International Symposium on Performance Science ISBN 978-2-9601378-0-4 The Author 2013, Published by the AEC All rights reserved The influence of performers stage entrance behavior on the audience s performance elaboration Friedrich Platz Hanover University of Music, Drama, and Media, Germany This study proposes a typology of the initial stage entrance behavior of performers based on audience members first impressions of selected video recordings at an international violin competition. Against the theoretical background of social interaction theory, it is assumed that performance evaluation can only be understood as an interaction between expectations of audience s sub-classes and observable behavior of groups of performers. An analysis of performer behavior was conducted in three steps: (1) based on methods of classical test theory and item response theory, a selection of six items that best describe performer behavior with regard to the audience s forming an impression was presented, (2) by means of a multi-level latent class analysis, responses could be described by one of three classes of the audience s first impression judgments ( appropriate, acceptable, and inappropriate stage entrance behavior) resulting in two groups of performers stage entrance behavior evaluation, (3) the association between audience first impression classes and the audience s motivation for performance continuation was used as an indicator for a more in-depth performance elaboration. Results suggest that for the adequate understanding of audiovisual performances, a model of performance elaboration can be an alternative to models of musical communication. Keywords: music performance; performance evaluation; multilevel latent class analysis; audience; music performance elaboration One way to look at music performance is by means of communication models (Juslin and Timmers 2010). With these widely distributed models, music performance can be summarized as an exchange of information between performer (as transmitter ) and audience (as receiver ) through the modifi-
346 WWW.PERFORMANCESCIENCE.ORG cation of physical space. This modification of space results from the performer s intentional operations for the successful realization of musical meanings (Davies 1994). Consequently, all auditive (Juslin 1997) as well as visual (Inge and Leman 2010) perceived modifications of the physical space within a performance were affiliated to the transmission of the compositional meaning. However, a recent meta-analysis (Platz and Kopiez 2012) revealed a medium effect size for an enhancement effect of the visual component, and the authors argue that the audience s music appreciation might not be sufficiently explained by music communication models in which the visual component s function has been marginalized to a supporting role. An alternative approach could be derived from social interaction theory (Goffman 1959), which explains the audience s music performance appreciation in terms of an impression formation embedded in a persuasion process (Knape 2003, Petty and Cacioppo 1986). Within this framework, all operations of both the performer and the audience result from their individual aims to regulate their objectives. According to Goffman (1959, p. 123), the audience s first impression at a concert event is made when a performer successfully takes on the front-stage mode by the putting on and taking off of character while entering the stage. Thus an impression type could therefore be defined as an audience member s preexisting mental-person schema (Fiske et al. 1999) of a performer s impression management (Goffman 1959). Additionally, every mental personschema is connected with preexisting evaluative judgments, such as emotional judgments or feelings. Depending on the audience member s success at allocating a performer to a unique category of person-perception, impression types could therefore have a different motivation on the audience s performance elaboration. Specifically, music performance elaboration is defined in this study as the extent to which a person deals with performance-relevant impressions and how long the audience wishes to continue paying attention to the performance. However, up until now, the importance of a performer s stage entrance behavior has been widely neglected in music performance research (Gabrielsson 1999), although McPherson and Schubert (2004, p. 71) mentioned a dominating influence of first impressions more as a result of an expectancy evaluation process on the part of the audience and less as the reflection of objective performer features. The aims of this current study are three-fold: first, to identify the best attributes that can describe the variety of performers initial stage entrance behavior; second, to develop a statistical model-based typology of stage entrance behavior that explains impression types of behavior as an interaction between audience sub-classes expectations and groups of performers be-
INTERNATIONAL SYMPOSIUM ON PERFORMANCE SCIENCE 347 havior; finally, to test the relationship between impression types and audience sub-classes motivation for ongoing performance elaboration. Participants METHOD Participants were recruited from a commercial online panel provider. In total, 1,530 participants took part in our online study. 528 subjects were excluded due to insufficient time on tasks, response bias, or stopping the video sequence too early. As a result, N=1002 subjects (50% females; mean age=38.27 years, SD=11.97 years) served as the sample in this study. Materials First, we created a corpus of video-recorded competitors at a German international violin competition (N=27; mean age=22.38 years, SD=3.52 years). After that, a new standardized video sequence was produced for the beginning of each competitor s first stage entrance. The end of a competitor s public (spoken) announcement was used as the video s starting point, whereas the end of the experimental video sequence was marked by the conclusion of the preparatory playing gesture. In the study, we presented twelve video sequences in which all musicians started their competition program with the same piece. Second, items were generated and validated in two pilot studies in terms of classical test theory as well as item-response theory to identify the preliminary attributes for the audience s impression of performer stage entrance behavior. Procedure Participants were first informed about the task of the online study, and then required to give their informed consent. Participants were then asked to give details about their socio-economic and demographic status, as well as their musical sophistication (Ollen 2006). After the one-time presentation of the video, participants rated their impressions of the musician s stage entrance behavior by means of ten dichotomous items resulting from the pilot studies. All video sequences were fully randomized between subjects. Finally, participants were asked to indicate whether they would like to continue with the musician s performance. The entire procedure took approximately 10 minutes.
348 WWW.PERFORMANCESCIENCE.ORG RESULTS The software Latent GOLD 4.5 (Vermunt and Magidson 2005) was used for multi-level latent class analysis (Vermunt 2003). According to Vermunt and Magidson (2004a, p. 549), a latent class model (LC model) can be defined as a statistical model in which parameters may differ between unobserved, socalled latent subgroups. Specifically, the response probability of an item represents such a parameter of a LC model, whereas the subgroups represent discrete categories of a latent variable (Vermunt and Magidson 2004b). Compared with conventional cluster analysis (Aldenderfer and Blashfield 1984), finite mixture modeling (McLachlan and Peel 2000) such as LC models offers more flexibility combined with lower degrees of freedom in the researcher s decision process, resulting in a more reliable and confident data analysis. As an example, only one parameter, namely the number of classes, has to be fixed a priori for an explorative LC model. The plausibility for the appropriate number of classes can be confirmed with regard to Bayesian information criteria (Burnham 2004) and further model fit indices (Vermunt and Magidson 2005). In contrast to a LC model, in a cluster analysis more than one parameter must be defined, and the problem of algorithm-dependent results remains unsolved. Further on, the main advantage of a LC model compared with cluster analysis is the statistically, and therefore theoretically oriented, decomposition process of the mixture distribution, depending on the items response format. To summarize, a LC model can be regarded as a top-down approach for statistical data modeling, while cluster analytical techniques represent a bottom-up approach. The data analysis consisted of five steps: (1) identifying the appropriate number of latent classes (level-one), (2) redeeming local independence assumption by removing four of the ten items with significant bivariate residuals (Magidson and Vermunt 2004), (3) obtaining the hierarchical structure of the data (individual ratings nested within musicians) by extending the preliminary model to a nonparametric random-coefficient LC model (level-two), (4) classifying audience sub-classes (level one) and groups of performers behavior (level-two), and finally (5) testing whether there was a significant, medium relationship between the audience s first impression types and the motivation for ongoing performance elaboration. The data analysis revealed that the observed response data could best be described by a nonparametric random-coefficient latent class model based on 6 items (see Table 1). The final model is comprised of three audience-related classes of appropriateness judgments for musician s stage entrance behavior (level-one).
INTERNATIONAL SYMPOSIUM ON PERFORMANCE SCIENCE 349 Table 1. The final 6-item questionnaire for the typology construction of the audience s first impression of musicians stage entrance behavior. No. Item I found it appropriate 1 how often this performer nodded at the audience 2 how often this performer changed his/her gaze 3 how often this performer touched her-/himself 4 which stance width the performer used for stage entrance 5 which step size the performer used for stage entrance 6 The musician s stage entrance behavior is resolute The response probabilities of the six items were high for the first class, which contained 45.17% of all participants. This class described the audience s impression of appropriateness, whereas the second class (44.10%) summarized all impressions of acceptance. The third class (10.73%) revealed all judgments based on perceived impressions of inappropriate stage entrances. Further on, all class sizes depended on two video groups representing the musicians stage entrances (level-two). The video groups differed from each other in their (level-one) class size proportions. In other words, stage entrances associated with Video Group 1 would be expected to be more appropriate (48.01%) in contrast to those associated with Video Group 2 (32.25%), which would primarily be judged as more acceptable (53.33%). Therefore, the first mixture component (Video Group 1) could be classified as musicians with successfully executed impression management in contrast to the musicians in the Video Group 2. Based on this classification, we observed a medium to strong relation between stage entrance judgment and a rater's motivation for the continuation of performance elaboration (X 2 (2)=36.64, p<0.01, V=0.46). DISCUSSION Against the background of music performance evaluation research, the results are in line with findings from previous studies that first impression types have an influence on music performance evaluation (McPherson and Schubert 2004). On the basis of the item selection for evaluative classification of the first impression of stage entrance behavior, a medium to strong effect size for the influence of first impression types on the motivation for performance elaboration continuation (V=0.46) was found. This was a quantification of an effect that appeared as an evaluation criterion for overall impression judg-
350 WWW.PERFORMANCESCIENCE.ORG ment (Stanley et al. 2002). Whereas first impression formation has been seen as an influencing factor at the beginning of a persuasive process (Ybarra 2001), up until now it has been discussed in music evaluation research only as a peripheral cue during the preparatory phase of a performance (McPherson and Schubert 2004). However, this study proposes a more psychologically oriented model of performance evaluation by assuming a general persuasion process (Petty and Cacioppo 1986). Furthermore, persuasion is based on music performance elaboration resulting in an involvement through optimal participation as the main purpose of experiencing music performance. It remains unanswered how sustainable the positively motivational effect of the appropriate stage entrance behavior type is for the whole performance elaboration. Additionally, the question of cross-validity between latent classes, as well as latent groups, and the winner of the international violin competition remained unanswered because the winner of the competition was not part of the video corpus. Future research in music performance evaluation should focus on music performance as a psychological function of impression formation perception. Acknowledgments I am indebted to Stephan Hintze (Hanover University of Music, Drama, and Media) for his expert comments on selected video sequences, to the anonymous reviewers for their helpful comments on a previous version of the manuscript, and to Maria Lehmann (Würzburg) for the final editing. For an extended version of this manuscript with more methodological detail and additional data analyses, see Platz and Kopiez (2013). Address for correspondence Friedrich Platz, Hanover University of Music, Drama, and Media, Germany, Emmichplatz 1, Hanover 30175, Germany; Email: friedrich.platz@hmtm-hannover.de References Aldenderfer M. S. and Blashfield R. K. (1984). Cluster Analysis. Newbury Park, California, USA: Sage Publications, Inc. Burnham K. P. (2004). Multimodel inference: Understanding AIC and BIC in model selection. Sociological Methods and Research, 33, pp. 261-304. Davies S. (1994). Musical Meaning and Expression. Ithaca, New York, USA: Cornell University Press.
INTERNATIONAL SYMPOSIUM ON PERFORMANCE SCIENCE 351 Fiske S. T., Lin M., and Neuberg S. L. (1999). The continuum model: Ten years later. In S. Chaiken and Y. Trope (eds.), Dual Process Theories in Social Psychology (pp. 231-254). New York: The Guilford Press. Gabrielsson A. (1999). Music performance. In D. Deutsch (ed.), The Psychology of Music (pp. 501-602). San Diego, California, USA: Academic Press. Goffman E. (1959). The Presentation of Self in Everyday Life. London: Penguin Books. Inge R. G. and Leman M. (2010). Musical Gestures. New York: Routledge. Juslin P. N. (1997). Emotional communication in music performance: A functionalist perspective and some data. Music Perception, 14, pp. 383-418. Juslin P. N. and Timmers R. (2010). Expression and communication of emotion in music performance. In P. N. Juslin and J. A. Sloboda (eds.), Music and Emotion (pp. 453-489). Oxford: Oxford University Press. Knape J. (2003). Persuasion. In G. Ueding and J. Walter (eds.), Historisches Wörterbuch der Rhetorik (pp. 1-34). Tübingen, Germany: Niemeyer. Magidson J. and Vermunt J. K. (2004). Latent class models. In D. Kaplan (ed.), The SAGE Handbook of Quantitative Methodology for the Social Sciences (pp. 175-198). London: Sage Publications Ltd. McLachlan G. and Peel D. (2000). Finite Mixture Models. New York: John Wiley & Sons, Inc. McPherson G. E. and Schubert E. (2004). Measuring performance enhancement in music. In A. Williamon (ed.), Musical Excellence (pp. 61-82). Oxford: Oxford University Press. Ollen J. E. (2006). A Criterion-related Validity Test of Selected Indicators of Musical Sophistication using Expert Ratings. Unpublished doctoral thesis, Ohio State University. Petty R. E. and Cacioppo J. T. (1986). The elaboration likelihood model of persuasion. Advances in Experimental Social Psychology, 19, pp. 123-205. Platz F. and Kopiez R. (2012). When the eye listens: A meta-analysis of how audio-visual presentation enhances the appreciation of music performance. Music Perception, 30, pp. 71-83. Platz F. and Kopiez R. (2013). When the first impression counts: Music performers, audience and the evaluation of stage entrance behavior, Musicae Scientiae, 17, pp. 167-197. Stanley M., Brooker R., and Gilbert R. (2002). Examiner perceptions of using criteria in music performance assessment. Research Studies in Music Education, 18, pp. 46-56. Vermunt J. K. (2003). Multilevel latent class models. Sociological Methodology, 33, pp. 213-239.
352 WWW.PERFORMANCESCIENCE.ORG Vermunt J. K. and Magidson J. (2004a). Latent class analysis. In M. S. Lewis-Beck, A. Bryman, and T. F. Liao (eds.), The Sage Encyclopedia of Social Sciences Research Methods (pp. 549-553). Thousand Oakes, California, USA: Sage Publications. Vermunt J. K. and Magidson J. (2004b). Latent variable. In M. S. Lewis-Beck, A. Bryman, and T. F. Liao (eds.), The Sage Encyclopedia of Social Sciences Research Methods (pp. 555-556). Thousand Oakes, California, USA: Sage Publications. Vermunt J. K. and Magidson J. (2005). Technical guide for Latent GOLD 4.0: Basic and Advanced. Belmont Massachusetts, USA: Statistical Innovations Inc. Ybarra O. (2001). When first impressions don't last: The role of isolation and adaptation processes in the revision of evaluative impressions. Social Cognition, 19, pp. 491-520.