Music Performance Anxiety in adolescence and early adulthood: Its relation with the age of onset in musical training

Abstract

Music Performance Anxiety (MPA) is one of the major emotional problems affecting music professionals and music students; it can only be addressed on a professional basis if a more profound knowledge of determining factors is acquired. This study examines one of these factors, scarcely investigated until now: the age at which an individual began music training. The relation of age of onset with MPA is examined here in two separate samples of music students/pupils in Spain: one of 437 advanced music students (mean age = 22.64 years) and another one consisting of 209 pupils enrolled in music schools (mean age = 12.09 years). The first sample was tested with the Spanish version of the Kenny Music Performance Anxiety Index (KMPAI), and the second sample was tested with the Performance Anxiety Questionnaire (PAQ) elaborated by Cox and Kenardy (1993). These two independent samples, tested with different evaluation tools, yield results that indicate a significant relation between the age of musical training onset and the individual’s currently perceived level of MPA. Those students/pupils who started at the age of 7 or younger report lower levels of MPA. The article concludes with a discussion of these results’ potential implications on a theoretical and practical level.

Keywords

age of musical training onset music conservatory Music Performance Anxiety music school music students

Music Performance Anxiety (MPA) is a phenomenon that can affect musicians’ playing skills to the point of threatening their professional careers (Yoshie, Kudo, Murakoshi, & Ohtsuki, 2009). Salmon has defined MPA as “the experience of persisting, distressful apprehension about and/or actual impairment of performance skills in a public context, to a degree unwarranted given the individual’s musical aptitude, training and level of preparation” (1990, p. 3). Papageorgi, Creech, and Welch (2011) have pointed out that MPA is one of the psychological variables that most affect the quality of musical performances in front of an audience; it can eventually lead to a series of cognitive, physiological, and motor difficulties (Studer, Danusera, Hildebrandt, Arial, & Gomez, 2011; Wilson, 2002). In musicians with low MPA levels, these symptoms are not highly interrelated; however, they interact intensely with one another in the most severe cases (Kenny, Davis, & Oates, 2004). Other factors can also have a direct influence on the emergence and persistence of MPA: cognitive components associated with peer group acceptance, combined with an individual’s own negative language and associated thoughts concerning his/her performance, previous mistakes, and his/her self-image (Kobori, Yoshie, Kazutoshi, & Ohtsuki, 2011).

MPA affects musicians on all continents (Yoshie et al., 2009). According to Dalia Cirujeda (2004), 20% of advanced music conservatory students eventually abandon their studies due to problems with MPA (Dalia Cirujeda, 2004), and 40%–60% of those who go on studying note that their musical performances deteriorate in quality due to MPA (Kaspersen & Götestam, 2002; Marchant-Haycox & Wilson, 1992). After graduation, MPA will continue to be a part of their professional lives (Fishbein, Middlestadt, Ottati, Straus, & Ellis, 1988; Van Kemenade, Van Son, & Van Heesch, 1995). Other studies have highlighted the importance of this problem in musician subgroups: classical, popular, and jazz (Kenny et al., 2004; Papageorgi et al., 2011; Ryan & Andrews, 2009). It seems to affect women more than men (Iusca & Dafinoiu, 2012; Thomas & Nettelbeck, 2014; Vaag, Bjørngaard, & Bjerkeset, 2016).

From a theoretical perspective, Barlow’s Triple Vulnerability Model (Barlow, 2000) explains the probable origins of anxiety and the reasons for its persistence throughout different stages of a person’s life. This model has proven to be one of the most consistent and reliable integrative theories of its kind, supported by a large corpus of empirical investigations carried out by Kenny (2009a, 2009b), Kenny et al. (2004), Papageorgi et al. (2011), Orejudo, Zarza, Casanova, Rodríguez, and Mazas (2016), Zarza, Casanova, and Orejudo (2016a, 2016b) and Zarza, Orejudo, Casanova, and Mazas (2016), among others. Barlow’s model postulates the presence of three risk factors for anxiety to develop: a generalized (heritable) biological vulnerability, a generalized psychological vulnerability (based on early experiences) and a specific psychological vulnerability. The first and second factors provoke a general sense of helplessness, which becomes a general risk factor, creating the necessary conditions for anxiety disorders and depression (Barlow, 2000). The third refers to specific contextual factors triggered by anxiety cues, especially in the case of phobias: it is associated with experiences in concrete situations (Osborne & Kenny, 2008). Specific psychological vulnerability is therefore a product of a person’s entire conditioning and experience.

So far, research has focused more on personal traits than on contextual conditions that could have an effect on specific psychological vulnerability. Thus, among specific factors, most studies have only singled out the type of instrument or ensemble, along with rehearsal and performance conditions (Cox & Kenardy, 1993; Kaspersen & Götestam, 2002; Kenny et al., 2004; Papageorgi, Hallam, & Welch, 2007).

In this study, we will be focusing our attention on one factor mostly ignored by research until now: the age of musical training onset and its relation with MPA. One should point out that the age of onset in musical training tends to vary widely. Although most children start to learn an instrument before the age of 12, there is great variety: almost 50% of children have started their training before the age of 7 (Nagel, 1993). Thus, it would be advisable to study the influence of age of onset on MPA in adulthood, as well as on the characteristics of MPA in childhood and adolescence.

A series of articles by Ryan (1998, 2004, 2005) have replicated the original study carried out by Simon and Martens (1979), which showed that music students experienced anxiety regarding public performance in a manner similar to the anxiety experienced by a sportsman or sportswoman before a match. It has been shown that performance anxiety is present very early in children, even at the age of 3 years (Boucher & Ryan, 2011). They experience MPA quite similarly to the way adults suffer from it prior to a performance – in terms of anticipatory anxiety, sensitivity to context, and regarding how difficult the music is (LeBlanc, Jin, Obert, & Siivola, 1997). Furthermore, Chan (2011) found that the youngest members of an orchestra in Hong Kong, 174 musicians aged 7 to 18, reported anxiety levels that even altered the quality of their performance in a highly stressful examination situation. On the other hand, although certain qualitative differences between adults and adolescents have been noted, no significant MPA differences are observable between boys and girls up to the pre-teen years, and a pronounced habituation effect sets in when two performances occur within short succession (Boucher & Ryan, 2011; Ryan, 2005).

None of the studies hitherto conducted has analyzed the relation between the age of musical training onset and subsequent manifestations of MPA in individuals. Boucher and Ryan (2011) are the only ones to have remarked that children with prior musical experience are less prone to MPA. Apart from that finding, the most relevant relation hitherto proposed between the age of onset and musical achievement is that the acquisition of certain technically demanding skills requires early initiation to music (Lehmann & Kristensen, 2014). In two independent studies, Bailey and Penhune (2010, 2012) compared musicians who started early (before the age of 7) with others who started playing music later on, and found that the two groups achieved significantly different skill levels in nonverbal tasks such as visual-motor and auditory motor synchronization. Improved short-term memory skills would also explain the higher performance levels achieved by those who started early, even taking other variables into account such as the number of years taking lessons, the number of musical practice years and total hours of practice. These authors hypothesize that there is a critical period in life for developing such skills. For their part, Moore, Burland, and Davidson (2003) note that a small group of successful music students had generally begun training earlier than those who dropped out; however, they had also had a close relationship with their professors in the first years, enjoyed significant family support and were encouraged to practice, and had offered public performances at a very early age. That study, however, did not analyze MPA as a variable. Thus, the current study’s objective is to analyze the relation between the age of musical training onset and MPA, while taking other possible variables into account that may contribute to this relationship, such as years of musical practice, age, sex or type of musical instrument. For that purpose, we selected a sample of music students enrolled in two different types of institutions. On the one hand, we gathered data from advanced university-level music academies (conservatorios superiores) in which students tend to enroll when they are 18 or older, with the purpose of attaining the highest degree of skill. On the other hand, we also gathered data from music schools (conservatorios profesionales) in which grade-school and secondary-level students enroll, either for recreational purposes or with hopes of going on to study for a professional career. Apart from the above-mentioned theoretical implications, the age of onset in musical training has practical implications, since the educational system in Spain regulates at which age children can enter music schools. As a hypothesis, we expected to find differences in MPA levels between participants depending on their age of onset, both in groups of younger pupils (attending music schools) and among older students (enrolled in advanced music academies). Since the number of practice years can also have an influence on such differences, we took that variable into account as well, despite the fact that previous studies had shown only slight differences between years of practice and level of MPA (Sadler & Miller, 2010). Although the number of years of musical practice has a significant bearing on a person’s degree of musical competence (Sloboda, Davidson, Howe, & Moore, 1996), it does not predict success any better than, say, the amount of public performance experience at a very young age (Moore et al., 2003).

Regarding the type of instrument, further studies have also revealed differences in personality traits or other psychopathological indicators, such as sleep problems (Buttsworth & Smith, 1995; Vaag, Bjørngaard, et al., 2016; Vaag, Saksvik-Lehouillier, Bjørngaard, & Bjerkeset, 2016).

Method

Participants

As indicated above, we worked with two separate samples. Sample I was composed of 437 students studying for an academic music degree; 54.2% of them were male and 45.8% female. Aspiring to become professional musicians and music teachers, they were enrolled in six advanced music academies (conservatorios superiores) in Spain. The sample mean age was 22.64 years (SD = 4.73), ranging between ages 16 and 51. The males in the sample were slightly older than the females, and with a wider age spread: males = 23.33 (5.15), females = 21.83 (4.02); Brown-Forsythe: F(1, 430.60) = 11.458, p = .001, η² = 0.025. Their specialties, including their instrument of specialty, are listed in Table 1.

Table 1.

Sample description.

	Sample I (University students)		Sample II (Music school students)
n	437		209
Sex	N (%)	Mean age (SD)	N (%)	Mean age (SD)
Female	54.2%	23.33 (5.15)	124 (59.3%)	11.74 (2.97)
Male	45.8%	21.83 (4.02)	85 (40.7%)	12.61 (5.55)
	Instrumental family
Woodwind	117	26.8	25	12.0
Brass instruments	64	14.6	19	9.1
Bowed string	126	28.8	49	23.6
Keyboard	66	15.1	75	36.1
Plucked string	28	6.4	20	9.6
Drums	12	2.7	20	9.6
Singing	18	4.1
Others	6	1.4
Questionnaire	KMPAI (Kenny et al., 2004)(Spanish Version)		PAQ (Cox & Kenardy, 1993)(Spanish Version)

Sample II comprised a total of 209 music school pupils, of which 40.7% were male and 59.3% were female. Their ages ranged from 8 to 46 years, but 92.3% of them were aged 15 or younger. The mean age was 12.09 years (SD = 4.22). In this case, although males were slightly older, the age difference between males and females was not statistically significant (Table 1, Brown-Forsythe: F(1, 117.12) = 1.743, p = .189, η² = 0.010); here the age spread among males was likewise wider than among females (Levene’s F(1, 207) = 4.387, p = .037). All these students were enrolled in the music school of Zaragoza, Spain (Conservatorio Elemental de Música de Zaragoza). The instrumental families occurring in the sample are described in Table 1.

Variables and tools

Music Performance Anxiety

In order to evaluate the participants’ degree of music performance anxiety (MPA), each of the two samples was tested with a different tool. For the advanced music academy sample, we applied the Spanish adaptation by Zarza, Orejudo, et al. (2016) of the Kenny Music Performance Anxiety Inventory (KMPAI), which features 26 items on a 7-point Likert scale (Kenny et al., 2004) and in which higher scores on KMPAI correspond to higher levels of MPA. The KMPAI tool groups the aspects related with vulnerability according to Barlow (2000) into three general factors reflecting experiences within a family context (3 items; Cronbach α = 0.56; e.g. “Excessive worrying is characteristic of my family”), a general vulnerability (10 items; Cronbach α = 0.78; e.g. “Sometimes I feel anxious for no reason”), and a specific factor related to musical performance cognition (11 items; Cronbach α = 0.86; e.g. “Prior to or during a performance I have feelings akin to panic”).

To test the sample of younger music school pupils, we applied the Spanish adaptation (Dalia Cirujeda, 2004) of the Performance Anxiety Questionnaire elaborated by Cox and Kenardy (1993), a 20-item tool that evaluates cognitive and physiological symptoms (Fehm & Schmidt, 2006) related to MPA on a 5-point Likert scale, along three different levels of performance conditions: study, practice or rehearsal either alone or in a group; performance as a member of a group, and concert performance as a soloist in which the higher scores the higher level of MPA. The Cox and Kenardy Performance Anxiety Questionnaire is generally considered more appropriate for younger-aged samples than KMPAI. Although participants are normally tested in all three areas (practice/rehearsal, group performance, solo performance), in this case we only tested the pupils in this sample on two conditions: group and solo performance, and we also eliminated two further ones (item number 5 “I feel nervous” and item number 19 “I need to urinate more often”) that were detrimental to internal consistency. We thereby differentiated a cognitive factor (10 items, α = 0.81 in group performance and α = 0.86 in solo performance; e.g. “I worry about my performance”) and a physiological factor (8 items, α = 0.77 group and α = 0.80 solo; e.g. “I feel tense in my stomach”). The four different conditions under consideration display significant intercorrelations of between 0.425 and 0.712.

In order to gather sociodemographic variables related with musical training and experience, we asked participants to indicate their age, gender, age of onset (musical training) and years of musical experience, along with the type of instrument they were learning.

Procedure

Both samples were collected within a formal music teaching context. The first sample was gathered according to availability, in terms of which institutions of higher learning were willing to participate, and also in terms of which students were willing to fill out the questionnaires. To that purpose, we contacted the directors of all 23 Spanish university-level music academies (conservatorios superiores), requesting collaboration in this study. Once affirmative answers arrived from six institutions, we agreed with them on dates and times for a member of our research team to carry out the survey on their respective premises. Pen-and-pencil questionnaires were administered to students; we emphasized that their participation was voluntary and that their responses would be kept anonymous.

The second survey was to be administered to minors; therefore, we first sent a letter to their parents, explaining the study’s main goals and asking if they objected to their child’s participation. Those children whose parents had stated they did not want their offspring to participate were not handed questionnaires in the classrooms. We also selected the music school in terms of its availability; the Conservatorio Elemental de Música de Zaragoza is the only state-sponsored music school of its kind located in the same town as our own institution. In terms of the responses we gathered, those students who indicated they had no experience in group performance because they played a solo instrument did not fill in the corresponding portion, and vice-versa for group-oriented instruments which do not tend to be played alone.

Results

Age of onset and years of practice

Among the university-level students (Table 2), the mean age at which they began playing an instrument was 8.80 years old (SD = 3.07), the minimum age of onset was 2 and the maximum 25. Broken down according to gender, both men and women differ in the mean and standard deviation, Brown-Forsythe: F(1, 433.79) = 7.241, p = .007, η² = 0.017, with men starting later and more heterogeneously (Table 2).

Table 2.

Students age of onset distribution. Samples I and II.

Age group (years)	Sample I (University students)			Sample II (Music school students)
Age group (years)	N	%		N	%
⩽ 7	146	33.4		126	60.3
8	124	28.4		30	14.4
9–10	82	18.8		24	11.5
⩾ 11	85	19.5		29	13.9
Sex
		M	SD		M	SD
Male		9.16	3.21		7.38	3.17
Female		8.38	2.85		7.22	2.36
Total		8.80	3.07		7.29	2.72
Instrumental family
		M	SD		M	SD
Woodwind		8.84	2.32		7.88	2.35
Brass instruments		9.25	2.15		8.05	3.17
Bowed string		7.65	1.99		6.24	2.21
Keyboard		8.11	2.83		7.47	2.70
Plucked string		9.68	3.76		8.25	2.57
Drums		10.83	4.37		6.70	3.42
Singing		15.28	5.49		–	–
Others		7.67	2.58		–	–
Total		8.80	3.07		7.28	2.72

Broken down into age blocks to test the association between those and sex, χ²(3) = 10.821, p = .013, φ = 0.157, a lower percentage of females had started at age 11 or thereafter (68.2% of boys, 31.8% of girls), whereas girls were more frequent in the age group where musical practice onset was at age 7 or under (46% of boys 53.4% of girls).

With respect to the number of years of music practice, the university-level students had spent a mean total of 13.86 years (SD = 4.09) learning music, and here there were no significant differences between males and females: males = 14.16 (4.34), females = 13.50 (3.74); Brown-Forsythe: F(1, 432.77) = 2.914, p = .089; once more, however, the group of boys was more heterogeneous in this aspect.

In the second sample, the mean age of onset was 7.29 years (SD = 2.72), with significant differences neither between the sexes, nor between averages, Brown-Forsythe: F(1, 145.70) = 0.161, p = .689, although variance was greater in the group of boys, Levene’s F(1, 207) = 10.667, p = .001, nor between age blocks, χ²(3) = 5.464, p = .141.

In accordance with their lower age, this group had been learning music for a mean of 4.80 years (SD = 3.87), with a slight difference in favor of the boys, males = 5.22(4.88), females = 4.51(2.98), a difference which nevertheless was not statistically significant, Brown-Forsythe: F(1, 126.93) = 1.418, p = .236.

A relationship between age of onset and type of instrument was found, F(7, 409) = 20.889, p < .001, η²_p = 0.254. Pairwise comparisons (Scheffé) showed three different homogeneous sub-groups. The first one was composed of singing students. They began studying music later due to the physiological requirements of their specialty: this was the group with the latest age of onset. Another subgroup included two groups of younger instrumentalists (bowed strings and the category of “other instruments”) and all other groups except voice students. Finally, the third group included students whose age of onset was intermediate, therefore not belonging to the singing group, to bowed strings or to “other instruments.” The results of each group are shown in Table 2.

Music performance anxiety (MPA)

University-level music students

In the age groups corresponding to musical practice onset (Table 3), the one-way ANOVA showed significant mean differences in the specific cognitions of MPA, F(3, 433) = 6.071, p < .001, η² = 0.04. This result remained significant after a MANCOVA control of age and years of study was applied, F(3, 425) = 9.066, p < .001, η² = 0.060. Related to gender, both a main effect, F(1, 425) = 24.672, p < .001, η² = 0.055, and an interaction with the age of onset, F(3, 425) = 6.887, p < .001, η² = 0.046, were found. However, no gender interaction with age was found, F(1, 428) = 3.014, p = .083, η² = 0.007, or years of study, F(1, 428) = 1.242, p = .266, η² = 0.003. By gender, the females scored higher on the specific factors of anxiety and helplessness (Table 3), but according to the interaction found, this trend did not occur in the group of participants who started at the age of 11 (males = 40.81, SD = 12.89; females = 36.54, SD = 10.52).

Table 3.

KMPAI factor differences associated with gender and age of onset in musical studies.

		N	M	SD	F	p	η²
MPA* (Specific cognitions)	Male	237	37.19	12.93	32.295	< .001	0.069
	Female	200	43.96	11.73	32.295	< .001	0.069
	age ⩽ 7	146	37.21	13.36	6.071	< .001	0.04
	age 8	124	41.75	11.66
	ages 9–10	82	44.15	13.10
	age ⩾ 11	85	39.73	12.16
Helplessness	Male	237	31.31	9.70	5.929	.015	0.013
	Female	200	33.55	9.42	5.929	.015	0.013
	age ⩽ 7	146	30.54	9.05	2.843	.037	0.019
	age 8	124	32.80	10.00
	ages 9–10	82	33.99	10.05
	age ⩾ 11	85	33.16	9.31
Early context	Male	237	8.97	3.78	0.948	.331	0.002
	Female	200	8.60	4.05	0.948	.331	0.002
	age ⩽ 7	146	8.02	3.49	7.100	< .001	0.047
	age 8	124	8.58	3.76
	ages 9–10	82	8.87	3.91
	age ⩾ 11	85	10.40	4.37

Music Performance Anxiety.

Post-hoc tests revealed that two groups stand out with a higher mean score (onset at ages 9 or 10) and a lower mean score (onset at age 7 or earlier). The other age groups are neither differentiated among themselves, nor with relation to the two aforementioned groups with more extreme scores.

With regard to the helplessness factor, a one-way ANOVA by gender revealed slight differences in favor of females (Table 1), whereas analysis according to age onset groups once more revealed significant differences, F(3, 433) = 2.843, p = .037, η² = 0.019. The same above-mentioned MANCOVA control revealed no relation with the participant’s current age, F(1, 428) = 0.111, p = .74, nor with the number of years of study, F(1, 428) = 0.061, p = .806, however, just as mentioned above in the ANOVA result, a relation with gender was indeed observed, F(1, 428) = 7.889, p = .005, η² = 0.018. In this case, the post-hoc tests revealed that the significant differences (p = .046) which can be observed only differentiate the group of students who started learning an instrument at age 7 or earlier from the group of students who started between ages 9 and 10.

With relation to early family context, Table 3 shows that there were no significant differences for gender, but they did exist among the four groups divided into age of musical practice onset, F(3, 433) = 7.1, p < .001, η² = 0.047. In this case, post-hoc comparisons suggested that the group of those who started playing music at age 11 or older was significantly different from the other groups, which did not differ among one another (Table 3). However, this result did not remain significant after the MANCOVA analysis, F(3, 428) = 0.687, p = 0.56. None of the co-variates attained statistical significance: age, F(1, 428) = 2.176, p = .141, gender, F(1, 428) = 0.014, p = .906, years of study, F(1, 428) = 0.241, p = 0.624.

To control the possible effect of type of instrument (typically soloist/typically orchestral), we carried out multiple regression in blocks (step-wise method). The variables we included, and the results, are compiled in Table 4. Block 2 included all instruments as dummy variables. As can be seen, the variables that were entered into the regression equation are those already mentioned in the ANOVA, without considering any variable relative to instruments.

Table 4.

MPA regression analysis (Sample I and Sample II).

Predictors
Criterion variable	Block 1		Block 2	Block 3
	Socio-demographics		Instrument (dummy)Years of study	Age of onset
Sample I (University students)
Specific cognition	Sex	0.277***		Sex	0.283***
				Age	0.108*
	Age	0.115**		Onset at 7 years	−0.211***
				Onset at 11 years	−0.108*
	R² =	0.080		R² =	0.112
Sample II (Music school students)
Cognitive-soloist	Sex	0.221**		Sex	0.235**
	Age	0.064*		Age	0.069
	Age	0.064*		Age of onset	0.203*
	R² =	0.069		R² =	0.101
Cognitive-group				Age of onset	0.223*
Cognitive-group				R² =	0.085

p ⩽ .05; **p ⩽ .01; ***p ⩽ .001.

Music school pupils

In this case, no significant differences were notable between mean results by gender in any factor within group performance conditions (Table 5). In solo performance conditions, however, differences were observed by gender, both in the cognitive factor, F(1, 207) = 7.866, p = .006, η² = 0.037, and in the physiological factor, F(1, 207) = 7.841, p = .006, η² = 0.036. In both cases, females obtained higher scores.

Table 5.

PAQ factor differences associated with gender and age of onset in musical studies (Sample II “Music school students”).

		N	M	SD	F	p	η²
Group (cognitive aspects)	Male	85	26.92	6.73	2.821	.095	0.013
	Female	124	28.59	7.23	2.821	.095	0.013
	age ⩽ 7	126	26.60	7.26	5.138	.002	0.070
	age 8	30	27.97	5.97
	ages 9–10	24	30.77	6.11
	age ⩾ 11	29	31.19	6.47
Group (physiological aspects)	Male	85	15.90	4.66	2.802	.096	0.013
	Female	124	17.17	5.80	2.802	.096	0.013
	age ⩽ 7	126	16.32	5.32	0.800	.495	0.012
	age 8	30	16.60	5.72
	ages 9–10	24	16.82	4.35
	age ⩾ 11	29	18.03	6.11
Soloist (cognitive aspects)	Male	85	31.83	8.39	7.866	.006	0.037
	Female	124	35.22	8.75	7.866	.006	0.037
	age ⩽ 7	126	32.61	9.33	3.417	.018	0.048
	age 8	30	33.47	7.99
	ages 9–10	24	37.88	6.12
	age ⩾ 11	29	36.25	7.46
Soloist (physiological aspects)	Male	85	19.03	6.06	7.841	.006	0.036
	Female	124	21.79	7.57	7.841	.006	0.036
	age ⩽ 7	126	20.40	6.97	0.275	.843	0.004
	age 8	30	20.53	8.15
	ages 9–10	24	21.78	6.92
	age ⩾ 11	29	21.01	7.02

On the other hand, an ANOVA on the age onset factor revealed no differences in function of that variable, in either physiological anxiety in group rehearsal/performance, F(3, 205) = 0.8, p = .495, or in solo performance, F(3, 205) = 0.275, p = .843. In the cognitive factor, however, we did find significant differences in both cases: group rehearsal/performance, F(3, 205) = 5.138, p = .002, η² = 0.070, and solo performance, F(3, 205) = 3.417, p = .018, η² = 0.048. Just as in the previous case, MANCOVA control analysis showed that the results persist in the cognitive factor, both in group performance, F(3, 132) = 4.559, p = .005, η² = 0.094, and in solo performance, F(3, 132) = 3.748, p = .013, η² = 0.078. But here the physiological factor was not significant in either the group situation, F(3, 132) = 1.998, p = .117, or in solo performance, F(3, 132 ) = 2.002, p = .117. With respect to the covariates, current age was not significant in any of these cases, whereas the above-mentioned gender and years of study effects persisted and were relevant in both the cognitive-group factor, F(1, 132) = 3.932, p = .049, η² = 0.029, and the physiological-soloist factor, F(1, 132) = 4.591, p = .034, η² = 0.034.

With respect to differences among different groups (Table 5), post-hoc tests indicated that the differences in mean scores in cognition as a member of a group emerged between those who started learning music at age 7 at the latest, those who started between 9 and 10 (p = .041) and between the young age of onset and those who started at age 11 or older (p = .008). In solo performance conditions, we found significant differences only emerging between those participants who began at age 7 or earlier and those who started between ages 9 and 10 (p = .038).

As in the case of the sample I, in order to control the effect of the instrument factor, we conducted a regression analysis on the performance anxiety variables (Table 5). Again, no variables related to the instrument were considered in the regression equations.

Conclusions and discussion

In this study, we have attempted to analyze the relation between age of onset and MPA, one of the afflictions most frequently suffered by students and professionals in the musical field. Most studies had hitherto focused on personality traits; however, theoretical models (Barlow, 2000; Kenny, 2009a, 2009b) had postulated the importance of contextual factors associated with life experiences. Thus, we confirm the notion that the age when pupils start learning to play music leads to significantly different levels of MPA later on. These results have been confirmed in two very different groups of students: older ones, studying at university level in conservatories and music academies to become professional musicians, therefore subjected to frequent stressors and a great amount of pressure to attain a high level of mastery, and a group of younger pupils in a music school with a population that includes participants who desire to become music professionals, mixed with others who wish to remain amateurs. Moreover, these results are independent of the number of years a student/pupil has been practicing an instrument or following music training or instrument played.

To summarize, these results suggest that musicians who start learning their instrument at the age of 7 or younger experience a lower level of anxiety than those who only started at ages 9 or 10. As to those who indicate having started at 8, music school pupils feature a profile similar to those who started earlier, but this finding is not replicated in the sample of older (university-level) participants. Finally, in the older sample, the group of those who started at age 11 or later show a profile similar to those who started at 7 or earlier, only now there are no significant differences between boys and girls. The theoretical implications of these findings are probably divergent for each case.

The children and adolescents in the younger sample are confronting the experience of performing in public at a time in their lives when the processing of performance-related information tends to change qualitatively with age. As Colonnesi, Engelhard, and Bögels (2010) have proposed, children can feel embarrassment at those ages, but the mechanisms with which they grasp their own reactions and those of others tend to differ according to age. They grow in complexity from the age of 8 onwards. That very age is generally associated with a series of mechanisms that help one learn to value one’s own mental representation of oneself in relation with an audience – either by adopting behavior that does not correspond with the established social norm or, on the other hand, behavioral tendencies that seek recognition on the part of others in a public situation (Banerjee, Bennett, & Luke, 2010; Naito & Seki, 2009). Something similar is suggested by Modigliani and Blumenfeld (1979), who propose that children at the age of 8 differentiate “primitive” embarrassment from mature embarrassment. The latter emotion displays a greater level of cognitive complexity since it stems from evaluating the discrepancy between one’s own self-image and the image one projects to an audience. Along the same lines, Bennett (1989) has ascertained that, thanks to that greater acquired cognitive ability, children from the age of 8 onwards can feel anxiety in front of an audience, even if its members are not particularly critical. Those different ways of confronting public situations at different ages, along with the fact that children in those ages already experience anxiety (Boucher & Ryan, 2011), can entail that a child younger than 8 who is about to confront an audience is thereby actually somewhat “vaccinated” against the stress often associated with such experiences. In later ages, conversely, the situation of having to perform can have an entirely different effect: anxiety, in fact, can continually reinforce itself in a vicious circle if such situations tend to occur on a regular basis (Colonnesi et al., 2010). Along those lines, Moore, Burland, and Davidson (2003), analyzed music students’ success profiles. They found that those who pursued their musical training and had access to advanced conservatory had tended to perform a greater number of public recitals at early age than those who were not able to continue music training or abandoned it. This even applied when there was no difference in terms of ages of onset.

Such situations could be affected, however, by reactions in the children’s environment, particularly on the part of their family and closest peers. Thus, family context and the support provided by close mentors can play a crucial role, not only in helping these children develop specific self-regulation skills, but also in identifying areas of interest that can be reinforced (Lehmann & Kristensen, 2014). One should point out that support on the part of adults is particularly relevant in activities requiring a great degree of specifically focused self-control, and thus requiring an early age of onset. Nevertheless, the role played by adults can also become a risk factor when their preparation as role models, teachers and mentors does not take children’s specific needs into account.

Another additional element could confirm the contextual effect on MPA: varying reactions on the part of professors, and the individual’s environment in those ages. Bastian (1989) found that the type of reaction on the part of adults present in the environment of younger children is warmer and more centered on them and their needs than at later ages. Apart from the different ways professors can treat children, one should also consider the different degrees of sensibility that youngsters in those ages display when reacting in public performances to feedback coming both from adults and from their peers (Mizokawa, 2015). This leads authors such as Lehmann and Kristensen (2014) to recommend that professors who introduce children to music should undergo thorough specialized training.

A further hypothesis could help explain our results. The notion of critical periods in life for the acquisition of skills such as music could corroborate the notion that those who initiate their training the earliest tend to attain greater mastery: this would lead the frequency of public performances to increase, thereby exposing them early on to audiences, and providing them with a greater sense of self-efficacy (Sichivitsa, 2007). Additionally, if this occurs in a family and/or social context of people highly interested in music, with parents that encourage the child’s spontaneous participation in musical activities while planning more numerous and better-quality activities, it can have a bearing on the age of onset, and we would thus be dealing with a combination of elements that would reinforce a child’s musical skill in a positive feedback loop (Moore et al., 2003). We should point out that the sense of mastering a specific skill acts as a protector against anxiety: in fact, if the skill is mastered at a high level, such anxiety can even turn into a positively favorable activator that encourages good performance (McCormick & McPherson, 2003; McPherson & McCormick, 2006; Ritchie & Williamon, 2011).

Another group of students present a series of traits that are significantly differentiated from the other groups, depending on age of onset: those who started playing an instrument at age 11 or older, and are currently studying at a university-level conservatory. There is probably a different explanation for this result, namely, those who suffer the most from MPA tend to have already abandoned musical studies altogether by that time. Fehm and Schmidt (2006) found that pupils at the pre-academic music school level who did not wish to continue their training tended to display higher levels of MPA than those who were still considering a musical career. In our study, the fact that differences related to age of onset can only be found in the university-level student sample, with no significant difference between boys and girls, would tend to corroborate that hypothesis.

We acknowledge some limitations of this study. Firstly, our results were gathered from the participants’ self-assessment: thus, the variables they indicated, such as age of onset of musical practice, could suffer from bias, and even more so since we are dealing with conditions and activities that have their origins in childhood. Secondly, the possibility of comparing the two sample groups is limited by the fact that different instruments were used to gather data from one sample and the other. Thirdly, a final limitation comes from the fact of having asked the participants when they started playing a musical instrument, although we actually presuppose, in theory, that those practices are associated with public appearances, i.e., situations in which a series of diverse processes of conditioning and sensitization to varying audience reactions can take place. In the future, this aspect will need to be addressed with greater care, despite the eventual danger of memory bias. At the same time, it would be advisable to study further chronological aspects of musical practice and formation: other studies have shown, for instance, that the number of previous years of practice is not the only predictor of success; another predictor is the manner in which practice and training are distributed along a chronological timeline (Biasutti & Concina, 2014; Moore et al., 2003).

Footnotes

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Programa Operativo FEDER Aragón 2014-2020 “Construyendo Europa desde Aragón.”

References

Bailey

J. A.

Penhune

V. B.

(2010). Rhythm synchronization performance and auditory working memory in early- and late-trained musicians. Experimental Brain Research, 204(1), 91–101. doi:10.1007/s00221–010–2299-y

Bailey

J. A.

Penhune

V. B.

(2012). A sensitive period for musical training: Contributions of age of onset and cognitive abilities. Annals of the New York Academy of Sciences, 1252, 163–170. doi:10.1111/j.1749–6632.2011.06434.x

Banerjee

Bennett

Luke

(2010). Children’s reasoning about the self-presentational consequences of apologies and excuses following rule violations. British Journal of Developmental Psychology, 28(4), 799–815. doi:10.1348/026151009X479475

Barlow

(2000). Unraveling the mysteries of anxiety and is disorders from the perspective of emotion theory. American Psychologist, 55(11), 1247–1263. doi:10.1037/0003–066X.55.11.1247

Bastian

H. G.

(1989). Leben für Musik. Eine Biographiestudie über musikalische (Hoch) Begabungen [Living for music. Biography of highly talented musicians.]. Mainz, Germany: Schott.

Bennett

(1989). Children’s self attribution of embarrassment. British Journal of Developmental Psychology, 7(3), 207–217. doi:10.1111/j.2044-835X.1989.tb00801.x

Biasutti

Concina

(2014). The role of coping strategy and experience in predicting music performance anxiety. Musicae Scientiae, 18(2), 189–202. doi:10.1177/1029864914523282

Boucher

Ryan

C. A.

(2011). Performance stress and the very young musician. Journal of Research in Music Education, 58, 329–345.

Buttsworth

Smith

(1995). Personality of Australian performing musicians by gender and by instrument. Personality and Individual Differences, 18(5), 595–603. doi:10.1016/0191-8869(94)00201-3

10.

Chan

M. Y.

(2011). The relationship between music performance anxiety, age, self-esteem, and performance outcomes in Hong Kong Music students (Unpublished doctoral dissertation). Durham University, Durham.

11.

Colonnesi

Engelhard

I. M.

Bögels

S. M.

(2010). Development in children’s attribution of embarrassment and the relationship with theory of mind and shyness. Cognition and Emotion, 24(3), 514–521. doi:10.1080/02699930902847151

12.

Cox

W. J.

Kenardy

(1993). Performance anxiety, social phobia, and setting effects in instrumental music students. Journal of Anxiety Disorders, 7(1), 49–60. doi:10.1016/0887-6185(93)90020-L

13.

Dalia Cirujeda

. (2004). Cómo superar la ansiedad escénica en músicos [How to overcome the Music Performance Anxiety]. Madrid, Spain: Mundimúsica Ediciones.

14.

Fehm

Schmidt

(2006). Performance anxiety in gifted adolescent musicians. Journal of Anxiety Disorders, 20(1), 98–109. doi:10.1016/j.janxdis.2004.11.011

15.

Fishbein

Middlestadt

S. E.

Ottati

Straus

Ellis

(1988). Medical problems among ICSOM musicians: Overview of a national survey. Medical Problems of Performing Artists, 3(1), 1–8.

16.

Iusca

Dafinoiu

(2012). Performance anxiety and musical level of undergraduate students in exam situations: The role of gender and musical instrument. Procedia – Social and Behavioral Sciences, 33, 448–452.

17.

Kaspersen

Götestam

K. G.

(2002). A survey of music performance anxiety among Norwegian music students. European Journal of Psychiatry, 16(2), 69–80.

18.

Kenny

D. T.

(2009a). The role of negative emotions in performance anxiety. In Juslin

Sloboda

(Eds.), Handbook of music and emotion: Theory, research, applications (pp. 425–452). Oxford, UK: Oxford University Press.

19.

Kenny

D. T.

(2009b). The factor structure of the revised Kenny Music Performance Anxiety Inventory. In Williamon

Pretty

Buck

(Eds.), Proceedings of the International Symposium on Performance Science 2009 (pp. 37–41). Utrecht, The Netherlands: Association Européenne des Conservatoires (AEC).

20.

Kenny

D. T.

Davis

Oates

(2004). Music performance anxiety and occupational stress amongst opera chorus artists and their relationship with state and trait anxiety and perfectionism. Journal of Anxiety Disorders, 18(6), 757–777. doi:10.1016/j.janxdis.2003.09.004

21.

Kobori

Yoshie

Kazutoshi

Ohtsuki

(2011). Traits and cognitions of perfectionism and their relation with coping style, effort, achievement, and performance anxiety in Japanese musicians. Journal of Anxiety Disorders, 25(5), 674–679. doi:10.1016/j.janxdis.2011.03.001

22.

LeBlanc

Jin

Y. C.

Obert

Siivola

(1997). Effect of audience on music performance anxiety. Journal of Research in Music Education, 45(3), 480–496.

23.

Lehmann

A. C.

Kristensen

(2014). “Persons in the Shadow” brought to light: Parents, teachers, and mentors – how guidance works in the acquisition of musical skills. Talent Development & Excellence, 6(1), 57–70.

24.

Marchant-Haycox

S. E.

Wilson

G. D.

(1992). Personality and stress in performing artists. Personality and Individual Differences, 13(10), 1061–1068. doi:10.1016/0191-8869(92)90021-G

25.

McCormick

McPherson

(2003). The role of self-efficacy in a musical performance examination: An exploratory structural equation analysis. Psychology of Music, 31(1), 37–51. doi:10.1177/0305735603031001322

26.

McPherson

McCormick

(2006). Self-efficacy and music performance. Psychology of Music, 34(3), 322–336. doi:10.1177/0305735606064841

27.

Mizokawa

(2015). Theory of mind and sensitivity to teacher and peer criticism among Japanese children. Infant and Child Development, 24(2), 189–205. doi:10.1002/icd.1877

28.

Modigliani

Blumenfeld

(1979). A developmental study of embarrassment and self-presentation. Unpublished Research Proposal. In Bennett

(1989), Children’s self attribution of embarrassment. British Journal of Developmental Psychology, 7(3), 207–217. doi:10.1111/j.2044-835X.1989.tb00801.x

29.

Moore

D. G.

Burland

Davidson

J. W.

(2003). The social context of musical success: A developmental account. British Journal of Psychology, 94(4), 529–549. doi:10.1348/000712603322503088

30.

Nagel

J. J.

(1993). Stage fright in musicians: A psychodynamic perspective. Bulletin of the Menninger Clinic, 57(4), 492–504.

31.

Naito

Seki

(2009). The relationship between second-order false belief and display rules reasoning: The integration of cognitive and affective social understanding. Developmental Science, 12(1), 150–164. doi:10.1111/j.1467-7687.2008.00748.x

32.

Orejudo

Zarza

F. J.

Casanova

Rodríguez

Mazas

(2016). The relation of music performance anxiety (MPA) to optimism, self-efficacy, and sensitivity to reward and punishment: Testing Barlow’s theory of personal vulnerability on a sample of Spanish music students. Psychology of Music. Advance online publication. doi:10.1177/0305735616674791

33.

Osborne

M. S.

Kenny

D. T.

(2008). The role of sensitizing experiences in music performance anxiety in adolescent musicians. Psychology of Music, 36(4), 447–462. doi:10.1177/0305735607086051

34.

Papageorgi

Creech

Welch

(2011). Perceived performance anxiety in advanced musicians specializing in different musical genres. Psychology of Music, 41(1), 18–41. doi:10.1177/0305735611408995

35.

Papageorgi

Hallam

Welch

(2007). A conceptual framework for understanding musical performance anxiety. Research Studies in Music Education, 28, 83–107. doi:10.1177/1321103X070280010207

36.

Ritchie

Williamon

(2011). Measuring distinct types of musical self-efficacy. Psychology of Music, 39(3), 328–344. doi:10.1177/0305735610374895

37.

Ryan

(1998). Exploring musical performance anxiety in children. Medical Problems of Performing Artists, 13(3), 83–88.

38.

Ryan

(2004). Gender differences in children’s experience of musical performance anxiety. Psychology of Music, 32(1), 89–103. doi:10.1177/0305735604039284

39.

Ryan

(2005). Experience of musical performance anxiety in elementary school children. International Journal of Stress Management, 12(4), 331–342. doi:10.1037/1072-5245.12.4.331

40.

Ryan

Andrews

(2009). An investigation into the choral singer’s experience of music performance anxiety. Journal of Research in Music Education, 57(2), 108–126. doi:10.1177/0022429409336132

41.

Sadler

M. E.

Miller

C. J.

(2010). Performance anxiety: A longitudinal study of the roles of personality and experience in musicians. Social Psychological and Personality Science, 1(3), 280–287. doi:10.1177/1948550610370492

42.

Salmon

(1990). A psychological perspective on Musical Performance Anxiety – a review of the literature. Medical Problems of Performing Artists, 5(1), 1–11.

43.

Sichivitsa

(2007). The influences of parents, teachers, peers and other factors on students’ motivation in music. Research Studies in Music Education, 29(1), 55–68. doi:10.1177/1321103X07087568

44.

Simon

J. A.

Martens

(1979). Children’s anxiety in sport and non-sport evaluative activities. Journal of Sport Psychology, 1, 160–169.

45.

Sloboda

J. A.

Davidson

J. W.

Howe

M. J. A.

Moore

D. G.

(1996). The role of practice in the development of performing musicians. British Journal of Psychology, 87(2), 287–309. doi:10.1111/j.2044-8295.1996.tb02591.x

46.

Studer

Danusera

Hildebrandt

Arial

Gomez

(2011). Hyperventilation complaints in music performance anxiety among classical music students. Journal of Psychosomatic Research, 70(6), 557–564. doi:10.1016/j.jpsychores.2010.11.004

47.

Thomas

J. P.

Nettelbeck

(2014). Performance anxiety in adolescent musicians. Psychology of Music, 42(4), 624–634. doi:10.1177/0305735613485151

48.

Vaag

Bjørngaard

J. H.

Bjerkeset

(2016). Symptoms of anxiety and depression among Norwegian musicians compared to the general workforce. Psychology of Music, 44(2), 234–248. doi:10.1177/0305735614564910

49.

Vaag

Saksvik-Lehouillier

Bjørngaard

J. H.

Bjerkeset

(2016). Sleep difficulties and insomnia symptoms in Norwegian musicians compared to the general population and workforce. Behavioral Sleep Medicine, 14(3), 325–342. doi:10.1080/15402002.2015.1007991

50.

Van Kemenade

J. F. L. M.

Van Son

M. J. M.

Van Heesch

N. C. A

. (1995). Performance anxiety among professional musicians in symphonic orchestras: A self-reported study. Psychological Reports, 77(2), 555–562. doi:10.2466/pr0.1995.77.2.555

51.

Wilson

G. D.

(2002). Psychology for performing artists (2nd ed.). London, UK: Whurr.

52.

Yoshie

Kudo

Murakoshi

Ohtsuki

(2009). Music performance anxiety in skilled pianists: Effects of social-evaluative performance situation on subjective, autonomic, and electromyographic reactions. Experimental Brain Research, 199(2), 117–126. doi:10.1007/s00221-009-1979-y

53.

Zarza

F. J.

Casanova

Orejudo

(2016a). Ansiedad escénica y constructos psicológicos relacionados. Estudiantes de cinco conservatorios superiores de música españoles [Music Performance Anxiety and related psychological constructs. Students of five Spanish music Conservatories.]. RIEM. Revista Internacional de Educación Musical, 4, 13–24. doi:10.12967/RIEM-2016-4-p013-024

54.

Zarza

F. J.

Casanova

Orejudo

(2016b). Estudios de música en los conservatorios superiores y ansiedad escénica en España [Music Studies at Conservatories in Spain and Music Performance Anxiety]. RECIEM. Revista Electrónica Complutense de Investigación en Educación Musical, 13, 50–63. doi:10.5209/RECIEM.49442

55.

Zarza

F. J.

Orejudo

Casanova

Mazas

(2016). Kenny Music Performance Anxiety Inventory: Confirmatory factor analysis of the Spanish version. Psychology of Music, 44(3), 340–352. doi:10.1177/0305735614567932