Alma Mater Studiorum University of Bologna, August 22-26 2006
Understanding performance anxiety in the adolescent musician: Approaches to instrumental learning and performance Ioulia Papageorgi Institute of Education, University of London
[email protected] their amount of weekly practice and quantity of annual public performances they carried out.
ABSTRACT Research on musical performance to date has tended to focus on adult musicians. Although there is anecdotal evidence that problems can begin at a younger age, with adolescence being a particularly critical period, there has been no systematic research into the condition in younger performers. The results of a research study exploring how performance anxiety affects adolescent instrumental musicians are presented. A new self-reporting questionnaire was piloted and administered to 410 adolescent musicians aged 12-19 years old in the UK and Cyprus. Principal components factor analysis was conducted to explore student approaches to instrumental learning and performance and to investigate the potential influence of performance anxiety. The analysis revealed four components that represented both positive and negative approaches. Positive approaches were ‘evidence for achievement, positive musical identity and effort’ and ‘tendency for self-reliance and intrinsic motivation’. Negative approaches were ‘suggestion for susceptibility to maladaptive anxiety’ and ‘indication of unsuccessful coping with anxiety, lack of motivation and low self-confidence’. Differences in nationality, age, gender and examination undertaking were observed across the four extracted components. Each component contributed to a different extent in the prediction of students’ level of attainment (evidenced by their last examination result),
Keywords Performance anxiety, adolescence, instrumental learning, musical performance
INTRODUCTION Research on musical performance anxiety has flourished over the last two decades. Numerous clinics, journals and conferences have devoted their efforts to studying the psychological and medical aspects of musical performance (Valentine, 2002). There are two key reasons for this. First, the high incidence of maladaptive pre-performance anxiety in musicians and second, the recognition of the debilitating effects that anxiety can have on performance (Steptoe, 2001) and the well-being of the performer. Research to date has mainly focused on the diagnosis, assessment and treatment of maladaptive performance anxiety and its effects on selected groups of musicians. This has facilitated the development within a clinical setting of a range of coping strategies (for reviews of techniques which can be used to alleviate anxiety in musical performance see Wilson, 2002; Gabrielsson, 1999; and Salmon and Meyer, 1998). Research in these domains has set the foundations on which subsequent research may be based. However, such research has mainly focused on adult professional musicians, disregarding how anxiety might affect younger performers and how their early experiences might affect their behaviour towards anxiety in subsequent years. Whilst there is anecdotal evidence that problems can begin in earlier years, with adolescence being a particularly critical period, there is little research on this. Adolescence is a time when self-consciousness is often acute and this may prove
Proceedings of the 9th International Conference on Music Perception & Cognition (ICMPC9). ©2006 The Society for Music Perception & Cognition (SMPC) and European Society for the Cognitive Sciences of Music (ESCOM). Copyright of the content of an individual paper is held by the primary (first-named) author of that paper. All rights reserved. No paper from this proceedings may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information retrieval systems, without permission in writing from the paper's primary author. No other part of this proceedings may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information re-
ISBN 88-7395-155-4 © 2006 ICMPC
970
ICMPC9 Proceedings
to be a contributory factor in relation to both performance anxiety and its effects on performance. Given that adolescence is also the time when decisions regarding career paths are consolidated, performance anxiety may be instrumental in determining choices which will have very long term implications. The research presented here forms part of a larger research study aiming to explore how anxiety affects this neglected age group of musicians. The results presented in this paper concentrate on an exploration of approaches to instrumental learning and performance of adolescent skilled musicians. These approaches are evidenced by applying a data reduction technique to questionnaire data, namely principal components factor analysis. The questionnaire focused on a range of learning and performance issues, which, through the factor analysis, reveal students’ underlying conceptions regarding learning and performance in music and the influences of performance anxiety. Scores on the extracted components from the factor analysis are subsequently used in further analyses, to explore differences in nationality, gender, age and exam taking (t-tests) and to investigate the contributions of the extracted components to reported data on attainment, amount of weekly practice and number of annual public performances of the students (multiple regressions).
to strongly agree. Each statement belonged to an overarching theme or category that was related to an aspect of learning or performance. The themes/categories for the questions derived from an extensive literature review and from a pilot study conducted prior to the main data collection. Students’ responses to theme-related statements of the questionnaire were summed to form an overall score in that theme. Each theme constituted a variable used in the analyses presented below. The themes will be set out in the findings section below.
RESULTS Principal components factor analysis An exploratory factor analysis using the principal components method was conducted on the data, which illustrated the approaches to learning and performance of these students as evidenced through their responses to the questionnaire. Factor analysis is a data-reduction technique designed to reduce a large number of variables to a smaller number of coherent subsets or factors (Green and Salkind, 2005). Its aim is to produce a smaller number of factors or components, by grouping variables based on their intercorrelations, in such a way that most of the variability in the pattern of correlations is accounted for (Pallant, 2001). Variables that correlate with one another and which are independent of other subsets of variables are combined into a single factor (Tabachnick and Fidell, 2006). Prior to conducting the analysis, the suitability of data for factor analysis was investigated. The KMO measure of sampling adequacy and Bartlett’s Test of Sphericity were explored. The KMO value was .801 and Bartlett’s Test of Sphericity was significant (p < .0005). It was, therefore, concluded that the data were suitable for factor analysis. The unrotated solution of the principal components analysis revealed six components with eigenvalues above the value of 1. The observation of the scree plot showed a break after the fourth component and it was decided to retain four components for further analysis. Factor rotation was conducted to aid the interpretability of the data. The Varimax rotation method was used to ensure that the extracted components were uncorrelated and to facilitate interpretation. The factor scores from the rotation of the four factors were saved using the Anderson-Ruben method, which ensures that factor scores are uncorrelated (Field, 2000) to enable their use in further analyses. The four components overall explained 45.1% of the variance. The factor scores were retained to be used in further analyses (see t-tests and multiple regressions below).
METHODOLOGY A questionnaire seeking information on a variety of issues relating to learning and performance was given to adolescent musicians attending junior conservatoires and/or playing in youth orchestras in two geographical locations, namely the UK and Cyprus. 410 students completed the questionnaire (199 from the UK and 210 from Cyprus). Student ages ranged from 12 to 19 years old. The mean sample age was 15.33, SD = 1.94. There was a slight imbalance in terms of gender, as females constituted a larger percentage in the sample (58%) compared to males (42%). Given that the participants were chosen randomly, as long as they fell between the desired age range, this gender imbalance was considered to represent the slight predominance of females in the approached conservatoires and youth orchestras at the time of collecting the data. The instrumental standards of the students ranged from Grade 1 (elementary) to post-Grade 8 levels. A significant number of students studied for advanced Grade levels, with 66% reporting to be between Grade 7 and post-Grade 8 levels. 39.3% of the students were piano/keyboard players, 26.2% were string players, 13.1% were woodwind players, 9.2% were guitar players, 5.2% were brass players, 2.5% were percussion players, 2.5% were singers, 1.5% were bouzouki players and .5% were harp players.
Component 1 was characterised most prominently by variables relating to the experience of high levels of performance anxiety, by negative performance experiences and the experience of the effects of anxiety during performance. The negative effects of anxiety were highlighted, together
In the questionnaire, students were asked to indicated their degree of agreement to a number of statements rated on a five-point Likert-type scale ranging from strongly disagree ISBN 88-7395-155-4 © 2006 ICMPC
971
ICMPC9 Proceedings
with negative outcome expectancies. This component was labelled ‘suggestion for susceptibility to maladaptive anxiety’ and was related to sensitivity to the degree of selfexposure (perception of high self-exposure leads to higher performance anxiety), to the fact that performing environment quality increased anxiety levels (threatening or uncomfortable venue), a need for perfectionism, low selfefficacy in music, a negative musical self-concept and a concern regarding others’ reactions to playing. This component accounted for 18.88% of the variance.
Table 2: Most prominent variables of component 2 Component 2:
Table 1: Most prominent variables of component 1
Evidence of achievement, positive musical identity and effort
Component 1: Suggestion of susceptibility to maladaptive performance anxiety Variable
Loading
Variable
Loading
Perception of positive feedback from teacher
.674
Evidence of pre-evaluation anxiety
.775
Development of musical identity
.617
Description of negative experiences in performance
.764
Perception of receiving pressure to continue with music lessons
.474
Experience of physiological symptoms of anxiety
.756
Effortful practice
.448
Entity view on ability
.430
Experience of heightened anxiety when audience present
.746 Positive self-concept in music
.401
Sensitivity to degree of self-exposure
.562
Need for perfectionism
.399
Negative outcome expectancies
.559
Focus on negative effects of anxiety
.548
Positive self-concept in music
-.442
Environment (venue) quality can increase anxiety
.421
Low self-efficacy in music
.434
Concern on others' reaction to playing
.386
Need for perfectionism
.338
Perception of positive feedback from teacher
-.308
Component 3 was concerned with negative perceptions regarding anxiety’s effects on career and related to perceptions of unsuccessful coping with anxiety. This variable, in combination with perceived low self-efficacy in music, lack of intrinsic motivation to learn a musical instrument, an entity (relatively fixed) view of ability and lack of effort in practice seemed to lead to unsuccessful coping with performance anxiety, reinforced by the perception of unsupportive parents. This component was named ‘indication of unsuccessful coping with anxiety, lack of motivation and low self-confidence’. This component accounted for 9.19% of the variance.
Component 2 focused on the perception of positive feedback and the development of a positive musical identity. Effortful practice was also present, along with the perception of pressure to continue with music lessons. These two variables may be connected in that perception of pressure is related to satisfying others’ expectations, which is why effort in practice was also present. Effort in practice could be the reason why positive feedback was one of the most predominant variables in component 2. Positive selfconcept in music also appeared, which may also be related to the perception of positive feedback and a general sense of achievement in music. Perfectionism and an entity view of ability also appeared in this component. This component was named ‘evidence for achievement, positive musical identity and effort’. This component accounted for 9.41% of the variance. ISBN 88-7395-155-4 © 2006 ICMPC
Table 3: Most prominent variables of component 3 Component 3: Indication of unsuccessful coping with anxiety, lack of motivation and low self-confidence Variable Negative perception of anxiety
972
Loading .695
Low self-efficacy in music
.550
Intrinsic motivation to learn a musical instrument
-.515
Entity view on ability
.475
Incremental view on ability
-.435
ICMPC9 Proceedings
Put effort in practice
-.407
Perception of supportive and encouraging parents
-.332
Differences in nationality were observed in the components ‘suggestion for susceptibility to maladaptive anxiety’ and ‘tendency for self-reliance and intrinsic motivation’, but not in ‘evidence for achievement, positive musical identity and effort’ and ‘indication of unsuccessful coping with anxiety, lack of motivation and low self-confidence’.
Component 4 reflected an apparent lack of concern regarding others’ opinions. A positive self-concept in music and intrinsic motivation to learn a musical instrument, coupled with a lack of concern regarding other people’s judgements illustrated strong self-reliance. This component was named ‘tendency for self-reliance and intrinsic motivation’. Nevertheless, the influence of parental attitudes in this component was obvious. This may imply that the development of self-reliance and autonomy as a performer is somehow related to the active involvement of parents, through offering support and constructive feedback (criticality in feedback can be constructive). A positive self-concept in music allied to an intrinsic motivation to learn a musical instrument were also present in this component, which accounted for 7.60% of the variance. A focus on the negative effects of anxiety on performance was also highlighted in this component, which could be associated with an infrequent experience of anxiety in performance (provided that there is tendency for self-reliance and intrinsic motivation to learn) and the development of the perception that if it were present, it would probably lead to negative results.
There was a significant difference in scores for ‘suggestion for susceptibility to maladaptive anxiety’ for Cypriot (M = .16, SD = 1.08) and British (M = -.19, SD = .89) students as evidenced by the t-test (t = 3.17, p = .002). The magnitude of the differences in the means of the two groups of students as measured by eta squared was .03, which indicated that 3% of the variance in ‘suggestion for susceptibility to maladaptive anxiety’ was explained by nationality, indicating a small effect. The Cypriot students showed more susceptibility to maladaptive anxiety compared to the British students. There was also a significant difference in scores for ‘tendency for self-reliance and intrinsic motivation’ for Cypriot (M = .53, SD = .94) and British (M = -.48, SD = .79) students as evidenced by the t-test (t = 10.15, p < .0005). The magnitude of the differences in the means of the two groups of students as measured by eta squared was .25, which indicated that 25% of the variance in ‘suggestion for susceptibility to maladaptive anxiety’ was explained by nationality, a large effect. The Cypriot students appeared to be more autonomous and have more intrinsic motivation. The British students seemed to be more dependent on others and less intrinsically motivated.
Table 4: Most prominent variables of component 4 Component 4: Tendency for self-reliance and intrinsic motivation Variable
Loading
Concern about others’ judgment
-.559
Concern on others' reaction to playing
-.514
Perception of critical parents with high expectations
.506
Perception of supportive and encouraging parents
.476
Positive self-concept in music
.398
Focus on negative effects of anxiety
.330
Intrinsic motivation to learn a musical instrument
.326
Differences in gender Differences in gender were observed in the components ‘suggestion for susceptibility to maladaptive anxiety’ and ‘tendency for self-reliance and intrinsic motivation’, but not in ‘evidence for achievement, positive musical identity and effort’ and ‘indication of unsuccessful coping with anxiety, lack of motivation and low self-confidence’. There was significant difference in scores for ‘suggestion for susceptibility to maladaptive anxiety’ between male (M = -.27, SD = .85) and female (M = .19, SD = 1.07) students as evidenced by the t-test (t = -4.18, p < .0005). The magnitude of the differences in the means of the two groups of students as measured by eta squared was .06, which indicated that 6% of the variance in ‘suggestion for susceptibility to maladaptive anxiety’ was explained by gender, which indicated a moderate effect. The results suggested that female students were more susceptible to maladaptive anxiety. The male students appeared to be less susceptible to maladaptive anxiety.
T-Tests Four independent sample t-tests were conducted to explore differences between British and Cypriot students, male and female students, younger and older students and students taking and not taking instrumental examinations across the four approaches to instrumental learning and performance revealed through the factor analysis. Differences in nationality ISBN 88-7395-155-4 © 2006 ICMPC
973
ICMPC9 Proceedings
There was also a significant difference in scores for ‘tendency for self-reliance and intrinsic motivation’ for male (M = -.12, SD = .91) and female (M = .18, SD = 1.05) students as evidenced by the t-test (t = -2.58, p = .01). The magnitude of the differences in the means of the two groups of students as measured by eta squared was .02, which indicated that 2% of the variance in ‘tendency for self-reliance and intrinsic motivation’ was explained by gender, which indicated a small effect. Female students appeared to be more self-reliant and be more intrinsically motivated. Male students seemed to be more externally reliant and be less intrinsically motivated.
students that took exams (M = .11 SD = 1.03) and students that did not take exams (M = -.34, SD = 1.02) was revealed by the t-test (t = 2.77, p = .006). The magnitude of the differences in the means of the two groups of students as measured by eta squared was .03, which indicated that 3% of the variance in ‘evidence for achievement, positive musical identity and effort’ was explained by whether students chose to take or not take instrumental examinations. This was a small effect. The results suggested that students that took instrumental examinations were more successful, had a more positive musical identity and put more effort into practice, whilst the students that chose not to take musical examinations were less successful, perhaps as an avoidance behaviour and, as a consequence, had a more negative musical identity and practised less.
Differences in age For the purposes of this analysis, students were categorised into two age groups. The first group consisted of students between the ages 12-15 and the second group consisted of students between the ages 16-19. The two groups consisted of a similar number of students and there was an equal range span. The first group represented the younger adolescent musicians and the second group denoted the older adolescent musicians. Differences in age were only observed in the component ‘evidence for achievement, positive musical identity and effort’.
Multiple regressions Three multiple regressions were conducted, which aimed at investigating the contributions of the extracted components to reported data on attainment, amount of weekly practice and number of annual public performances of the students. The aim was to explore how well the four extracted components were able to predict attainment as evidenced through students’ marks in their last instrumental exam, their amount of weekly practice of the students (calculated by multiplying days of practising per week with the length of each practice session in minutes) and the quantity of annual public performances (calculated by adding up the number of solo performances with the number of group performances carried out in the previous year). Higher scores on these variables indicated higher attainment, more practice and a higher number of public performances respectively.
There was significant difference in scores for ‘evidence for achievement, positive musical identity and effort’ between younger (M = .16, SD = 1.06) and older (M = -.09, SD = .98) students as evidenced by the t-test (t = 2.12, p = .035). The magnitude of the differences in the means of the two groups of students as measured by eta squared was .01, which indicated that 1% of the variance in ‘evidence for achievement, positive musical identity and effort’ was explained by age, which indicated a small effect. This finding suggested that younger students tended to see themselves as more successful, have a more positive musical identity and put more effort into practice. This may be explained by the fact that younger students were most probably working towards lower Grade levels, which were less demanding and required less amount of practice compared to more advanced Grades of the older students.
Prediction of last examination result The regression model as a whole was statistically significant [F = 7.83, df = 4, p < .0005). The effect size, as calculated by the multiple R was R = .343, R2 = .118 and the adjusted R2 = .103. This indicated that the model explained 10.3% of the variance in mark in the last exam taken. Results showed that the significant predictors were ‘suggestion for susceptibility to maladaptive anxiety’ and ‘evidence for achievement, positive musical identity and effort’. ‘Indication of unsuccessful coping with anxiety, lack of motivation and low self-confidence’ and ‘tendency for self-reliance and intrinsic motivation’ did not make a significant contribution to the model. The strongest contribution was made by ‘evidence for achievement, positive musical identity and effort’, with higher values in this variable being related to higher marks in examinations. ‘Suggestion for susceptibility to maladaptive anxiety’ was negatively related to reported attainment in last examination, which
Differences in examination undertaking Students were asked to state whether they took instrumental examinations or not and were categorised in groups according to whether the responded ‘Yes, I do take music exams’ or ‘No, I do not take music exams’. The t-test results showed that there were significant differences between students that took and students that did not take instrumental examinations in ‘evidence for achievement, positive musical identity and effort’. A statistically significant difference in scores for ‘evidence for achievement, positive musical identity and effort’ for ISBN 88-7395-155-4 © 2006 ICMPC
974
ICMPC9 Proceedings
indicated that susceptibility to maladaptive anxiety was, as expected, related to lower achievement.
Results indicated that the significant predictors were ‘suggestion for susceptibility to maladaptive anxiety’, ‘evidence for achievement, positive musical identity and effort’ and ‘indication of unsuccessful coping with anxiety, lack of motivation and low self-confidence’. ‘Tendency for selfreliance and intrinsic motivation’ did not reach statistical significance, although it was very close (p = .056). The strongest contribution was made by ‘evidence for achievement, positive musical identity and effort’, with higher values in this variable being related to a higher amount of weekly practice on behalf of the students. ‘Suggestion for susceptibility to maladaptive anxiety’ and ‘indication of unsuccessful coping with anxiety, lack of motivation and low self-confidence’ were both negatively related to the amount of weekly practice. This implied that students were less likely to practise if they were susceptible to maladaptive performance anxiety and less successful at dealing with it. This lack of practice could be related to them feeling less motivated to put effort if they had negative experiences due to their susceptibility to maladaptive performance anxiety and ineffectiveness at coping with it.
The contributions of the independent predictor variables included in the regression model are summarised in table 5 below: Table 5: Contribution of individual variables in prediction of last examination result Beta weight
t
Sig.
Suggestion of susceptibility to maladaptive anxiety
-.148
-2.41
.017
Evidence for achievement, positive musical identity and effort
.300
4.89