Abstract
Recent research has found that neuroticism (i.e., trait emotional instability) may dispose people to use music listening as a strategy to regulate their emotions. To estimate the magnitude of this relationship, we performed a meta-analysis (random effects model) of the extant 13 correlational studies (k = 13) for a total of 2641 participants. Results indicated a significant small-to-medium summary effect (r =.22, 95% CI [0.17, 0.27]) for the positive correlation between neuroticism and emotion regulation through music listening. Furthermore, there was no evidence of significant heterogeneity in effect sizes across studies. Overall, we conclude that the putative effect of neuroticism on musical emotion regulation is relatively moderate. Findings may suggest that people higher in neuroticism are more prone to use music listening as an accessible resource to regulate their negative emotions or manage whatever affects their mood in everyday life.
Introduction
In the last 20 years, there have been several discoveries and innovations in research on personality and music (e.g., Bonneville-Roussy, Rentfrow, Xu, & Potter, 2013; Chamorro-Premuzic & Furnham, 2007; Delsing, ter Bogt, Engels, & Meeus, 2008; Rentfrow & Gosling, 2003). Recently, research has found that neuroticism (i.e., trait emotional instability) may predict the use of music listening as a strategy to regulate emotions (e.g., Chamorro-Premuzic & Furnham, 2007). The initial theorization behind this finding was that because people with higher neuroticism are more negative and unstable emotionally they might thereby be more sensitive to the emotional effects of music and perhaps even more receptive to the emotional uses of music (Chamorro-Premuzic & Furnham, 2007). However, the magnitude of the association between neuroticism and the use of music listening for emotion regulation (e.g., small, moderate or large effect size) is still an open question. Therefore, to answer this question, our objective was to conduct the first meta-analysis of studies examining the relationship between neuroticism and emotion regulation through music.
Neuroticism within personality
There is substantial agreement that personality traits can be aptly described by a Five-Factor Model of personality also known as the ‘Big Five’ (i.e., extraversion, agreeableness, conscientiousness, neuroticism, and openness), which is essentially assessed through self-reports (John, Naumann, & Soto, 2008). These five broad traits capture relatively stable individual differences that develop gradually from childhood to late adulthood (Roberts, Walton, & Viechtbauer, 2006). These five traits are also widely and primarily studied by researchers as dispositions notably because they have been shown to consistently predict meaningful life outcomes (Ozer & Benet-Martínez, 2006).
More specifically, within the Big Five, higher neuroticism is more maladaptive because it involves having a continual negative and disproportionate emotional response to everyday challenges and adversities, which makes the environment seem more threatening and distressing (Caspi, Roberts, & Shiner, 2005; Lahey, 2009). For example, some of its main specific traits include depression and anxiety, but also hostility, impulsiveness, self-consciousness, and vulnerability (McCrae, Costa, & Martin, 2005; Soto & John, 2009). As such, neuroticism is a risk factor given that it may increase susceptibility to develop different mental health problems (Kotov, Gamez, Schmidt, & Watson, 2010).
Emotion regulation through music listening
Music is a form of communication that is inherently about conveying and eliciting emotions in listeners (Juslin & Sloboda, 2010). Music may, as such, represent a vast and sustainable resource to regulate emotions in many moments of life. More specifically, emotion regulation through music occurs when people engage in music listening to create, change, or maintain positive and negative emotions (Baltazar & Saarikallio, 2016). Like for the Big Five, emotion regulation is typically assessed through self-reports whether in music psychology (e.g., Chamorro-Premuzic & Furnham, 2007) or in psychology in general (e.g., Gross & John, 2003).
Emotion regulation through music is a relatively new concept that may have implications for mental health. In particular, it seems that music engagement may involve either more adaptive (e.g., maintenance and enhancement of positive emotions) or more maladaptive (e.g., venting negative emotions) forms of emotion regulation, which respectively promote and affect mental health (Thomson, Reece, & Di Benedetto, 2014). However, there is a lack of research on the personality determinants that may lead to the use of emotion regulation through music.
Neuroticism as a disposition toward emotion regulation through music
Chamorro-Premuzic and Furnham (2007) posited that because higher neuroticism involves intense, negative, and unstable emotionality, it should, in turn, increase sensitivity to music’s emotional properties and incite the use of music to regulate one’s difficult emotional states. This hypothesis was supported by findings that were replicated across a number of their studies (e.g., Chamorro-Premuzic & Furnham, 2007; Chamorro-Premuzic, Fagan, & Furnham, 2010; Chamorro-Premuzic, Gomà-i-Freixanet, Furnham, & Muro, 2009a; Chamorro-Premuzic, Swami, & Cermakova, 2012; Chamorro-Premuzic, Swami, Furnham, & Maakip, 2009b).
That being said, the interrelations between neuroticism and emotion regulation can be more complex, notably because there are different forms of emotion regulation. Notably, reappraisal is usually considered to be a more adaptive emotion regulation strategy, whereas suppression and rumination to be more maladaptive emotion regulation strategies (John & Gross, 2004). For example, Gullone and Taffe (2012) found that neuroticism was moderately correlated with both less reappraisal (r = −21) and also with more suppression (r = 0.26) in adolescence. However, Gross and John (2003) found neuroticism to be moderately and negatively correlated with reappraisal (r = −20), but not with suppression (r = 0.03) in young adulthood. Moreover, neuroticism has also been found to have a medium positive correlation with rumination (r = 0.32) and a moderate negative correlation with reappraisal (r = −0.22), but to have a non-significant correlation with suppression (r = 0.07) in middle adulthood (Yoon, Maltby, & Joormann, 2013). In sum, neuroticism might lose its significant association with suppression during adulthood.
Potential moderators
There are at least two important reasons to examine potential moderators in the present meta-analysis. First, personality research can gain nuance by examining moderators that may condition the potential predictive effects of personality traits on outcomes (Hampson, 2012). Second, it is also standard practice to examine potential moderators when conducting a meta-analysis (Borenstein, Hedges, Higgins, & Rothstein, 2009). We decided to focus on age and type of measure as potential moderators because these are some of the most common moderators that are routinely tested in recent meta-analyses relevant to either personality or emotion regulation (e.g., Compas et al., 2017; Sibley & Duckitt, 2008; Wilson & Dishman, 2015). Gender is also typically tested as a moderator; however, the studies we found did not report enough information to examine potential gender differences.
Age. Age is a marker of developmental level that may moderate the association between neuroticism and emotion regulation through music. More specifically, we expected that the effect size should be significantly larger during adolescence (10–21 years of age) than in adulthood. 1 The main rationale supporting our hypothesis is that listening to music is a particularly consequential, meaningful, and ubiquitous activity during adolescence (Miranda, 2013).
Type of measure. Different self-report measures of emotional regulation through music may yield heterogenous findings from which it would be difficult to calculate an average effect size. To test if there could be such measurement heterogeneity, we explored if the effect size would be different if we compared the emotional use subscale of the Uses of Music Inventory (UMI; Chamorro-Premuzic & Furnham, 2007) to the rest of self-report measures of musical emotion regulation. The main reason is that from our reading of research on neuroticism and emotion regulation via music, the UMI is the most thoroughly validated self-report measure and the most utilized. Hence, the absence of this moderation would suggest that findings are comparable across studies using the UMI or other self-report measures, whereas the presence of this moderation would caution of possible heterogeneity in self-report measures across studies.
Research question
In this meta-analysis, our research question is to estimate the magnitude of the association between neuroticism and the use of music listening for emotion regulation. It is an important research question from the perspective of at least two arguments. First, in terms of mental health in everyday life, neuroticism can increase the probability of developing many different psychopathologies (Kotov et al., 2010; Lahey, 2009). As such, identifying whether meaningful, common, and omnipresent daily behaviors (e.g., music listening behaviors) are associated with neuroticism can be a step forward to better understand how the negative effects of this trait actually unfold in everyday life. Second, in terms of mental health in therapeutic settings, people who consult for therapy can have high neuroticism prior to beginning the intervention (Lüdtke, Roberts, Trautwein, & Nagy, 2011). Hence, because neuroticism can also be related to more time spent listening to music (Kawase, 2016), working on music listening behaviors may represent a practical intervention strategy to help clients with high neuroticism.
Method
We used five different strategies to search for studies. First, we searched for published articles on PsycINFO (from 1806 to the fifth week in April 2018). Second, we searched for theses on ProQuest Dissertations & Theses Global (through 7 May 2018). Third, from citations of the seminal Chamorro-Premuzic and Furnham (2007) article in Google Scholar, we searched for published articles or unpublished theses that used the UMI (searching the term “Uses of Music Inventory” within citing articles through 4 May 2018). Fourth, we consulted the recent systematic literature review on emotion regulation through music by Baltazar and Saarikallio (2016). Fifth, we examined if there were any more references cited within our selected studies.
Literature search
Keywords. The following combination of keywords was used on PsycINFO and ProQuest Dissertations & Theses Global: Music AND emotion regulation OR emotional regulation OR affect regulation OR mood regulation OR mood management OR emotional management AND neuroticism OR emotional stability OR big five OR personality OR traits.
Inclusion criteria. Inclusion criteria for searched studies consisted of: Quantitative studies, English language, music reasons, or motivations or behaviors to manage emotions with music, music in everyday life as studied via correlational research, self-report, community or student samples, peer-reviewed articles, or theses with a formal academic evaluation (doctoral, masters, or bachelor’s honors levels).
Exclusion criteria. Exclusion criteria for searched studies included: Qualitative studies, music reactions or music genres or music preferences, research with an experimental design or in laboratory settings, physiological markers, clinical samples or interventions, no formal scientific evaluation (either scholarly peers or academic committee), and suspected predatory journals.
Search results
PsycINFO (n = 30), ProQuest Dissertations & Theses Global (n = 9), Google Scholar (n = 30), and the Baltazar and Saarikallio (2016) review (n = 35) yielded a total of 104 references. Using our inclusion and exclusion criteria, and after removing duplicate studies, a total of 13 pertinent studies were found.
Studies characteristics and data coding
Table 1 presents the characteristics of each study included in the meta-analysis. All studies used self-report data through a correlational design. Twelve studies were cross-sectional and one was longitudinal. Ten studies were published articles and the three others were unpublished theses. Six studies were conducted among middle or late adolescents, whereas seven were conducted with adults and older adults. Seven studies used the UMI, whereas the others used various other self-report measures of musical emotion regulation.
Characteristics of Studies Included in the Meta-Analysis (K = 13).
N = 232 for the correlation provided to us by the author of that study.
USA: United States of America; SD: standard deviation; NEO-FFI: NEO Five-Factor Inventory; ERM: Emotion Regulation through listening to Music; IPIP: International Personality Item Pool; PANAS: Positive and Negative Affect Schedule; BFI: Big Five Personality Inventory; TIPI: Ten-Item Personality Inventory; BFQ: Big Five Questionnaire; MMR: Music in Mood Regulation scale; FFMRF: Five-Factor Model Rating Form; IMLS: Inventory of Music Listening Situations.
Originally, the Laukka (2007) study did not report the correlation between their measure of neuroticism and emotion regulation via music. However, we were able to obtain the Pearson correlation for this specific association. The longitudinal study by Miranda, Gaudreau, and Morizot (2010) reported Spearman correlations for girls and boys. Nonetheless, we were able to obtain the Pearson correlation for all participants at baseline for sake of comparability with the other cross-sectional studies. The study of Getz, Chamorro-Premuzic, Roy, and Devroop (2012) used the UMI, but they measured negative affect instead of neuroticism. We kept this study nonetheless primarily because their use of the UMI was highly consistent with prior work by Chamorro-Premuzic and colleagues but also because negative affectivity is a core element in neuroticism. The study of Carlson et al. (2015) reported the correlation between neuroticism and one (i.e., discharge) of seven subscales from the Music in Mood Regulation scale (Saarikallio, 2008). One study (Chamorro-Premuzic et al., 2009b) reported a standardized beta (β) in lieu of a Pearson correlation (r). Hence, the β was converted to an r by adding 0.05 to the coefficient (as per Peterson & Brown, 2005). 2
Plan of analysis
Meta-analytical estimates were performed with the software Comprehensive Meta-Analysis, Version 3 (Borenstein, Hedges, Higgins, Rothstein, 2013). We selected a random-effects model because we did not assume that all studies were identical in methodological procedure, measurement, or population. In a random-effects model, the postulate is that each study has a true effect that contributes to a distribution and a mean of true effects; whereas in a fixed-effect model, the proviso is that all studies share the same true effect and variation is due to measurement error within each study (Borenstein et al., 2009). In this meta-analysis, the index for the weighted mean of effect sizes is the Pearson (r) correlation. 3
Heterogeneity of effect sizes. A significant Q statistic was used as an indicator of significant heterogeneity of effect sizes across studies (Borenstein et al., 2009). To further examine the degree of effect size heterogeneity across studies, the I2 index was used as it reports the percentage of the total variability in effect sizes that corresponds to actual effect size heterogeneity between studies (Higgins & Thompson, 2002; Higgins, Thompson, Deeks, & Altman, 2003). Lastly, in the event that the Q statistic was significant, it was used as a preliminary step to justify the examination of moderation effects.
Publication Bias. Potential publication bias was assessed in three ways. First, we visualized a Funnel Plot for signs of asymmetry that would suggest publication bias (Borenstein et al., 2009). Second, Kendall’s Tau rank order correlation and Egger’s regression intercept tested the significance of the Funnel Plot asymmetry (Borenstein et al., 2009). Third, in terms of publication status, we conducted a subgroup analysis (mixed effects) to examine whether the subgroup effect sizes were significantly different across peer-reviewed articles and unpublished theses.
Results
The 13 studies (k = 13) included a total of 2641 participants. The weighted mean correlation indicated a significant small-to-medium summary effect (r =.22, 95% CI [0.17, 0.27], z = 8.09, p <.001) estimated from a random-effects model. 4 Overall, there was no evidence of significant heterogeneity in effect sizes across all 13 studies (Q (12) = 20.28, p =.06). Also, the degree of heterogeneity was moderate (I2 = 40.83; Higgins et al., 2003). Hence, based on these grounds, moderation analyses were not deemed necessary for age nor type of measure.
Figure 1 presents the forest plot with the effect sizes and 95% confidence interval for each study. 5 Three studies had confidence intervals (CIs) that included a null effect (r = 0). However, two of those three studies were Bachelor’s honors theses that had a substantial amount of error variance. The other study was that of Laukka (2007), which had a larger sample. Nonetheless, relative to the other studies, it may represent an outlier insofar as it selected a very brief measure of neuroticism, it used a measure of emotion regulation that did not undergo psychometric validation, and it was conducted in a population of older adults. 6 For sake of comparison, results without the Laukka (2007) study would have also yielded a significant small-to-medium summary effect (r =.24, 95% CI [0.20, 0.28], z = 11.85, p <.001, k = 12, n = 2409). 7
Forest Plot of Studies’ Effect Sizes, Confidence Intervals, and Relative Weights.
Publication bias
Overall, there was no evidence of publication bias. Figure 2 presents the Funnel Plot from which a visual inspection did not suggest asymmetry. The study of Laukka (2007), however, is the one falling outside the funnel, but it may represent an outlier. In fact, Kendall’s Tau rank order correlation (−0.03, p =.90) and Egger’s regression intercept (−0.14, p = 0.90) did not detect asymmetry given that they were both non-significant. Finally, subgroup analysis (mixed effects) indicated that the subgroup effects were not significantly different across peer-reviewed articles (k = 10) and unpublished theses (k = 3): Q (1) = 0.11, p = 0.74.
Funnel Plot of Standard Error by Fisher’s Z.
Discussion
Our meta-analytical findings indicated a significant small-to-medium summary effect (r = 0.22) for the cross-sectional positive relationship between neuroticism and emotion regulation through music listening. The absence of evidence for significant heterogeneity in effect sizes among studies did not justify conducting subsequent moderation analyses for age and type of measure. Thus far, the extant studies did not yield evidence in support of our hypotheses that the effect size for the association between neuroticism and musical emotion regulation would be larger in adolescents compared to adults. However, the Q statistic can be sensitive to lack of statistical power in meta-analyses with a small number of studies (Borenstein et al., 2009). Hence, the present findings point to a lack of heterogeneity across studies; however, more studies need to be meta-analyzed to properly test a moderation by age or type of measure.
Our summary effect is somewhat in the range of what has been found in previous studies when neuroticism was significantly associated with emotion regulation (e.g., Gross & John, 2003; Gullone & Taffe, 2012; Yoon et al., 2013). Compared with other meta-analyses on individual differences, our summary effect size is of a typical magnitude (r =.10 is small, r =.20 is typical, and r =.30 is large; Gignac & Szodorai, 2016). More specifically, the present summary effect is greater than for that of neuroticism and music preferences (ranging from r = −.04 to r =.04; Schäfer & Mehlhorn, 2017) but lesser than that of neuroticism and psychopathology (e.g., r = 0.39 across depressive, anxiety, and substance use disorders; Kotov et al., 2010). Interestingly, our summary effect is larger than those reported in two prior meta-analyses on neuroticism and emotion regulation. Joseph and Newman (2010) conducted a meta-analysis in which emotional stability (neuroticism reversed) had a small relationship (r =.14) with emotion regulation as a facet of emotional intelligence. Moreover, Connor-Smith and Flachsbart (2007) conducted a meta-analysis in which neuroticism was not associated at all (r =.00) with emotion regulation as a component of coping. Therefore, perhaps the relationship between neuroticism and musical emotion regulation is stronger because of the emotional nature and ubiquitous presence of music in everyday life.
Neuroticism and adaptive versus maladaptive musical emotion regulation
The present meta-analysis indicates that people higher in neuroticism make more use of music for emotion regulation but it does not reveal how they do it. For instance, the fact that neuroticism is associated with more emotion regulation through music does not necessarily suggest that emotion regulation through music is a maladaptive consequence of trait emotional instability. Indeed, this would be a disturbing thought given the emotional importance of music to so many people. Rather, it is important to reiterate that there are both adaptive and maladaptive forms of emotion regulation through music (Thomson et al., 2014). Unfortunately, the vast majority of studies on neuroticism and emotion regulation through music did not operationalize both adaptive and maladaptive forms of musical emotion regulation. As acknowledged by Chamorro-Premuzic and collaborators (2009a, 2009b), future studies on neuroticism and music should distinguish between positive and negative forms of musical emotion regulation. In fact, within those measures of musical emotion regulation reviewed in this meta-analysis, the negative aspects of musical emotion regulation often outweighed the positive ones. For instance, among the five items of emotion regulation in the Uses of Music Inventory (UMI), only one is clearly about positive emotionality, namely happiness (Chamorro-Premuzic & Furnham, 2007, p. 179). For sake of comparison, besides the UMI, among the items of the emotional coping measure used by Miranda et al. (2010), most are about decreasing negative emotionality. In sum, given that neuroticism is linked to maladaptive emotion regulation strategies in general (Gullone & Taffe, 2012; Yoon et al., 2013), it is thereby conceivable that neuroticism would also be associated with maladaptive forms of musical emotion regulation (e.g., suppression or rumination). Moreover, conversely, perhaps trait emotional stability (low neuroticism) may also be associated with adaptive forms of musical emotion regulation (e.g., reappraisal).
Music listening as an emotional self-help strategy against neuroticism
People high in neuroticism might be more inclined to use music listening as an emotional self-help strategy. If that is the case, it would be an instantiation of music listening potentially impacting neuroticism rather than the other way around. Indeed, there is preliminary evidence that music listening may predict the development of either more or less neuroticism (Miranda et al., 2010). Hence, the cross-sectional positive relationship between neuroticism and emotion regulation via music may hide that some people higher in neuroticism might actually be using music listening as an attempt to self-regulate toward gradually reaching trait emotional stability. This would be compatible with recent research suggesting that people can set themselves goals to change and improve their personality traits, notably neuroticism (Hudson & Fraley, 2016; Robinson, Noftle, Guo, Asadi, & Zhang, 2015). For example, future research could examine if people can purposively reduce their neuroticism levels when resorting to more adaptive musical emotion regulation (e.g., reappraisal), as well as examine whether people may inadvertently increase their neuroticism levels when engaging in maladaptive musical emotion regulation (e.g., rumination, venting). Therefore, this calls for future longitudinal studies to verify the extent to which people are successful (or unsuccessful) at self-regulating their levels of neuroticism through different musical emotion regulation strategies.
Neuroticism and musical emotion regulation: Two hypothetical explanations
We propose two hypothetical explanations that future studies could examine as possible mediators explaining the relationship between neuroticism and emotion regulation through music. However, it is important to note that these two hypothetical explanations might be complementary and that they do not aim to represent a complete theoretical model elucidating the interface between neuroticism and musical emotion regulation.
First, we suggest the emotional overload hypothesis, according to which people higher in neuroticism might use music to manage their daily excess of negative emotionality. As previously mentioned, neuroticism involves negative and disproportionate emotional response to everyday challenges and adversities (Caspi et al., 2005; Lahey, 2009). As such, neuroticism can flare-up an overproduction of negative emotions following any given circumstance perceived as negative (Hervas & Vazquez, 2011). Consequently, given that the positive emotion of pleasure is a natural and typical reaction from music (Menon & Levitin, 2005), and that neuroticism can be tied to experiencing more pleasure from music in everyday life (Juslin, Liljeström, Västfjäll, Barradas, & Silva, 2008), those who are higher in neuroticism may be attempting to compensate for their negative emotional overload through the immediate pleasure of listening to the music they enjoy. However, given that positive and negative emotions belong to two distinct affective systems (Watson & Tellegen, 1985), this compensation strategy could be illusory as increasing positive emotions would not necessarily decrease the negative ones inherent to neuroticism. Alternatively, given that some people are also known to manage their negative emotionality by discharging (e.g., venting) their negative emotions through music (Saarikallio, 2012), this might be a more direct strategy used by people high in neuroticism to mitigate their overproduction of negative emotions. Indeed, neuroticism is associated with discharging (i.e., venting) negative emotions when listening to music (Carlson et al., 2015). Unfortunately, as highlighted earlier, discharging/venting negative emotions via music tends to be a maladaptive form of emotion regulation (Thomson et al., 2014).
Second, we propose the emotional self-focus hypothesis, which means that people higher in neuroticism may increase in self-focus on their negative emotionality when listening to music. More specifically, neuroticism is known to be associated with a tendency to ruminate about negative feelings and thoughts (Yoon et al., 2013). Furthermore, rumination in general (Aldao, Nolen-Hoeksema, & Schweizer, 2010) and rumination through music are both maladaptive forms of emotion regulation (Saarikallio, Gold, & McFerran, 2015). Rumination can also be associated with a tendency to listen to sad music (Garrido & Schubert, 2013), which could also increase exposure to negative themes that may beget even more rumination. Besides, the fact that people who are higher in neuroticism can experience more negative emotional reactions (e.g., sad, tense, anxious, regretful) when listening to music (Juslin, Liljeström, Laukka, Västfjäll, & Lundqvist, 2011) might be partially explained by rumination while listening to music. Lastly, because many people report seeking consolation from music and song lyrics (ter Bogt, Vieno, Doornwaard, Pastore, & van den Eijnden, 2017), perhaps those who are high in neuroticism are also listening to music to find solace from the many negative thoughts about which they ruminate on a daily basis. However, their engagement in musical consolation may possibly sustain the salience of their negative thoughts and thus maintain their rumination.
Limitations
Our meta-analysis did not examine extraversion although this trait can include among its specific facets a tendency toward positive emotionality (McCrae et al., 2005), which could in turn predict a greater use of music for positive emotion regulation. However, the relationship between extraversion and emotion regulation via music has yielded mix findings. For instance, the association between extraversion and musical emotion regulation has been found to be negative (e.g., Chamorro-Premuzic & Furnham, 2007), to be positive (e.g., Chamorro-Premuzic et al., 2009a), and ultimately it has even been excluded from hypothesized links within structural equation modeling (e.g., Chamorro-Premuzic et al., 2010).
Gender could not be considered as a moderator in this meta-analysis because only two studies reported analyses in both female and male participants (Carlson et al., 2015; Miranda et al., 2010). It could have been anticipated that the summary effect would be larger in female than in male participants. This could be expected because female participants tend to be higher in neuroticism (Costa, Terracciano, & McCrae, 2001) and also tend to make more use of music listening for emotion regulation (Chamorro-Premuzic et al., 2009a; Miranda & Claes, 2009).
Our meta-analysis is based on cross-sectional findings. Methodologically, this does not allow us to confirm that neuroticism is a predictor of musical emotion regulation. Yet, it is well-established that personality traits can represent dispositions and, as such, predictors of consequential outcomes (Ozer & Benet-Martínez, 2006), including functions of music listening (Chamorro-Premuzic & Furnham, 2007). This is why we primarily considered neuroticism as an antecedent of musical emotion regulation rather than its outcome. Nevertheless, more longitudinal studies are necessary to examine if and to which extent neuroticism predicts musical emotion regulation (and/or vice versa) across everyday life situations and the lifespan.
Of course, factors other than neuroticism are probably also impacting emotion regulation through music. This is evident by the fact that—although significant—the present small-to-medium summary effect of neuroticism still leaves much unexplained variance in musical emotion regulation. The literature suggests several possible determinants of emotion regulation that are not necessarily personality traits. For example, cognitive development in childhood and adolescence also shapes emotion regulation skills (Steinberg, 2005). The social context within the family (e.g., parenting style) can also impact the development of emotion regulation strategies (Morris, Silk, Steinberg, Myers, & Robinson, 2007). Moreover, socialization outside the family (e.g., at school or in the work place) can also influence how people regulate their emotions (John & Gross, 2004). Therefore, future research could also examine—in conjunction with personality traits—how social and cognitive factors may also act as potential determinants of emotion regulation through music.
Conclusion
In conclusion, much of the present findings may suggest that people higher in neuroticism are more prone to use music listening as an accessible resource to regulate their negative emotions or manage whatever affects their mood in everyday life. This is compatible with the notion that neuroticism is a continual negative and disproportionate emotional response to everyday challenges and adversities (Caspi et al., 2005; Lahey, 2009). Hence, it is in agreement with the hypothesis that neuroticism acts as a disposition toward using music to regulate unstable emotional states (Chamorro-Premuzic & Furnham, 2007). However, it is not possible to discern at this point if higher neuroticism in itself increases sensitivity to the emotional effects of music. In sum, the putative effect of neuroticism on musical emotion regulation seems relatively moderate. However, more research is justified notably to examine whether people higher in neuroticism are successfully (or unsuccessfully) using music listening as a daily emotional self-help strategy to gradually attain trait emotional stability.
