and control conditions. Respondents in the placebo and control groups did not differ from one another. Table 1 presents pre and post-test means for all measures. Of particular note, is that the pre/post difference scores for agitation are 4.98 which is substantially greater than the 3.0 difference reported by Ayres and Sonandre (2003), or the .20 difference reported by Ayres and Hopf (1992).

Discussion

As expected people exposed to performance visualization reported lower CA than those exposed to placebo or control conditions. Performance visualization was also linked to substantial reductions in rigidity and agitation in these two investigations. The mean differences in this study were 4.4 and 4.6 respectively for rigidity and agitation which compares to the .20 to 3.0 mean differences reported in related research. It appears that treating these mutually exclusive variables independently is of consequence. This finding suggests that performance visualization may be more powerful and consistent than previous research indicates. Previous work tended to ignore the fact that agitation and rigidity are mutually exclusive (if you are rigid you will not be agitated and vice versa). Operationally, performance visualization will fail to impact rigidity for the agitated person because the agitated person cannot be rigid. In turn, performance visualization will fail to impact agitation for the rigid person because he or she is not agitated. The findings in this investigation suggest that performance visualization is quite effective in reducing rigidity for those who display rigid behavior and in reducing agitation for those who display agitated behavior.

Given these findings, it may prove useful to test ways to improve performance visualization by adjusting it to target specific behaviors/thoughts. In the athletic arena, for instance, a person having a problem making free throws in basketball does not envision a jump shot. He or she is exposed to a person who is an excellent free throw shooter and patterns his or her images after that person (Bandura, 1997). It might be useful to test such things in the speech arena. For instance, a rigid speaker exposed to performance visualization featuring a particularly fluid speaker may find it more useful than the general form of visualization employed in this investigation.

Research in the physical education arena also indicates that the most effective model is often the person him or herself (Gonzales & Dowrick, 1982; Scraba, 1990). This is accomplished by taping the person doing something several times, editing the tape to generate a good model of the activity, and exposing the person to the edited tape. It might be profitable to examine whether a similar approach would help speakers (e.g., editing out rigid behaviors for instance).

Related literature (Gray & McNaughton, 2000) on why people display fight, flight, or freeze responses when confronted with a threatening situation suggests an interesting parallel with the rigidity (freeze?) and agitation (flight? fight?) dimensions of Mulac and Sherman's (1974) behavioral disruption scale. If rigidity and agitation are reasonable indicators of flight, fight and freeze behaviors in humans, they provide a means by which one may be able to test Gray and McNaughton's model with public speakers. In the main, Gray and McNaughton argue that animals (including humans) constantly scan the environment. Things in the environment are compared with known repositories of information. When something falls into a threatening category, the comparator engages the behavioral inhibition system (i.e., fight, flight, freeze). By viewing CA from this perspective, one may be able to provide reasons why people exhibit agitated or rigid behaviors and why interventions reduce such behaviors. Of course, this connection is mere speculation at this point and will require considerable research to document.

Overall the results of this investigation support the following conclusions. Performance visualization reduces CA, rigidity and agitation more so than a placebo or no treatment. Unlike previous work, the results for rigidity and agitation in this study were consistent and large. It appears that treating these mutually exclusive variables independently is of consequence. Thus, unlike some previous work, these data suggest that performance visualization is an effective way to reduce rigidity and agitation for those who display high rigidity/agitation prior to exposure to visualization.

References

Allen, M., Hunter, J. E., & Donohue, W. A. (1989). Meta-analysis of self-report data on the effectiveness of public speaking anxiety treatment techniques. Communication Education, 38, 54-76.

Ayres, J., & Hopf, T. S. (1985). Visualization: A means of reducing speech anxiety. Communication Education, 34, 318-323.

Ayres, J., & Hopf, T. S. (1987). Visualization, systematic desensitization, and rational emotive therapy: A comparative evaluation. Communication Education, 36, 236-240.

Ayres, J., & Hopf, T. S. (1990). The long-term effect of visualization in the classroom. Communication Education, 39, 283-291.

Ayres, J., & Hopf, T. S. (1992). Visualization: Reducing speech anxiety and enhancing performance. Communication Reports, 5, 1-10.

Ayres, J., & Sonandre, D. M. (2003). Performance visualization: Does the nature of the speech model matter. Communication Research Reports, 20, 260-268.

Ayres, J., Heuett, B., & Sonandre, D. A. (1998). Testing a refinement in an intervention for communication apprehension. Communication Reports, 11, 73-86.

Ayres, J., Hopf, T., & Ayres, D. M. (1997). Visualization and performance visualization: Applications, evidence, and speculation. In J. A. Daly, J. C. McCroskey, J. Ayres, T. Hopf, & D. M.

Ayres (Eds.), Avoiding communication (2nd ed., pp. 401-402). Cresskill, NJ: Hampton Press.

Ayres, J., Ayres, D. M., Grudzinskas, G., Hopf, T., Kelly, E., & Wilcox, A. A. (1995). A component analysis of performance visualization. Communication Reports, 8, 185-192.

Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W. H. Freeman and Co.

Beatty, M. J., Dobos, J. A., Balfantz, G. L., & Kuwabara, A. Y. (1991). Communication apprehension, state anxiety, and behavioral disruption: A causal analysis. Communication Quarterly, 39, 48-57.

Daly, J. A., Caughlin, J. P., & Stafford, L. (1997). Correlates and consequences of social-communicative anxiety. In J. A. Daly, J. C. McCroskey, J. Ayres, T. Hopf, & D. M. Ayres (Eds.), Avoiding communication (2nd ed., pp. 21-74). Cresskill, NJ: Hampton Press.

Friedrich, G., Goss, B., Cunconan, T., & Lane, D. (1997). Systematic desensitization. In J. A. Daly, J. C. McCroskey, J. Ayres, T. Hopf, & D. M. Ayres (Eds.), Avoiding communication (2nd ed., pp. 305-329). Cresskill, NJ: Hampton.

Gonzales, F. P. & Dowrick, P. W. (1982, November). The mechanism of self-modeling: Skills acquisition versus raised self efficacy. Paper presented at 16th annual convention of the Association for Advancement of Behavior Therapy, Los Angeles.

Gray, J. A. & McNaughton, N. (2000). The neuropsychology of anxiety. Oxford: Oxford University Press.

Jaremko, M. E. (1980). The use of stress inoculation training in the reduction of public speaking anxiety. Journal of Clinical Psychology, 36, 735-738.

Kelly, L. (1997). Skills training as a treatment for communication problems. In J. A. Daly, J. C. McCroskey, J. Ayres, T. Hopf, & D. M. Ayres (Eds.), Avoiding communication (2nd ed., pp. 331-365). Cresskill, NJ: Hampton Press.

Kuruvilla, A. (1989). Does cognitive training influence anxiety behavior? Unpublished Master's thesis, Washington State University, Pullman, WA.

Levine, T. R., & McCroskey, J. C. (1990). Measuring trait communication apprehension: A test of rival measurement models of the PRCA-24. Communication Monographs, 57, 62-72.

McCroskey, J.C. (1997). Self-report measurement. In J. A. Daly, J. C. McCroskey, J. Ayres, T. Hopf & D. M. Ayres (Eds.), Avoiding communication (2nd ed., pp. 191-216). Cresskill, NJ: Hampton Press.

Mulac, A., & Sherman, A. R. (1974). Behavioral assessment of speech anxiety. Quarterly Journal of Speech, 60, 134-143.

Robinson II, T.E. (1997). Communication apprehension and the basic public speaking course: A national survey of in-class treatment techniques. Communication Education, 46, 188-197.

Scraba, P. J. (1990). Self-modeling for teaching swimming to persons with physical disabilities (Doctoral dissertation, University of Connecticut, 1990). Dissertation Abstracts International, 50, 2830A.

Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1970). Manual for the state-trait anxiety inventory. Palo Alto, CA: Consulting Psychologists Press.

Trussell, R. (1978). Use of graduated behavior rehearsal, feedback, and systematic desensitization for speech anxiety. Journal of Counseling Psychology, 25, 14-20.

Wilcox, A. K., (1997). Cognitive components of communication apprehension: What are they thinking? In J. A. Daly, J. C. McCroskey, J. Ayres, T. Hopf & D. M. Ayres (Eds.), Avoiding communication (2nd ed., pp. 367-378). Cresskill, NJ: Hampton Press.

Joe Ayres is a Professor Emeritus in the School of Communication at Washington State University, Pullman, WA 99164-2520, USA; E-mail: fal@moscow.com. Appreciation is extended to Frances Ayres and Tara Farley for their assistance with various aspects of this investigation.

Table 1 Pre and Post-tests Means and Standard Deviations Across
Three Dependent Variables in Two Studies

                   Study One

              Pretest        Post-test

              M      SD       M      SD

Rigidity
  Control   7.8     2.0     7.4     1.9
  Placebo   7.7     1.9     7.7     1.9
  PV        7.9     2.1     3.8     1.9

Trait CA
  Control   20.2    3.1     18.7    3.4
  Placebo   19.8    2.9     18.2    3.5
  PV        17.9    2.9     15.4    2.6

State CA
  Control   19.0    3.1     18.6    3.3
  Placebo   19.1    2.9     18.3    3.1
  PV        19.8    2.9     13.6    2.7

                   Study Two

Agitation
  Control   7.4     2.0     7.7     1.9
  Placebo   7.5     2.5     7.4     2.4
  PV        7.9     2.2     3.5     1.8

Trait CA
  Control   19.7    2.9     18.6    2.7
  Placebo   19.3    2.7     18.3    2.4
  PV        18.9    3.1     15.1    2.1

State CA
  Control   20.1    3.3     18.9    2.8
  Placebo   19.3    2.9     18.3    2.7
  PV        20.2    3.1     14.6    2.9

Note: PV = Performance Visualization using a great speech;
N = 33 in all cells.