Development of the four-dimensional Motives for Listening to Music Questionnaire (MLMQ) and associations with health and social issues among adolescents

Abstract

In order to develop the Motives for Listening to Music Questionnaire (MLMQ) and to confirm its construct and concurrent validity, data from a nationally representative sample of 4,524 adolescents in Switzerland were used. The results confirmed the MLMQ’s four-dimensional factor structure (i.e., social, enhancement, coping and conformity motives) in general and across gender, age and linguistic region. Girls listened to music more frequently for coping, enhancement and social motives than boys. Structural equation modelling confirmed the hypothesized associations between coping motives and health-related outcomes (somatic complaints, aggressive and depressed mood, school pressure, low life satisfaction etc.), between social motives and peer-related activities (spending evenings with friends, bullying others etc.), and between conformity motives and being depressed and a victim of bullying. To conclude, confirming its psychometric qualities, the MLMQ is a valid and reliable instrument for assessing music motives in adolescent populations. In health care, the MLMQ could be used to identify individuals and to help applying music therapy in accordance with their motives for listening to music.

Keywords

adolescents confirmatory factor analysis music motives scale development

Music is a central element in adolescence and beyond. Research among young people (13–25-year-olds) has shown that listening to music can help to master developmental tasks (e.g. in peer affiliating processes and confrontation with daily life stressors like school pressure), regulate emotions, manage difficulties and provide points of reference in the process of constructing personal and social identities (Arnett, 1995; Chamorro-Premuzic, Gomà-i-Freixanet, Furnham, & Muro, 2009; Laiho, 2004; Lonsdale & North, 2011; Miranda & Claes, 2009; North, Hargreaves, & O’Neill, 2000; Ruud, 1997; Saarikallio & Erkkilä, 2007; Tarrant, North, & Hargreaves, 2000; Thoma, Scholz, Ehlert, & Nater, 2012). While young people listen to music for different reasons, mood management is often cited as one of the most important motives (Boer & Fischer, 2012; Miranda & Claes, 2009; North et al., 2000; Saarikallio & Erkkilä, 2007; Schäfer & Sedlmeier, 2009; Tarrant et al., 2000; ter Bogt, Mulder, Raaijmakers, & Nic Gabhainn, 2011). According to Lonsdale and North (2011), the emotional function of music is even more important than its identity and social functions. This is supported by Saarikallio and Erkkilä (2007), for whom mood management encompasses a wide variety of personal needs that can be related to identity, gender or age. Following on from this, studying motives for listening to music from the perspective of mood regulation can be a good way of better understanding the psychological functioning and mental health of adolescents (Miranda, 2013; Miranda & Claes, 2009; Ruud, 1997). One model that offers a comprehensive theory of how motives relate to behaviour is the Motivational Model of Alcohol Use (Cox & Klinger, 1988, 2004). Central to the model is goal pursuit that is initiated by the affective change that people expect from attaining the goal. In other words, individuals strive for things or circumstances that will make them feel better, by either giving them pleasure or relieving their discomfort. Thus, the pursuit of an affective change constitutes the basic condition for individuals to recognize the value of a potential goal. Generally, people are motivated either to acquire positive incentives to achieve a positive affective change, or they strive to reduce negative incentives that create discomfort (Klinger & Cox, 2004).

Thus, the first dimension in which motivation can be classified in general is the affective change that can either have a positive valence (i.e., to increase positive feelings) or a negative one (i.e., to decrease negative feelings). The second dimension is the source of the affective change that can either be internal (e.g. in respect to one’s own bodily sensations) or external (e.g. in respect to significant others). Crossing these two dimensions results in four broad motive categories (Table 1).

Table 1.

Classification of motives according to the valence and source of expected affective change.

		Valence
		Positive reinforcement	Negative reinforcement
Source	Internal	Enhancement motives	Coping motives
		(e.g. because it is fun)	(e.g. to improve bad mood)
	External	Social motives	Conformity motives
		(e.g. to enjoy a party)	(e.g. to fit in with a group)

Studies across various countries and age groups have confirmed the validity and usefulness of the Motivational Model to better understand alcohol use (e.g. Kuntsche, Stewart, & Cooper, 2008; Mazzardis, Vieno, Kuntsche, & Santinello, 2010; Németh, Kuntsche, Urbán, Farkas, & Demetrovics, 2011), and for other psychoactive substances besides alcohol use (Cooper, Kuntsche, Levitt, Barber, & Wolf, in press). Originally developed to better understand alcohol consumption, the Motivational Model has subsequently been transferred to other domains of human behaviour such as gambling (Stewart & Zack, 2008), sexual risk-taking behaviour (Cooper, Shapiro, & Powers, 1998) and internet use (Bischof-Kastner, Kuntsche, & Wolstein, 2014). We believe that this model could also be adapted to study motives for listening to music. Indeed, in their study, Lonsdale and North (2011) use two dimensions that are very similar to the ones in the Motivational Model to describe motives for listening to music, that is, to manage their mood, ‘people use music both as a means to cope with and alleviate negative feelings, as well as to create and optimize a positive mood’ (p. 111). Moreover, previous studies used motives for listening to music that are consistent with the four categories obtained by crossing the two dimensions (cf. Table 1). These are:

(A) enhancement motives in the sense of personal growth or well-being promotion such as enjoyment (Boal-Palheiros & Hargreaves, 2001; North et al., 2000, England, mean age 13), self-actualization (e.g. to create an image of the self and to express feelings: Tarrant et al., 2000, UK and US, mean age 15), achieving or optimizing positive moods and feelings (Chamorro-Premuzic et al., 2009, Spain, mean age 20; Lonsdale & North, 2011, UK, mean age 20; Schäfer & Sedlmeier, 2009, Germany, mean age 25);

(B) coping motives such as relieving daily stress, e.g. school pressure (Miranda & Claes, 2009), reducing loneliness (Tarrant et al., 2000), relieving tension or boredom (North et al., 2000; Tarrant et al., 2000), alleviating negative feelings (Chamorro-Premuzic et al., 2009; Lonsdale & North, 2011), relaxing (Schäfer & Sedlmeier, 2009), feeling better and improving mood (Saarikallio & Erkkilä, 2007, Finland, age 14 and 17) and getting through difficult times (North et al., 2000);

(C) social motives such as improving social relationships reported, for instance, among 9–14-year-old British and Portuguese students (Boal-Palheiros & Hargreaves, 2001) and strengthening and maintaining personal relationships (e.g. to talk to others: Laiho, 2004; Lonsdale & North, 2011);

(D) conformity motives such as fitting in with peers (Laiho, 2004; Miranda & Claes, 2009, French-speaking Canada, mean age 16), social needs (e.g. to be popular and to please friends) reported by American and British adolescents (Tarrant et al., 2000), creating an external impression (e.g. to appear trendy or cool: North et al., 2000).

The purpose of this study was to develop the four-dimensional Motives for Listening to Music Questionnaire (MLMQ) based on the Short Form of the Drinking Motive Questionnaire Revised (DMQ-R SF: Kuntsche & Kuntsche, 2009) that was developed based on Cox and Klinger’s Motivational Model (1988, 2004).

Two steps were undertaken to test construct validity (Allen & Yen, 2002), that is, the degree to which the MLMQ measures the theoretical construct that it was designed to measure. The first aim was to confirm the hypothesized four-dimensional structure of the MLMQ. We expected a good model fit of the four-dimensional structure of music motives (H1.1) and high factor loadings of the corresponding items (H1.2) to provide evidence for construct validity (Allen & Yen, 2002). We also expected the internal consistencies to be greater than .7 for each music motive dimension factor (H1.3) to demonstrate satisfactory consistency (Nunnally & Bernstein, 1994).

Second, we tested whether the four-dimensional factor structure of the MLMQ was invariant for different subgroups characterized by gender, age and linguistic region. This is important to provide indications as to what extent the scale can be used for cross-group comparisons. We expected to find no structural differences in the MLMQ between boys and girls (H2.1), younger and older adolescents (H2.2) or between the linguistic regions in Switzerland (H2.3).

Additionally, we tested mean differences in the four music motive scores in the subgroups. In accordance with previous research (Chamorro-Premuzic et al., 2009; Miranda & Claes, 2009; North et al., 2000; ter Bogt et al., 2011), we expected girls to be more likely to listen to music due to coping and enhancement motives than boys (H3.1) and boys to be more likely to listen to music for conformity motives than girls (H3.2).

Subsequently, we investigated concurrent validity of the MLMQ defined as the correlation between test and criterion scores when both are obtained at the same time (Allen & Yen, 2002). Thus, we tested links between the four music motive factors and different health issues (i.e., somatic complaints, depressed mood, aggressive mood, physical powerlessness, self-rated health and life satisfaction). We also investigated whether social aspects (i.e., frequency of being bullied, frequency of bullying others and evenings spent out with friends) are associated with music motives.

Based on previous findings (Lonsdale & North, 2011; Miranda & Claes, 2009; Saarikallio & Erkkilä, 2007; Siedliecki & Good, 2006), we expected that those who listen to music to cope with negative emotions are likely (H4.1a) to have more health problems (somatic complaints, depressed mood, aggressive mood and physical powerlessness), (H4.1b) to rate their health lower, (H4.1c) to have lower life satisfaction and (H4.1d) to perceive more school pressure than those who listen to music for other reasons.

We also expected that adolescents who listen to music due to social motives spend more evenings with friends (Lonsdale & North, 2011; Tarrant et al., 2000) (H4.2). According to the definition, adolescents who listen to music for conformity motives do so because they like to belong to a certain peer group and not to feel left out, that is, they are most likely afraid of or already a target of peer rejection. In turn, such social marginalization is related to emotional problems and victimization through bullying (Juvonen, Graham, & Schuster, 2003). Therefore, we explored the hypothesis of a positive relationship between conformity motives and the frequency of being bullied (H4.3a) and depressed mood (H4.3b).

Since various aspects of previous research (Boal-Palheiros & Hargreaves, 2001; Chamorro-Premuzic et al., 2009; Lonsdale & North, 2011; North et al., 2000; Schäfer & Sedlmeier, 2009; Tarrant et al., 2000), such as enjoyment, self-actualization, achieving or optimizing positive moods, can be theoretically subsumed under the dimension ‘enhancement motives,’ it was not possible to formulate any specific hypotheses for this dimension and the links to health and social issues therefore remain exploratory.

Methods

Study design

Data were taken from the Swiss participation in the international survey ‘Health and Behaviour in School-Aged Children’ (www.HBSC.org). Since 1982, this international survey has been conducted every 4 years and now comprises more than 40 countries (predominantly in Europe). Adolescents attending state schools in Switzerland ranging from the fifth to the ninth grade participated in this study. Classes were randomly selected within a random cluster sampling design, which was proportionate to the size of the participating cantons (federal states in Switzerland) with a response rate of 88%. Permission for the HBSC survey to be conducted was granted by the educational authorities of each participating canton. In addition, before data were collected, permission was granted by the school principals of the randomly-selected school classes. For the purpose of data collection, students completed the questionnaire (taking approximately 45 minutes) independently and on a voluntary basis between January and April 2010 in a classroom setting. To guarantee anonymity and privacy, names were not written on the questionnaires, and after completion, students were asked to put their questionnaire in an unmarked envelope and seal it. The Human Research Ethics Committee (Canton of Vaud Protocol no. 173/09) approved the study.

Sample and missing values

The sampling was based on comprehensive lists of Swiss schools from fifth to ninth grade compiled by the Swiss Federal Statistical Office. For comprehension and ethical reasons, the fifth- to seventh-graders received a shortened version of the questionnaire in which music motives were not included. Thus, the original sample contained 4,644 eighth- and ninth-grade adolescents from state schools in all regions of Switzerland (i.e., German-, French- and Italian-speaking). Participants who answered only three or fewer out of 12 questions on music motives (N = 77, 1.6 %) were excluded from the analyses. Missing values on gender (N = 20, 0.4 %) and age (N = 28, 0.8%) were also excluded. Remaining missing values (usually below 1.5% except 2.4% for the variable ‘self-rated health’) were taken into account using the full information maximum likelihood (FIML) option of Mplus software (Muthén & Muthén, 1998–2010) when estimating confirmatory factor analyses or regression analyses. For bivariate analyses (t tests, correlations etc.) remaining missing values were excluded on a case by case basis. The analysed data consists of 4,524 adolescents (mean age = 14.5, SD = 0.9; age range 12–18 years; 48.9% boys). The majority of the adolescents came from the German-speaking region of Switzerland (69.8%), with 25.6% from the French-speaking region and 4.6% from the Italian-speaking region.

Measures

Motives for listening to music

To develop the motives for listening to music questionnaire (MLMQ), 12 motive items were adapted from the DMQ-R SF (Kuntsche & Kuntsche, 2009). The scale was constructed on the basis of the four-factor model of drinking motives (Cooper, 1994) with the dimensions Coping, Enhancement, Social and Conformity. The wording of the introductory sentence was: ‘How often do you listen to music for the following reasons?’ The formulation of all items can be found in Table 1. The different motives for listening to music were rated on a 5-point relative frequency scale. The answer categories were: ‘almost never/never’ (coded as 1), ‘some of the time’ (coded as 2), ‘half of the time’ (coded as 3), ‘most of the time’ (coded as 4), and ‘almost always/always’ (coded as 5).

Life satisfaction

The Cantrill ladder (Cantrill, 1965) was given to the adolescents with the following instruction: ‘Here is a picture of a ladder. The top of the ladder “10” is the best possible life for you and the bottom “0” the worst possible life for you. In general, where on the ladder do you feel you stand at the moment?’ The adolescents then had to rate their own perceived life satisfaction on a 10-point scale ranging from ‘worst possible life’ (coded as 0) to ‘best possible life’ (coded as 10).

School pressure

To assess school pressure, an item was developed to measure the global feeling of school-associated pressure due to the demands of schoolwork including homework (Danielsen, Samdal, Hetland, & Wold, 2009; Ravens-Sieberer et al., 2008). Following the question ‘How pressured do you feel by the schoolwork you have to do?’ adolescents could indicate whether they did not experience any school pressure (answer category ‘not at all,’ coded as 1) or if they had experienced school pressure ‘a lot’ (coded as 4). Other answer categories were ‘a little’ (coded as 2), ‘some’ (coded as 3).

Self-rated health

To measure self-perceived health status, an item from epidemiology (Idler & Benyamini, 1997) was included. The question ‘Would you say your health is…’ was given to the adolescents. They had to rate their health on a 4-point scale with the answer categories ‘poor’ (coded as 1), ‘fair’ (coded as 2), ‘good’ (coded as 3), and ‘excellent’ (coded as 4).

Health complaints

The Kidscreen Symptom Checklist (SCL: Ravens-Sieberer et al., 2008) that measures physical and psychological symptoms in non-clinical samples was used. The adolescents were asked: ‘In the last 6 months: how often have you had the following…?’ Subsequently, 12 items had to be rated on a 5-point scale that was coded to represent monthly frequencies: ‘about every day’ (coded as 30), ‘more than once a week’ (coded as 9), ‘about every week’ (coded as 4.5), ‘about every month’ (coded as 1), and ‘hardly ever’ (coded as 0). We split the scale into four subscales to measure (1) somatic complaints comprising ‘headache,’ ‘stomach ache,’ and ‘backache’ (α = .58), (2) aggressive mood comprising ‘irritability and bad temper’ and ‘felt annoyed and angry’ (α = .76), (3) depressed mood comprising ‘feeling low,’ ‘feeling nervous,’ ‘feeling anxious or worried’ (α = .72), and (4) physical powerlessness comprising ‘difficulties in getting to sleep,’ ‘feeling dizzy,’ and ‘feeling tired’ (α = .55).

Frequency of being bullied and bullying others

Two questions assessed the frequency of bullying from the victim and the offender perspective (Olweus, 1993). The victim question was: ‘How often have you been bullied at school in the past couple of months?’ The offender question was: ‘How often have you taken part in bullying another student(s) at school in the past couple of months?’ Adolescents answered on a 5-point scale that was coded to represent frequencies in the last two months: ‘I have not been bullied / I have not bullied another student at school in the past couple of months’ (coded as 0), ‘It has only happened once or twice’ (coded as 1.5), ‘2 or 3 times a month’ (coded as 5), ‘About once a week’ (coded as 9) and ‘Several times a week’ (coded as 18).

Evenings spent with friends

The question was: ‘How many evenings per week do you usually spend out with your friends?’ On a 7-point scale ranging from ‘0 evenings’ (coded as 0) to ‘7 evenings’ (coded as 7), the adolescents were asked to indicate the corresponding number of evenings spent with friends.

Statistical analysis

First, to confirm construct validity, we used confirmatory factor analysis to test the hypothesized four-factor structure of the Motives for Listening to Music Questionnaire (MLMQ). Residual item variance was only allowed to correlate in the event of high similarity in item formulation (i.e., between the items ‘to be liked’ and ‘so you won’t feel left out’ and between ‘because it helps you to enjoy a party’ and ‘because it improves parties and celebrations’). To evaluate model fit, we used the CFI (Comparative Fit Index), the TLI (Trucker Lewis Index), the RMSEA (Root Mean Square Error of Approximation) and the SRMR (Standardized Root Mean Square Residual). CFI and TLI relate to the total variance accounted for by the model, where values close to 1, i.e., higher than .9, were sought (Kline, 2005). SRMR and RMSEA relate to the residual variance, where values close to 0, i.e., lower than .1, were sought (Kline, 2005). The internal consistencies for each music dimension were assessed using Cronbach’s Alpha in the software SPSS16. Values greater or equal to .9, .8, .7, .6, and .5 were considered as excellent, good, acceptable, questionable, and poor, respectively (Nunnally & Bernstein, 1994).

Second, apart from validating the four-dimensional structure of music motives in general, we tested whether it was also valid for different subgroups. Therefore, nested models of confirmatory factor analyses with increasing constraints were estimated for gender, age (median split: 12–14-year-olds versus 15–18-year-olds) and linguistic region (German speaking (69.8%) versus French or Italian speaking (30.2%). We tested two forms of measurement invariance: the configural and metric invariance (e.g. Steenkamp & Baumgartner, 1998) for each grouping variable (gender, age group and linguistic region). Configural invariance is supported if the unconstrained model has an acceptable fit to the data. Metric invariance requires that the factor loadings are equivalent between the groups (λ-constrained model). Subsequently, we investigated whether the fit indices remained the same when, first, variances and, second, factor loadings, variances and correlations were constrained to equivalence between the groups. Due to the large sample size, the chi-square differences may have been biased, whereby even very small differences across groups can become significant (Meredith, 1993). We therefore decided not to perform chi-square difference tests. We examined the values of the CFI, the TLI, the RMSEA, and the SRMR in the unconstrained and the constrained model instead. Regarding confirming the metric invariance, the model fit between the unrestricted models and models with increasing restrictions should not substantially change, i.e., CFI and TLI values should be close to 1 (higher than .90, preferably close to .95), and SRMR and RMSEA values close to 0 (lower than .10, preferably close to .08) even in the most restrictive model (Hooper, Coughlan, & Mullen, 2008; Kline, 2005; Marsh, Hau, & Wen, 2004).

Third, to test mean difference in the four music motive factors between boys and girls, younger and older adolescents and those from the German-speaking and the French/Italian-speaking parts of Switzerland, we conducted independent sample t tests in SPSS16. For this purpose, we computed summary scores for the four motive factors. To counteract artificial enhancement of test power due to the cluster sampling of school classes (i.e., students in the same class tend to behave in a similar way resulting in artificially-small standard errors), we down-weighted the sample by 0.83 corresponding to an average sample design effect of 1.2 (Roberts et al., 2007).

Fourth, to test concurrent validity of the scale, we estimated a multivariate structural equation model including the four music dimensions as independent variables, gender and age as control variables. Dependent variables were: life satisfaction, school pressure, self-rated health and somatic complaints, depressed mood, aggressive mood, physical powerlessness, frequency of being bullied, frequency of bullying others and evenings spent with friends (a correlation matrix and descriptive statistics of the variables included in the model is given in the Appendix). Unless stated otherwise, all analyses were conducted with Mplus 6 software (Muthén & Muthén, 1998–2010) by using the MLR (Maximum Likelihood Robust) estimator to account for deviations from normal distribution and for the dependency of observations (i.e., adolescents nested within school classes).

Results

Confirming the four-factor structure

Confirming H1.1 and H1.2, the confirmatory factor analysis revealed highly significant factor loadings ranging from λ = .55 to λ = .90 (Table 2). The Coping factor had the highest item loadings (ranging from λ = .69 to λ = .90), followed by the Conformity factor (λ = .64 to λ = .85). Consequently, these two factors also had the highest internal consistency (i.e., values above .8; confirming H1.3). The Enhancement factor had the lowest loadings (λ = .55 to λ = .73) and also the lowest internal consistency. The enhancement, social and coping items revealed mean values of around 3, whereas conformity mean values were much lower (i.e., ranging from 1.01 to 1.25). The four factors were also highly correlated. The highest correlations were found between social and enhancement (r = .78) and the lowest between coping and conformity (r = .20). There was a good fit to the data as demonstrated by CFI and TLI values above .93 and RMSEA and SRMR values below .07 (Table 2).

Table 2.

Item factor loadings, item means, inter-factor correlations, and internal consistencies as results of the confirmatory factor analysis to test the four-factor structure of music motives.

How often do you listen to music…	Enhancement	Coping	Social	Conformity	Means (SD)
because you like the feeling?	.73				3.52 (1.46)
to get high?	.55				1.70 (1.24)
because it is fun?	.58				4.20 (1.15)
because it helps you when you feel depressed or nervous?		.89			3.39 (1.47)
to cheer you up when you are in a bad mood?		.90			3.55 (1.39)
to forget about your problems?		.69			2.78 (1.54)
because it helps you to enjoy a party?			.61		3.50 (1.54)
because it makes social gatherings more fun?			.80		2.84 (1.50)
because it improves parties and celebrations?			.66		3.37 (1.60)
to be liked?				.70	1.25 (0.77)
to fit in with a group you like?				.85	1.30 (0.90)
so you won’t feel left out?				.64	1.25 (0.80)

Correlation with the factor ‘Coping’	.67
Correlation with the factor ‘Social’	.78	.50
Correlation with the factor ‘Conformity’	.33	.20	.33

Internal consistencies (Cronbach’s α)	.61	.86	.82	.83

Note. The figures shown are standardized factor loadings that are all significant at the 0.01% error level; model fit: CFI =. 953; TLI =. 931; RMSEA =. 064; SRMR = .049.

The four-factor structure in different subgroups

To test construct validity of the four-factor model according to gender, age group and linguistic region, nested models of confirmatory factor analyses with increasing degrees of freedom were conducted for each of the subgroups. Confirming H2.1-3, configural invariance was supported for all subgroups, i.e., in all unconstrained baseline models, CFI and TLI values were above .9 and RMSEA and SRMR values below .08 (Table 3). Metric invariance was also supported for all subgroups due to a minor difference in model fit from the unconstrained to the λ-constrained models, i.e., the factor loadings did not considerably differ between the subgroups. Subsequently, we tested whether the fit indices changed when variances were additionally constrained to equivalence between the subgroups (third model) and when factor loadings, correlations and variances were constrained (fourth model). However, in all cases, all four fit indices remained basically the same (Table 3). Sometimes, there was even a slightly better fit with increasing constraints, which was due to a higher gain in degrees of freedom in comparison to a modest increase in chi-square. Even in the most restrictive models, CFI and TLI values were above .9 and RMSEA and SRMR values were below .08 (with the exception of the gender restricted models, for which the SRMR values were only slightly higher than this threshold).

Table 3.

Model fit according to gender, age and linguistic region in models with increasing constraints.

	χ²	df	CFI	TLI	RMSEA	SRMR
Gender (boys vs. girls)
Unconstrained model	927.8	92	.951	.929	.063	.049
λ constrained model	1036.5	106	.945	.931	.062	.052
Plus variance constrained	1188.3	110	.936	.923	.066	.094
Plus correlations constrained	1312.3	116	.935	.926	.065	.095
Age groups (12–14 vs. 15–18)
Unconstrained model	923.1	92	.953	.933	.063	.050
λ constrained model	993.2	106	.950	.938	.061	.052
Plus variance constrained	982.0	110	.951	.941	.059	.060
Plus correlations constrained	1000.0	116	.950	.943	.058	.060
Linguistic region (Ger. vs. FR/IT)
Unconstrained model	945.8	92	.955	.935	.064	.053
λ constrained model	1271.7	106	.938	.923	.070	.061
Plus variance constrained	1210.9	110	.942	.930	.067	.065
Plus correlations constrained	1263.9	116	.939	.931	.066	.067

Note. Due to maximum likelihood robust estimation, the chi-square values are not strictly comparable to each other.

Testing mean differences between the subgroups

The results in Table 4 show that female adolescents scored higher on the enhancement, coping and social motives, but lower on conformity motives than male adolescents (confirming H3.1 and H3.2). Similar results were found for the age groups, i.e., higher mean scores for older adolescents on enhancement, coping and social motives compared to younger adolescents. Only the means for conformity motives did not significantly differ between the two age groups. Concerning linguistic regions, German-speaking adolescents scored higher on the enhancement motives than their French- and Italian-speaking counterparts. There was no significant difference for the other three motive factors.

Table 4.

Means (Standard Deviations in brackets) of the four music motive dimensions according to gender, age and linguistic region and independent sample t tests.

	Enhancement	Coping	Social	Conformity
Gender
Boys	3.04 (1.05)	2.79 (1.31)	3.10 (1.36)	1.33 (0.80)
Girls	3.25 (0.91)	3.67 (1.13)	3.37 (1.24)	1.28 (0.59)
t value	6.74***	22.0***	6.4***	−5.4***
Age
12–14-year-olds	3.04 (0.99)	3.14 (1.32)	3.05 (1.34)	1.26 (0.67)
15–18-year-olds	3.26 (0.97)	3.33 (1.27)	3.42 (1.25)	1.28 (0.74)
t value	6.7***	4.6***	8.8***	1.1
Linguistic region
German	3.27 (0.95)	3.26 (1.27)	3.26 (1.34)	1.26 (0.71)
French/Italian	2.86 (1.02)	3.17 (1.35)	3.19 (1.23)	1.29 (0.70)
t value	−11.6***	−1.9	−1.6	1.2

Note. * p < .05; ** p < .01; *** p < .001; t tests were performed based on the down-weighted sample (see Statistical Analysis section for further information).

Concurrent validity of the MLMQ

Results from the structural equation model revealed that the more frequently participants listened to music for enhancement purposes, the higher their life satisfaction was and the fewer evenings they spent out with friends (Table 5). Confirming H4.1a–d, coping motives were significantly related to all health issues, i.e., those who frequently listened to music to cope had more health complaints, were less satisfied with their lives and experienced a higher level of school pressure than those listening for other motives. In addition, coping motives showed a significant positive relationship with the frequency of being bullied. Confirming H4.2, social motives revealed a particularly strong relationship between social aspects, i.e., the more frequently participants listened to music for social purposes, the more evenings they spent out with friends and the more often they took part in bullying others but the less likely they were to be a victim of bullying. In addition, social motives showed a significant positive relationship to self-rated health. For conformity motives (H4.3a–b), there was a significant positive link to somatic complaints and depressed mood. Adolescents who listened to music for conformity motives were likely to rate their health as good or excellent but indicated a low life satisfaction and reported being victims of bullying. The explained variance of the outcome variables varied between 2% and 18%.

Table 5.

Health and social aspects regressed on music motive, gender, and age (standardized regression coefficients and explained variance).

	Enhancement	Coping	Social	Conformity	Gender	Age	R ²
Health issues
Somatic complaints	−.09	.16***	.07	.06*	.10***	.05**	5.0%
Aggressive mood	−.02	.20***	.02	.04	.03	.01	4.7%
Depressed mood	−.10	.25***	−.02	.14***	.10***	−.00	8.0%
Physical powerlessness	−.01	.21***	−.02	.04	.08***	.01	5.6%
Self-rated health	−.04	−.13***	.13**	.05*	−.08***	−.07***	3.3%
Life satisfaction	.17**	−.25***	−.02	−.06*	−.03	−.02	4.2%
School pressure	−.06	.13***	.06	.04	−.01	−.02	1.9%
Social issues
Frequency of being bullied	−.08	.16***	−.12*	.07**	−.07***	−.03	2.7%
Frequency of bullying others	.05	−.04	.20***	.05	−.23***	.00	10.2%
Evenings spent with friends	−.14*	−.01	.50***	.02	−.14***	.09***	17.8%

Note. *p < .05; **p < .01; ***p < .001; model fit: CFI = .954; TLI = .910; RMSEA = .045; SRMR = .033.

Discussion

The purpose of this study was to develop the four-dimensional Motives for Listening to Music Questionnaire (MLMQ) and to test associations with health and social issues. The first aim was to test construct validity. Results from the confirmatory factor analysis revealed a good fit for the hypothesized four-factor model for music motives in our nationally representative sample of adolescents in Switzerland (H1.1) and high factor loadings of the corresponding items (H1.2). Interestingly, the highest inter-factor correlation was found between social and enhancement, i.e., those who listen to music because it is fun and because they like the feeling tend to do so to improve and enhance their enjoyment of parties and celebrations, which is consistent with findings for drinking motives (Cooper, 1994; Kuntsche, Knibbe, Gmel, & Engels, 2005; Kuntsche & Kuntsche, 2009).

Generally, high internal consistencies were found (Nunnally & Bernstein, 1994) especially considering that only three items were used to measure each factor (H1.3). Only for enhancement motives was the Cronbach’s Alpha value slightly below the threshold of .7. The confirmatory factor analysis in the subgroups demonstrated the invariance of the four-factor structure of the MLMQ across gender (H2.1), age (H2.2) and linguistic region (H2.3). Analysing mean differences across subgroups revealed that, in line with previous research (Chamorro-Premuzic et al., 2009; Miranda & Claes, 2009; North et al., 2000; ter Bogt et al., 2011), girls listened to music for enhancement motives, and to cope with unpleasant emotions more frequently than their male counterparts (H3.1). In addition, girls scored higher on social motives than boys but not higher on enhancement motives, which partially contradicts our hypothesis H3.1. However, confirming hypothesis H3.2, boys listened to music for conformity motives more frequently than girls.

Concerning concurrent validity of the MLMQ, results from the structural equation model confirmed hypothesis H4.1a–d that coping motives were not only related to a wide range of health issues (e.g. Lonsdale & North, 2011; Saarikallio & Erkkilä, 2007; Siedliecki & Good, 2006) but also to school pressure (e.g. Miranda & Claes, 2009) and being a victim of bullying. Listening to music appears to be a way for adolescents to compensate negative experiences, to cope with problems and complaints and to alleviate negative moods and feelings. As expected (e.g. Lonsdale & North, 2011; Tarrant et al., 2000), social motives were related to the frequency of peer-related activities such as spending evenings out with friends (H4.2) and bullying others, which is per definition also an activity together with peers (Olweus, 1993). Adolescents who are victims of bullying, by contrast, were expected to listen to music in order to gain access to a (new) peer group. Confirming hypotheses H4.3a–b, the results showed that conformity motives were related to victimization and bullying and also to depressed mood.

Although we did not have any specific hypotheses, the results show that those who listen for enhancement motives had a higher life satisfaction but did not spend many evenings with friends. This is consistent with the definition of enhancement motives (Cox & Klinger, 1988, 2004) and increasing positive mood internally, i.e., not in a social context (cf. Table 1).

Limitations and recommendations for future research

With the lowest factor loading and a lower mean than the other two enhancement items, it appears that the item ‘to get high’ relates more to psychoactive changes (e.g. resulting from alcohol use) and might be too strongly formulated to capture motives for listening to music. Moreover, other important aspects of listening to music, such as help constructing own identity (Arnett, 1995), expressing personal values (Schäfer & Sedlmeier, 2009), bringing back memories of events, emotions, relationships, life stages, and beloved persons (Boer & Fischer, 2012), that are beyond the four general motivational categories obtained by crossing the dimensions valence (positive vs. negative) and source (internal vs. external), were not included in this 12-item instrument.

Having established concurrent validity of the MLMQ based on cross-sectional data, longitudinal studies are needed to establish predictive validity of the MLMQ, i.e., whether the MLMQ predicts behaviour of adolescents in the future. Such replications will not only establish test-retest validity of the MLMQ but also enable testing of the causal relationship between motives for listening to music and health and social issues like somatic complaints, depressed mood, aggressive mood, physical powerlessness, self-rated health and life satisfaction or bullying. Although this study used a nationally representative sample and confirmed the invariance of the factor structure indicating the usefulness of the MLMQ, e.g. for different age groups, it remains unclear whether the MLMQ can also be used across different countries or for adult populations. Cross-cultural research is needed to investigate whether music motives differ across adolescents from different continents and cultural traditions. Moreover, we used a non-clinical sample to develop and validate the MLMQ. Thus, it remains unclear to what degree the MLMQ can be applied to a clinical sample.

Theoretical and practical implications

Having found links between the four music motive factors and different health and social aspects, the validation of the MLMQ demonstrates that consideration of the Motivational Model of Alcohol Use (Cox & Klinger, 1988, 1990, 2011) can be successfully transferred to another domain of human functioning (cf. also Cooper et al., 1998; Stewart & Zack, 2008). The MLMQ is concise and easy to administer and has proven to be construct-valid and reliable.

Including the MLMQ or other instruments assessing motives for listening to music in future research appears to be a promising approach, for example to shed further light on the finding that listening to music is linked to both positive and negative aspects of adolescent functioning (McFerran, Garrido, & Saarikallio, 2013; McFerran, Roberts, & O’Grady, 2010; Miranda & Claes, 2009). For health care and therapy, it may also be the case that using music in medical intervention or treatment, e.g. to decrease anxiety in patients (Gold, Voracek, & Wigram, 2004; McCaffre & Good, 2000; McFerran et al., 2010; Pacchetti et al., 2000; Wang, Kulkarni, Dolev, & Kain, 2002) would be even more effective when the kind of affective change that adolescents expect to obtain by listening to music (e.g. measured by the MLMQ) is taken into account. This is particularly promising since the motivation for listening to music was found to explain more variance in well-being than the mere behaviour of listening to music (Morinville, Miranda, & Gaudreau, 2013). However, clearly more research is needed to better understand the potential links between music motives and the effects of music therapy.

Footnotes

Acknowledgements

The authors would like to thank Marina Delgrande Jordan, Edith Bacher, Christiane Gmel, Aurélie Archimi and Béat Windlin for their contributions to the 2010 Swiss HBSC study and Gemma Brown for the English copy editing.

Funding

This research was primarily funded by the Swiss Federal Office of Public Health (grant no. 09.000925).

References

Allen

M. J.

Yen

W. M.

(2002). Introduction to measurement theory. Prospect Heights, IL: Waveland Press.

Arnett

J. J.

(1995). Adolescents’ uses of media for self-socialization. Journal of Youth and Adolescence, 24(5), 519–533.

Bischof-Kastner

Kuntsche

Wolstein

(2014). Development and validation of a four-dimensional Internet Motive Questionnaire for Adolescents (IMQ-A). Journal of Medical Internet Research, 16(10), e230.

Boal-Palheiros

G. M.

Hargreaves

D. J.

(2001). Listening to music at home and at school. British Journal of Music Education, 18(2), 103–118.

Boer

Fischer

(2012). Towards a holistic model of functions of music listening across cultures: A culturally decentred qualitative approach. Psychology of Music, 40(2), 179–200.

Cantrill

(1965). The pattern of human concern. New Brunswick, NJ: Rutgers University Press.

Chamorro-Premuzic

Gomà-i-Freixanet

Furnham

Muro

(2009). Personality, self-estimated intelligence, and uses of music: A Spanish replication and extension using structural equation modeling. Psychology of Aesthetics, Creativity, and the Arts, 3, 149–155.

Cooper

M. L.

(1994). Motivations for alcohol use among adolescents: Development and validation of a four-factor-model. Psychological Assessment, 6(2), 117–128.

Cooper

M. L.

Kuntsche

Levitt

Barber

Wolf

(in press). A motivational perspective on substance use: Review of theory and research. In Sher

K. J.

(Ed.), Oxford handbook of substance use disorders. New York, NY: Oxford University Press.

10.

Cooper

M. L.

Shapiro

C. M.

Powers

A. M.

(1998). Motivations for sex and risky sexual behavior among adolescents and young adults: A functional perspective. Journal of Personality and Social Psychology, 75(6), 1528–1558.

11.

Cox

W. M.

Klinger

(1988). A motivational model of alcohol use. Journal of Abnormal Psychology, 97(2), 168–180.

12.

Cox

W. M.

Klinger

(1990). Incentive motivation, affective change, and alcohol use: A model. In Cox

W. M.

(Ed.), Why people drink – Parameters of alcohol as a reinforcer (pp. 291–314). New York, NY: Gardner Press, Inc.

13.

Cox

W. M.

Klinger

(2004). A motivational model of alcohol use: Determinants of use and change. In Cox

W. M.

Klinger

(Eds.), Handbook of motivational counseling: Concepts, approaches, and assessment (pp. 121–138). Chichester, UK: John Wiley & Sons, Ltd.

14.

Cox

W. M.

Klinger

(2011). Handbook of motivational counseling. Goal-based approaches to assessment and intervention with addiction and other problems. West Sussex, UK: John Wiley & Sons, Ltd.

15.

Danielsen

A. G.

Samdal

Hetland

Wold

(2009). School-related social support and students’ perceived life satisfaction. Journal of Educational Research, 102, 303–318.

16.

Gold

Voracek

Wigram

(2004). Effects of music therapy for children and adolescents with psychopathology: A meta-analysis. Journal of Child Psychology and Psychiatry, 45(6), 1054–1063.

17.

Hooper

Coughlan

Mullen

M. R.

(2008). Structural equation modelling: Guidelines for determining model fit. Electronic Journal of Business Research Methods, 6(1), 53–60.

18.

Idler

E. L.

Benyamini

(1997). Self-rated health and mortality: A review of twenty-seven community studies. Journal of Health and Social Behavior, 38(1), 21–37.

19.

Juvonen

Graham

Schuster

M. A.

(2003). Bullying among young adolescents: The strong, the weak and the troubled. Pediatrics, 112(6), 1231–1237.

20.

Kline

R. B.

(Ed.). (2005). Principles and practice of structural equation modeling (2nd ed.). (Methodology in social sciences). New York, NY: The Guilford Press.

21.

Klinger

Cox

W. M.

(2004). Motivation and the theory of current concerns In Cox

W. M.

Klinger

(Eds.), Handbook of motivational counseling (pp. 3–27). Chichester, UK: John Wiley & Sons.

22.

Kuntsche

Knibbe

R. A.

Gmel

Engels

R. C. M. E.

(2005). Why do young people drink? A review of drinking motives. Clinical Psychology Review, 25(7), 841–861.

23.

Kuntsche

(2009). Development and validation of the Drinking Motive Questionnaire Revised Short Form (DMQ–R SF). Journal of Clinical Child and Adolescent Psychology, 38(6), 899–908.

24.

Kuntsche

Stewart

S. H.

Cooper

M. L.

(2008). How stable is the motive-alcohol use link? A cross-national validation of the Drinking Motive Questionnaire Revised among adolescents from Switzerland, Canada, and the United States. Journal of Studies on Alcohol and Drugs, 69(3), 388–396.

25.

Laiho

(2004). The psychological functions of music in adolescence. Nordic Journal of Music Therapy, 13(1), 47–63.

26.

Lonsdale

A. J.

North

A. C.

(2011). Why do we listen to music? A uses and gratifications analysis. British Journal of Psychology, 102, 108–134.

27.

Marsh

H. W.

Hau

K.-T.

Wen

(2004). In search of golden rules: Comment on hypothesis-testing approaches to setting cutoff values for fit indexes and dangers in overgeneralizing Hu and Bentler’s (1999) findings. Structural Equation Modeling: A Multidisciplinary Journal, 11(3), 320–341.

28.

Mazzardis

Vieno

Kuntsche

Santinello

(2010). Italian validation of the Drinking Motive Questionnaire Revised Short Form (DMQ-R SF). Addictive Behaviors, 35(10), 209–305.

29.

McCaffre

R. G.

Good

(2000). The lived experience of listening to music while recovering from surgery. Journal of Holistic Nursing, 18(4), 378–390.

30.

McFerran

Garrido

Saarikallio

(2013). A critical interpretive synthesis of the literature linking music and adolescent mental health. Youth & Society. Advance online publication. doi:0044118X13501343

31.

McFerran

Roberts

O’Grady

(2010). Music therapy with bereaved teenagers: A mixed methods perspective. Death Studies, 34(6), 541–565.

32.

Meredith

(1993). Measurement of invariance, factor analysis and factorial invariance. Psychometrika, 58(4), 525–543.

33.

Miranda

(2013). The role of music in adolescent development: Much more than the same old song. International Journal of Adolescence and Youth, 18(1), 5–22.

34.

Miranda

Claes

(2009). Music listening, coping, peer affiliation and depression in adolescence. Psychology of Music, 37(2), 215–233.

35.

Morinville

Miranda

Gaudreau

(2013). Music listening motivation is associated with global happiness in Canadian late adolescents. Psychology of Aesthetics, Creativity, and the Arts, 7(4), 384–390.

36.

Muthén

L. K.

Muthén

B. O.

(1998–2010). Mplus User’s Guide (6th ed.). Los Angeles, CA: Muthén & Muthén.

37.

Németh

Kuntsche

Urbán

Farkas

Demetrovics

(2011). Why do festival goers drink? Assessment of drinking motives using the DMQ-R SF in a recreational setting. Drug and Alcohol Review, 30(1), 40–46.

38.

North

A. C.

Hargreaves

D. J.

O’Neill

S. A.

(2000). The importance of music to adolescents. British Journal of Educational Psychology, 70(2), 255–272.

39.

Nunnally

J. C.

Bernstein

I. H.

(1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill.

40.

Olweus

(1993). Bullying at school: What we know and what we can do. Oxford, UK: Blackwell.

41.

Pacchetti

Mancini

Aglieri

Fundarò

Martignoni

Nappi

(2000). Active music therapy in Parkinson’s disease: An integrative method for motor and emotional rehabilitation. Psychosomatic Medicine, 62(3), 386–393.

42.

Ravens-Sieberer

Gosch

Rajmil

Erhart

Bruil

Power

, . . . KIDSCREEN Group. (2008). The KIDSCREEN-52 quality of life measure for children and adolescents: Psychometric results from a cross-cultural survey in 13 European countries. Value in Health, 11(4), 645–658.

43.

Roberts

Currie

Samdal

Currie

Smith

Maes

(2007). Measuring the health and health behaviours of adolescents through cross-national survey research: Recent developments in the Health Behaviour in School-aged Children (HBSC) study. Journal of Public Health, 15(3), 179–186.

44.

Ruud

(1997). Music and the quality of life. Nordic Journal of Music Therapy, 6(2), 86–97.

45.

Saarikallio

Erkkilä

(2007). The role of music in adolescents’ mood regulation. Psychology of Music, 35(1), 88–109.

46.

Schäfer

Sedlmeier

(2009). From the functions of music to music preference. Psychology of Music, 37(3), 279–300.

47.

Siedliecki

S. L.

Good

(2006). Effect of music on power, pain, depression and disability. Journal of Advanced Nursing, 54(5), 553–562.

48.

Steenkamp

J.-B. E. M.

Baumgartner

(1998). Assessing measurement invariance in cross-national consumer research. Journal of Consumer Research, 25, 78–90.

49.

Stewart

S. H.

Zack

(2008). Development and psychometric evaluation of a three-dimensional gambling motives questionnaire. Addiction, 103(7), 1110–1117.

50.

Tarrant

North

A. C.

Hargreaves

D. J.

(2000). English and American adolescents’ reasons for listening to music. Psychology of Music and Education, 28(2), 166–173.

51.

ter Bogt

T. F. M.

Mulder

Raaijmakers

Q. A. W.

Nic Gabhainn

(2011). Moved by music: A typology of music listeners. Psychology of Music, 39(2), 147–163.

52.

Thoma

M. V.

Scholz

Ehlert

Nater

U. M.

(2012). Listening to music and physiological and psychological functioning: The mediating role of emotion regulation and stress reactivity. Psychology & Health, 27(2), 227–241.

53.

Wang

S.-M.

Kulkarni

Dolev

Kain

Z. N.

(2002). Music and preoperative anxiety: A randomized, controlled study. Anesthesia and Analgesia, 94(6), 1489–1494.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB