School Music and the Transition to College

Abstract

This study explored the transition from secondary to postsecondary education among a national sample of students who had or had not studied music in high school. Using evidence from the High School Longitudinal Study of 2009, a nationally representative longitudinal study of 21,440 American high school students who were ninth graders in the 2009–2010 school year, music and nonmusic students were compared for college admission outcomes. Specifically, music and nonmusic students were compared in terms of participation in the college admission process, the selectivity of colleges applied to and attended, scholarship and grant receipt, and election of either an arts or STEM (science, technology, engineering, mathematics) major. Comparisons controlled for the well-documented preexisting differences among those students who do and do not elect high school music study. Results showed that music and nonmusic students dropped out of high school, applied to college, attended college, received college scholarships and grants, and majored in STEM fields at statistically similar rates. However, music students were considerably more likely to major in a visual or performing arts field than nonmusic students. These results suggest that school music study does not disadvantage students in the transition to college even when compared with peers who elected additional “academic” subjects in lieu of music.

Keywords

music enrollment course-taking persistence in school college admission transition to postsecondary education

Accumulating research literature across diverse fields of inquiry—such as arts education, economics, psychology, and neuroscience—suggests there are, or should be, demonstrable and measurable academic benefits that accrue to arts students. For example, evidence from the prior literature across these fields (e.g., Catterall, 1997, 2009; Catterall et al., 1999, 2012; Deasy, 2002; Elpus, 2013; Fitzpatrick, 2006; Johnson & Memmott, 2006; Kraus & Chandrasekaran, 2010; Martin et al., 2013; Moreno et al., 2011; Young et al., 2014) indicates that high-quality arts education may be associated with artistic, social, and academic benefits for students, but the results are somewhat mixed. Comparisons of academic outcomes between arts and nonarts students in the research literature are sometimes problematic given that the abundance of extant correlational evidence cannot account for preexisting population differences between students who do and do not pursue music and arts course work (e.g., Elpus & Abril, 2011, 2019).

Among the various arts subdisciplines, the study of music has received the broadest attention from researchers. Music also has the most well-developed theoretical basis for potentially yielding academic benefits; however, the science remains far from settled. Researchers have theorized that music study may promote specific neurobiological adaptations that have implications for improved academic achievement (Gordon et al., 2015; Kraus & Chandrasekaran, 2010). Psychological research has shown that joint music-making promotes prosocial skills (Ilari et al., 2019; Kirschner & Tomasello, 2010; Schellenberg et al., 2015) and that formal music education in schools may help channel adolescents into more prosocial peer groups (Fredricks & Eccles, 2005; Nurmi, 2004); both of these findings may underpin observations of better social and academic outcomes for music students. Despite this research evidence that music education is associated with positive psychological, neurobiological, and educational benefits, contradictory results are also reported in the literature. Rigorous experimental and quasi-experimental studies by Costa-Giomi (2004), Elpus (2013), and Mehr et al. (2013) have shown that music students may not consistently outperform nonmusic students on a variety of academic measures.

Although population differences may account for some of the more widely touted academic consequents of music and arts education, newer research with more rigorous analytic approaches promotes promising theories beyond the 1990s-era correlational linkage of arts education to academic outcomes (often, test scores) in K–12 schools. Researchers engaged in this emerging line of inquiry seek to determine whether and how arts education might be linked to students’ successful progress through the educational system. Rather than merely comparing the standardized test scores of arts and nonarts students, this research explores a theory suggesting that arts education promotes better social/emotional learning (Cohen & Sandy, 2007) and creates greater engagement with school (Dwyer, 2011), which in turn has the academic benefits of increasing school persistence toward achieving high school graduation with no dropout episodes and promoting greater postsecondary aspirations and attainment (Elpus, 2018).

Persistence in school and school engagement are, in and of themselves, critical academic indicators that have been linked with better long-term academic, social, and economic outcomes (Fredricks et al., 2004, 2005). Music and arts education researchers have only recently begun to explore school engagement among music and arts students. For example, Catterall and colleagues (2012) showed that “at-risk” students hailing from “arts-rich” K–12 schools were more likely to have attended college than their peers from schools with fewer arts opportunities. Research using administrative education data from Texas (Thomas et al., 2015) suggests that music and arts students are considerably less likely to drop out of school than are nonarts students. Martin et al. (2013) found that arts participation among youths in Australia was linked with greater academic aspirations, greater enjoyment of school, and an overall sense of life satisfaction. Likewise, Guhn and colleagues (2020) showed that students in British Columbia, Canada, generally performed better than their nonmusic peers in provincial exams of several school subjects, rejecting the notion that elective participation in music courses poses an opportunity cost of lost time to study English, math, and science.

In an earlier study (Elpus, 2018), I established initial evidence for a theory explicitly linking music and arts education course-taking to greater postsecondary academic persistence, possibly through the mechanism of school engagement. Using data from the Education Longitudinal Study of 2002 base year and attendant follow-up waves, I found that American students who had taken arts course work in their freshman, sophomore, and/or junior years of high school—that is, prior to college applications in the senior year—were more likely than their nonarts peers to participate in the college admissions process by applying to at least one postsecondary institution. This effect persisted even after statistically controlling for a host of known preexisting population differences between arts and nonarts students. In addition, students who had enrolled in one or more music or arts education class applied to and were admitted at colleges of equal selectivity to their nonmusic and arts peers when matched on grade point average (GPA) and standardized test scores, implying that there was no college admission penalty for students to opt for arts courses in place of, say, doubling up on additional math or science courses. Among the various arts disciplines pursued by that cohort of students in the United States, music study was most strongly linked with postsecondary outcomes.

Taken together, the Catterall et al. (2012), Thomas et al. (2015), Martin et al. (2013), Elpus (2018), and Guhn et al. (2020) studies do somewhat strongly suggest that K–12 music and arts education experiences may be linked to successful progress and persistence through the educational system. Still, the science supporting this notion remains far from settled. A solid theoretical rationale focusing on arts study for why arts students should outperform nonarts students on academic performance measures unrelated to the arts has yet to be convincingly articulated. Although school engagement may be one mechanism through which these benefits accrue to music and arts students, the precise causal chain through which these positive educational benefits accrue has not yet been fully explained in the literature. Examined individually, each unique educational benefit linked to music and arts education also has relatively few—if any—studies showing that these effects replicate across varying social, economic, and geographic contexts or across different cohorts. The present study, part of this burgeoning line of research, will add to the evidentiary base in this area.

Purpose of the Present Study and Research Questions

Given that the line of research associating music education to successful progress through the educational system is still emerging, the purpose of the present study was to explore the transition from secondary to postsecondary school for a nationally representative cohort of American high school music students. This study extends the existing research line with evidence from a newer cohort of students than previously explored in the Elpus (2018) study while leveraging research insights about population differences between music and nonmusic students in the sample cohort (Elpus & Abril, 2019). The specific research questions guiding the study were (1) How does successful progress (i.e., no dropout episodes) through the K–12 education system and participation in the college admissions process vary for high school music and nonmusic students? (2) Among those who applied to a postsecondary institution, how does the selectivity ranking of colleges applied to vary for high school music and nonmusic students? Among those who attended a postsecondary institution, how does college selectivity vary for high school music and nonmusic students? (3) Among those who attended a postsecondary institution, how does the rate of college scholarship and grant receipt vary among music and nonmusic students? and (4) Among those who attended a postsecondary institution, how does the rate of electing a major in a visual or performing arts field vary among music and nonmusic students? How does the rate of electing a major in a STEM (science, technology, engineering, or math) field vary among music and nonmusic students?

Method

Data Source

In this study, I analyzed restricted-use data from the base year and three follow-up waves of the High School Longitudinal Study of 2009 (HSLS), which, at the time of this writing, is the most recently available nationally representative longitudinal survey study of American secondary and postsecondary education conducted by the National Center for Education Statistics (NCES). HSLS sample members—approximately 21,440 American high school students who were enrolled in ninth grade during the 2009–2010 school year—were drawn in a complex multistage sampling procedure that selected students probabilistically (but not randomly) from approximately 940 American public and private high schools that had been randomly selected from among the universe of all U.S. schools serving ninth graders in 2009. As with all the longitudinal studies of education the NCES produces, HSLS data include survey weights, which must be employed in analyses to ensure that estimates made from HSLS data are nationally representative. The weights also adjust for certain unique features of the HSLS sample—for example, to ensure their adequate representation in the sample for subgroup analyses, students of Hispanic or Latino ethnicity and those who identify as Asian American were overrepresented in the sample compared with their proportion in the population of American students. Properly weighting the analyses adjusts population estimates for subgroup oversampling, nonresponse, and other features of the HSLS sampling procedure. Full details of the HSLS samples and weighting procedures are documented in Duprey et al. (2018). HSLS sample members, their parents, their math and science teachers, their school counselors, and school administrators were surveyed first when the sample members were ninth graders in 2009–2010 (the base year). Sample members, parents, counselors, and administrators (but not teachers) were again surveyed when the modal sample member was an 11th grader in the spring of 2012 (the first follow-up). Sample members and their parents were surveyed yet again when the modal sample member was a graduating senior in summer/fall 2013 (the 2013 update). Sample members only were once more surveyed in 2016–2017 (the second follow-up). Administrative data were collected from high schools in the base year, first follow-up, and 2013 update waves and from postsecondary institutions in the second follow-up.

I identified high school music students and nonmusic students using the same method of complete high school transcript analysis previously employed by Elpus and Abril (2019) and established as valid in earlier work (Elpus, 2017). First, I manually verified that each transcript course coded as a music course in the data set was correctly coded. Next, I classified any student with one or more verified, credit-earning music courses on their transcript as a “music student.” Unlike the classification used in Elpus and Abril, I did not limit the term music students to those enrolled in an ensemble. Here, I considered music students all students with any performance or nonperformance music course appearing on their transcripts. Because of an emphasis in the prior literature on ensemble classes, I created three binary indicator variables of music status: (a) music student (set to 1 for a student who had taken any kind of music class and to 0 otherwise), (b) instrumental music student (set to 1 for a student who had enrolled in an instrumental music ensemble and to 0 otherwise), and (c) choral music student (set to 1 for a student who had enrolled in a choral music ensemble and to 0 otherwise). I did not create a specific category for students who had taken more than one type of music course. For example, choral students who also performed in a wind band are included in all three categories (music student, instrumental student, and choral student).

Empirical Approach

For all analyses reported in the article, I controlled for the following preexisting population differences between music and nonmusic students identified in prior research as both related to music enrollment status and successful postsecondary transitions: (a) race/ethnicity; (b) gender (i.e., birth-assigned sex); (c) family socioeconomic status; (d) whether the student was born in the United States; (e) whether the student was a native speaker of English; (f) the educational attainment of the student’s highest achieving parent; (g) whether the student lived in a single- or dual-parent/guardian household; (h) academic achievement of the student, measured by eighth-grade science grades, a standardized math test administered in freshman year to all sample members, and 11th-grade GPA; (i) whether the student played a sport in high school; and (j) how many hours per week the student was engaged in extracurricular activities during high school. Each of these variables is observable in the HSLS data set and was entered as a covariate into all the regression models described in the next section. As is typical of regression estimation, all continuous and binary covariates were entered into the regression model directly, whereas categorical covariates were entered as a series of dummy dichotomous variables omitting one category as a comparison group.

I estimated the relation between music study and the outcomes of interest using appropriate survey-adjusted regression techniques. I employed the proper regression link function for each analysis based on the scaling of the outcome variable. For selectivity rankings of the colleges where students applied or attended, I used an ordered logit model (McCullagh, 1980). Ordered logit, also known as the proportional odds model, is a regression method appropriate for ordinal outcomes such as college selectivity rankings, where higher selectivity can be logically ordered from lowest to highest but monotonic intervals between successive ranks are not assured. In this application, the ordered logit models yielded estimates of the proportional odds that music students would apply to or be admitted at more selective colleges than their nonmusic peers while holding all covariates in the model constant. For the binary outcomes, I used standard logistic regression. All analyses were computed using survey-adjusted procedures accounting for the sampling structure and weights in the data; all standard errors reported are robust to the complex nature of the HSLS sample. Alpha for null hypothesis significance testing is set at the customary .05 level.

Results

The logistic regression and ordinal models for music students enrolled in any music class are shown in Table 1. Tables 2 and 3 show the results for instrumental music students and choral music students, respectively. In all three tables, Research Question 1 is addressed in Models 1 and 2, Research Question 2 is addressed in Models 3 and 4, Research Question 3 is addressed in Model 5, and Research Question 4 is addressed in Models 6 and 7. In the prose presentation of the results, I structure this section of the article by outcome rather than by type of music studied and interpret results by outcome across all three tables. Where appropriate, I interpret the results using marginal predicted probabilities¹ rather than by simply repeating the odds ratios shown in the tables because odds ratios, being multiplicative, can sometimes obscure the true magnitude of changes in the absolute probability estimated for a specific event.

Table 1.

Outcome Comparisons for Students Who Enrolled in Any High School Music Course.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Ever Dropped Out	Applied to Any College	Selectivity of Colleges Applied to	Selectivity of College Attended	Received a Scholarship or Grant	Majored in an Arts Field	Majored in a STEM Field
	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p
Music student (any type)	1.09	1.21	1.03	1.09	1.04	1.80***	1.18
	(0.19)	(0.15)	(0.08)	(0.07)	(0.09)	(0.27)	(0.14)
	.622	.127	.718	.199	.653	.000	.166
Covariates
Race/ethnicity
Black	0.84	3.15***	1.69***	1.41**	1.92***	1.18	1.22
	(0.14)	(0.82)	(0.27)	(0.18)	(0.33)	(0.35)	(0.21)
	.297	.000	.001	.009	.000	.572	.252
Latinx	1.02	1.97**	0.84	0.82	1.02	1.27	1.47**
	(0.33)	(0.42)	(0.13)	(0.09)	(0.21)	(0.49)	(0.20)
	.952	.002	.251	.082	.941	.533	.006
Asian/Pac Islander	1.26	1.31	1.48*	1.46*	0.78	0.67	1.79*
	(0.41)	(0.74)	(0.28)	(0.23)	(0.16)	(0.30)	(0.46)
	.491	.629	.042	.016	.237	.367	.025
Native American	1.29	0.47	0.68	0.58	2.84	0.50	0.99
	(0.94)	(0.20)	(0.32)	(0.19)	(2.00)	(0.38)	(0.66)
	.726	.084	.417	.089	.138	.364	.982
Two or more races	1.06	1.18	0.94	0.62*	1.33	1.28	1.13
	(0.19)	(0.24)	(0.13)	(0.12)	(0.20)	(0.34)	(0.17)
	.750	.409	.677	.012	.064	.351	.440
Male	1.03	0.59***	0.99	0.91	0.86	0.97	2.84***
	(0.15)	(0.07)	(0.09)	(0.08)	(0.07)	(0.15)	(0.25)
	.846	.000	.902	.240	.053	.821	.000
SES	0.71**	1.65***	1.33***	1.49***	.63***	1.27	1.08
	(0.09)	(0.24)	(0.11)	(0.11)	(0.06)	(0.22)	(0.10)
	.009	.001	.001	.000	.000	.172	.445
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Ever Dropped Out	Applied to Any College	Selectivity of Colleges Applied to	Selectivity of College Attended	Received a Scholarship or Grant	Majored in an Arts Field	Majored in a STEM Field
	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p
Student born in the United States	0.79	1.36	1.16	1.19	0.98	1.55	0.79
	(0.18)	(0.39)	(0.27)	(0.31)	(0.20)	(0.59)	(0.20)
	.287	.279	.521	.496	.922	.252	.337
Native English speaker	1.26	0.60*	0.63*	0.72*	0.99	0.80	0.98
	(0.36)	(0.14)	(0.12)	(0.09)	(0.19)	(0.26)	(0.20)
	.426	.029	.013	.012	.976	.506	.917
Parental education
High school diploma	0.71	0.91	1.79*	1.64*	1.49	2.33	1.17
	(0.17)	(0.27)	(0.50)	(0.32)	(0.46)	(1.63)	(0.39)
	.161	.753	.037	.010	.200	.227	.640
Associate’s degree	0.65	1.34	1.62	1.40	1.52	2.40	0.86
	(0.24)	(0.46)	(0.45)	(0.32)	(0.57)	(2.00)	(0.38)
	.247	.395	.085	.141	.266	.294	.732
Bachelor’s degree	0.61	1.60	1.67	1.65*	1.19	3.30	1.36
	(0.17)	(0.60)	(0.44)	(0.34)	(0.40)	(2.40)	(0.53)
	.084	.213	.055	.016	.607	.101	.430
Master’s degree	0.70	1.58	2.22**	2.22***	1.11	3.62	1.38
	(0.19)	(0.60)	(0.63)	(0.48)	(0.34)	(2.75)	(0.51)
	.203	.228	.005	.000	.734	.090	.388
Educational specialist diploma	1.17	1.31	2.89***	2.91***	1.23	3.56	1.47
	(0.39)	(0.56)	(0.83)	(0.66)	(0.40)	(2.76)	(.58)
	.639	.520	.000	.000	.517	.102	.336
PhD/JD/MD, etc.	0.62	3.31	3.59***	3.78***	0.95	2.87	1.30
	(0.25)	(2.14)	(1.11)	(1.00)	(0.33)	(2.27)	(0.54)
	.238	.064	.000	.000	.881	.183	.521
Dual-parent household	0.60***	0.84	0.85	0.86	0.82	0.96	0.73
	(0.09)	(0.13)	(0.08)	(0.08)	(0.10)	(0.23)	(0.13)
	.001	.238	.081	.126	.120	.871	.069
11th-grade GPA	0.39***	2.64***	2.42***	2.73***	1.28***	1.30*	1.32**
	(0.03)	(0.21)	(0.20)	(0.16)	(0.09)	(0.17)	(0.12)
	.000	.000	.000	.000	.001	.038	.003
Standardized test score	0.99	1.03***	1.06***	1.06***	1.01	1.03**	1.04***
	(0.01)	(0.01)	(0.01)	(0.00)	(0.01)	(0.01)	(0.01)
	.069	.000	.000	.000	.178	.002	.000
8th-grade science grade	0.99	1.03**	1.03***	1.03***	1.00	.99	1.01
	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)
	.227	.002	.000	.000	.893	.337	.121
Athlete	0.76	1.55**	1.40***	1.39***	1.02	0.50***	0.96
	(0.12)	(0.23)	(0.13)	(0.11)	(0.10)	(0.08)	(0.12)
	.084	.003	.000	.000	.846	.000	.720
ECA (hours per week)	1.01	1.04	1.07*	1.08***	.99	1.04	0.99
	(0.04)	(0.05)	(0.03)	(0.02)	(0.03)	(0.05)	(0.03)
	.866	.409	.011	.000	.865	.444	.668

Note. Estimates reported as odds ratios. Binary outcomes for Models 1, 2, 5, 6, and 7 were estimated with logistic regression. Ordinal outcomes for Models 3 and 4 were estimated with an ordered logit model. Estimates and significance tests are adjusted for the High School Longitudinal Study complex sampling design. Survey-adjusted standard errors reported in parentheses. Categorical variable results are interpreted as comparisons to the following omitted reference groups: for race/ethnicity, White; for parental education, less than high school diploma. STEM = science, technology, engineering, mathematics; SES = socioeconomic status; GPA = grade point average; ECA = extracurricular activities.

p < .05. **p < .01. ***p < .001.

Table 2.

Outcome Comparisons for Students Who Enrolled in High School Instrumental Music.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Ever Dropped Out	Applied to Any College	Selectivity of Colleges Applied to	Selectivity of College Attended	Received a Scholarship or Grant	Majored in an Arts Field	Majored in a STEM Field
	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p
Instrumental music student	0.76	1.64*	0.91	0.98	1.25	1.72***	1.26*
	(0.13)	(0.35)	(0.09)	(0.08)	(0.15)	(0.27)	(0.14)
	.100	.019	.297	.831	.059	.001	.037
Covariates
Race/ethnicity
Black	0.83	3.15***	1.68**	1.40*	1.93***	1.20	1.22
	(0.14)	(0.81)	(0.27)	(0.18)	(0.33)	(0.36)	(0.21)
	.288	.000	.001	.011	.000	.547	.254
Latinx	1.01	1.97**	0.84	0.81	1.02	1.21	1.46**
	(0.32)	(0.43)	(0.13)	(0.10)	(0.21)	(0.46)	(0.20)
	.985	.002	.235	.076	.923	.620	.006
Asian/Pac Islander	1.26	1.33	1.47*	1.45*	0.78	0.68	1.78*
	(0.42)	(0.75)	(0.28)	(0.23)	(0.16)	(0.29)	(0.46)
	.487	.616	.043	.018	.238	.363	.027
Native American	1.31	0.46	0.68	0.57	2.83	0.47	0.96
	(0.95)	(0.20)	(0.32)	(0.18)	(1.99)	(0.36)	(0.64)
	.714	.070	.415	.077	.140	.323	.956
Two or more races	1.06	1.18	0.94	0.62*	1.33	1.29	1.13
	(0.19)	(0.23)	(0.13)	(0.12)	(0.20)	(0.34)	(0.17)
	.725	.417	.671	.012	.064	.340	.445
Male	1.03	0.58***	0.99	0.90	0.86*	0.93	2.81***
	(0.14)	(0.07)	(0.09)	(0.08)	(0.07)	(0.14)	(0.24)
	.853	.000	.909	.228	.044	.620	.000
SES	0.72**	1.63***	1.33***	1.49***	0.63***	1.23	1.07
	(0.09)	(0.24)	(0.11)	(0.11)	(0.06)	(0.21)	(0.10)
	.009	.001	.001	.000	.000	.229	.486
Student born in the United States	0.79	1.37	1.17	1.20	0.98	1.59	0.78
	(0.18)	(0.39)	(0.27)	(0.31)	(0.20)	(0.62)	(0.20)
	.306	.268	.508	.484	.913	.230	.337
Native English speaker	1.26	0.60*	0.63*	0.72*	0.99	0.79	0.98
	(0.36)	(0.14)	(0.12)	(0.09)	(0.19)	(0.26)	(0.20)
	.423	.029	.012	.011	.979	.474	.925
Parental education
High school diploma	0.71	0.93	1.78*	1.64*	1.50	2.42	1.19
	(0.17)	(0.28)	(0.49)	(0.31)	(0.47)	(1.70)	(0.40)
	.166	.794	.038	.010	.191	.209	.604
Associate’s degree	0.64	1.38	1.62	1.40	1.53	2.53	0.88
	(0.24)	(0.47)	(0.45)	(0.32)	(0.58)	(2.12)	(0.39)
	.245	.348	.084	.139	.259	.267	.768
Bachelor’s degree	0.62	1.61	1.67	1.65*	1.20	3.51	1.39
	(0.17)	(0.61)	(0.44)	(0.34)	(0.41)	(2.55)	(0.53)
	.086	.207	.054	.015	.597	.085	.391
Master’s degree	0.70	1.63	2.22**	2.22***	1.12	3.87	1.41
	(0.20)	(0.62)	(0.63)	(0.48)	(0.35)	(2.92)	(0.52)
	.206	.202	.005	.000	.714	.074	.354
Educational specialist diploma	1.17	1.34	2.90***	2.93***	1.24	3.84	1.50
	(0.40)	(0.57)	(0.83)	(0.66)	(0.40)	(2.98)	(0.59)
	.643	.491	.000	.000	.507	.083	.301
PhD/JD/MD, etc.	0.62	3.34	3.60***	3.81***	0.95	3.10	1.34
	(0.25)	(2.16)	(1.11)	(1.00)	(0.33)	(2.43)	(0.55)
	.244	.063	.000	.000	.890	.149	.479
Dual-parent household	0.61**	0.83	0.85	0.86	0.82	0.97	0.73
	(0.09)	(0.12)	(0.08)	(0.08)	(0.10)	(0.23)	(0.13)
	.001	.222	.087	.129	.114	.906	.072
11th-grade GPA	0.40***	2.63***	2.44***	2.75***	1.27***	1.31*	1.32**
	(0.03)	(0.21)	(0.20)	(0.17)	(0.09)	(0.17)	(0.12)
	.000	.000	.000	.000	.001	.036	.003
Standardized test score	0.99	1.03***	1.06***	1.06***	1.01	1.03**	1.04***
	(0.01)	(0.01)	(0.01)	(0.00)	(0.01)	(0.01)	(0.01)
	.088	.000	.000	.000	.211	.003	.000
8th-grade science grade	0.99	1.02**	1.03***	1.03***	1.00	0.99	1.01
	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)
	.263	.002	.000	.000	.851	.371	.122
Athlete	0.75	1.55**	1.38***	1.37***	1.03	0.51***	0.97
	(0.12)	(0.23)	(0.13)	(0.11)	(0.10)	(0.08)	(0.11)
	.074	.003	.001	.000	.757	.000	.773
ECA (hours per week)	1.02	1.04	1.07**	1.08***	0.99	1.03	0.98
	(0.04)	(0.05)	(0.03)	(0.02)	(0.03)	(0.06)	(0.03)
	.693	.450	.008	.000	.741	.527	.563

p < .05. **p < .01. ***p < .001.

Table 3.

Outcome Comparisons for Students Who Enrolled in High School Choral Music.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Ever Dropped Out	Applied to Any College	Selectivity of Colleges Applied to	Selectivity of College Attended	Received a Scholarship or Grant	Majored in an Arts Field	Majored in a STEM Field
	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p	b/(SE)/p
Choral music student	1.49	1.14	1.05	1.09	1.03	1.53**	0.99
	(0.38)	(0.21)	(0.10)	(0.10)	(0.12)	(0.23)	(0.20)
	.122	.489	.603	.333	.793	.005	.966
Covariates
Race/ethnicity
Black	0.84	3.16***	1.69***	1.40**	1.93***	1.18	1.21
	(0.15)	(0.81)	(0.27)	(0.18)	(0.33)	(0.35)	(0.21)
	.330	.000	.001	.009	.000	.586	.280
Latinx	1.03	1.95**	0.84	0.82	1.01	1.22	1.45**
	(0.32)	(0.42)	(0.13)	(0.09)	(0.21)	(0.46)	(0.20)
	.927	.002	.259	.079	.948	.605	.008
Asian/Pac Islander	1.26	1.33	1.48*	1.46*	0.78	0.70	1.78*
	(0.41)	(0.76)	(0.28)	(0.23)	(0.16)	(0.30)	(0.47)
	.477	.622	.041	.016	.243	.404	.030
Native American	1.35	0.47	0.68	0.57	2.84	0.49	0.96
	(0.97)	(0.21)	(0.32)	(0.18)	(2.00)	(0.37)	(0.65)
	.680	.088	.420	.085	.139	.349	.947
Two or more races	1.05	1.18	0.95	0.62*	1.33	1.28	1.12
	(0.19)	(0.24)	(0.13)	(0.12)	(0.20)	(0.34)	(0.17)
	.786	.411	.680	.012	.063	.344	.451
Male	1.07	0.59***	0.99	0.91	0.86	1.01	2.83***
	(0.16)	(0.07)	(0.10)	(0.08)	(0.07)	(0.15)	(0.25)
	.660	.000	.935	.284	.056	.969	.000
SES	0.71**	1.64***	1.33***	1.49***	0.63***	1.25	1.07
	(0.09)	(0.24)	(0.11)	(0.11)	(0.06)	(0.22)	(0.10)
	.008	.001	.001	.000	.000	.202	.495
Student born in the United States	0.77	1.37	1.16	1.19	0.98	1.61	0.79
	(0.17)	(0.39)	(0.28)	(0.31)	(0.20)	(0.62)	(0.20)
	.252	.269	.526	.492	.930	.215	.361
Native English speaker	1.26	0.60*	0.63*	0.72*	0.99	0.78	0.98
	(0.35)	(0.14)	(0.12)	(0.09)	(0.19)	(0.26)	(0.20)
	.409	.029	.013	.011	.970	.457	.920
Parental education
High school diploma	0.70	0.91	1.78*	1.64*	1.49	2.31	1.17
	(0.17)	(0.27)	(0.49)	(0.31)	(0.46)	(1.62)	(0.39)
	.137	.764	.037	.011	.201	.233	.639
Associate’s degree	0.64	1.35	1.62	1.40	1.52	2.50	0.87
	(0.24)	(0.46)	(0.45)	(0.32)	(0.57)	(2.10)	(0.38)
	.236	.386	.084	.140	.264	.276	.750
Bachelor’s degree	0.61	1.61	1.67	1.65*	1.19	3.41	1.38
	(0.17)	(0.60)	(0.44)	(0.34)	(0.41)	(2.48)	(0.53)
	.080	.205	.055	.016	.605	.091	.408
Master’s degree	0.70	1.61	2.22**	2.22***	1.11	3.71	1.39
	(0.19)	(0.61)	(0.63)	(0.48)	(0.35)	(2.80)	(0.51)
	.199	.215	.005	.000	.730	.084	.366
Educational specialist diploma	1.17	1.33	2.89***	2.91***	1.24	3.63	1.49
	(0.39)	(0.56)	(0.83)	(0.66)	(0.40)	(2.81)	(0.59)
	.650	.502	.000	.000	.514	.097	.312
PhD/JD/MD, etc.	0.62	3.35	3.59***	3.79***	0.95	3.04	1.34
	(0.25)	(2.17)	(1.11)	(1.00)	(0.33)	(2.40)	(0.55)
	.244	.063	.000	.000	.889	.159	.479
Dual-parent household	0.61***	0.84	0.85	0.86	0.82	0.97	0.73
	(0.09)	(0.12)	(0.08)	(0.08)	(0.10)	(0.24)	(0.13)
	.001	.245	.083	.126	.120	.890	.076
11th-grade GPA	0.39***	2.64***	2.42***	2.74***	1.28***	1.33*	1.34**
	(0.03)	(0.21)	(0.20)	(0.16)	(0.09)	(0.17)	(0.12)
	.000	.000	.000	.000	.001	.025	.002
Standardized test score	0.99	1.04***	1.06***	1.06***	1.01	1.04**	1.04***
	(0.01)	(0.01)	(0.01)	(0.00)	(0.01)	(0.01)	(0.01)
	.076	.000	.000	.000	.174	.002	.000
8th-grade science grade	0.99	1.03**	1.03***	1.03***	1.00	0.99	1.01
	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)
	.239	.002	.000	.000	.905	.422	.112
Athlete	0.75	1.53**	1.40***	1.38***	1.02	0.48***	0.94
	(0.12)	(0.22)	(0.13)	(0.11)	(0.10)	(0.08)	(0.11)
	.071	.003	.000	.000	.865	.000	.619
ECA (hours per week)	1.01	1.05	1.07*	1.08***	1.00	1.06	0.99
	(0.04)	(0.05)	(0.03)	(0.02)	(0.03)	(0.06)	(0.03)
	.875	.350	.010	.000	.884	.284	.771

p < .05. **p < .01. ***p < .001.

As previously reported by Elpus and Abril (2019), 34% of students in this cohort graduated high school having taken at least one music course of any type. Choral music ensembles were pursued by 13% of the cohort, band by 11% of the cohort, and orchestra by 2% of the cohort. Because of small cell sizes, there is insufficient statistical power to analyze the orchestra students in isolation; thus, these analyses combine orchestra with band students in a category I refer to as “instrumental music” students.

Likelihood of Ever Dropping Out of High School

Music students were neither more nor less likely than nonmusic students to have a dropout episode in high school. In examining the entire model, only socioeconomic status, GPA, and family composition were significantly related to the likelihood that a student would have ever dropped out of high school. These effects persist regardless of whether the student was enrolled in music. Each standard deviation increase in socioeconomic status decreases the probability that a student would have ever dropped out of high school by .02. Each standard deviation increase in GPA reduced the probability of dropping out by .05. Students from single-parent homes had a .13 chance of ever experiencing a dropout episode, which was statistically significantly greater than the .09 probability of dropout experienced by students from dual-parent families.

Disaggregated analyses by instrumental and choral music were also not statistically significant. In terms of absolute predicted probabilities, however, instrumental students did have a lower absolute probability of experiencing a dropout episode—.09—compared to noninstrumental students, who had a .11 probability of experiencing a dropout episode. Again, this .02 difference was not statistically significant (p = .10). Interestingly, the probability that choral students would experience a dropout episode (.13) was greater than that of nonchoral students (.10), but this difference was not statistically significant (p = .16).

Likelihood of Applying to Any College

Overall, music students were more likely than nonmusic students to apply to at least one college (.90 probability for music students compared to .89 probability for nonmusic students). Still, the slight difference in absolute terms was not statistically significant (p = .13). However, disaggregated analyses show that instrumental music students (probability = .92) were significantly more likely to apply to any college than were noninstrumental music students (probability = .89), p = .02. Choral music students were also slightly more likely (probability = .90) to apply to any college than were nonchoral students (probability = .89), but this .01 difference in probability was not statistically significant (p = .49).

Selectivity of Colleges Where Students Applied

Controlling for academic and sociodemographic factors, music students applied to colleges that were no more and no less selective than did their nonmusic peers. This result suggests that music enrollment status was not associated with a student applying to colleges that were either more selective or less selective than would be otherwise predicted by their standardized test scores, race/ethnicity, high school grades, athletic participation, hours per week in high school extracurricular activities, and socioeconomic status. Perhaps unsurprisingly, the strongest predictor of the selectivity of colleges applied to was parental educational attainment. The pattern of nonassociation between music status and application selectivity held for all three models; neither aggregated nor disaggregated music study was significantly associated with applying to a more selective postsecondary institution.

Unlike the lack of observed effect for music study, students who had participated in athletics did apply to colleges with greater selectivity rankings than their grades, socioeconomic, and other factors would otherwise predict.² The probability that an athlete would apply to a college with the most selective admissions was .30, compared to a .25 chance for nonathletes, and athletes were similarly less likely to apply to colleges with lower selectivity than nonathletes. As seen in Tables 1, 2, and 3, the estimated effect for athletes persists irrespective of whether the athlete pursued ensemble or nonensemble music courses.

Likelihood of Attending Any College

Music students were slightly but not statistically significantly more likely to have ever attended college by 2 years past high school graduation than were nonmusic students. The probability for music students (.80) was just .01 higher than the probability for nonmusic students (.79), and this slight difference was not significant (p = .57). As might be reasonably expected, socioeconomic status, 11th-grade GPA, and standardized test scores were all associated with the likelihood that a student would ever attend any college. Athletes were significantly more likely than nonathletes to attend college regardless of the athlete’s music enrollment status. Athletes had a .81 probability of attending any college, compared to a .76 probability for nonathletes; the difference is significant (p < .001).

Selectivity of College Attended

Music students were slightly more likely than nonmusic students to attend colleges with greater selectivity ratings, but the slight difference was not statistically significant (p = .20). Greater selectivity of the college a student attended was significantly related to race/ethnicity, socioeconomic status, parental educational attainment, GPA, standardized test scores, hours per week in extracurricular activities during high school, and athletic participation.

Ever Received a Scholarship or Grant for College

Compared to nonmusic students, music students were marginally more likely to have received a merit-based scholarship or grant to help defray the cost of attending a postsecondary institution, but the .005 change in probability was not statistically significant, p = .65. Instrumental students were about 3 percentage points more likely than noninstrumental students to have obtained a scholarship or grant, but the difference was not statistically significant (p = .06). The difference between choral students and nonchoral students (.004 percentage points) was virtually indistinguishable. As socioeconomic status increased, the likelihood of scholarship receipt decreased, but the likelihood rose with 11th-grade GPA.

Majored in an Arts Field

Of all the postsecondary outcomes in this analysis, music enrollment status was most highly associated with the likelihood that the student would major in an arts field. Although the likelihood that any student would major in an arts field was somewhat low, the increase in music students’ likelihood was comparatively large. Nonmusic students had a .03 probability of majoring in an arts field, whereas music students had a .05 probability—meaning roughly one in 20 college-bound high school music students went on to major in one of the arts. The increase in probability was statistically significant, p < .001. Predicted probabilities and the differences between the music and nonmusic students were the same across any music, instrumental music, and choral music groups.

Majored in a STEM Field

When aggregating all forms of music courses, music and nonmusic students majored in STEM fields at statistically indistinguishable rates. The probability for nonmusic students to major in a STEM field (.20) was lower than, but not statistically significantly lower (p = .17) than, the probability that music students would major in a STEM field (.22). In contrast to the overall analysis, instrumental students were significantly more likely than noninstrumental music to pursue a STEM field major (.24 probability for instrumental students vs. .20 for noninstrumental students), p = .04. Choral students showed virtually no difference (.206 probability) from nonchoral students (.204 probability), and this small difference was not statistically significant, p = .97.

Discussion

The results of the present study occupy an interesting position in the research literature examining the academic consequents of music education. With appropriate statistical control for preexisting population differences applied, the comparisons between music and nonmusic students were not the overwhelmingly positive results that tend to be trumpeted by music education advocates. Instead, the picture that emerges from this analysis is one of relative sameness: For most of the outcomes examined, music and nonmusic students were indistinguishable in terms of their transitions from secondary to postsecondary education. Thus, the present study does not meaningfully support logical postsecondary assumptions that might be made about the Canadian secondary students studied by Guhn et al. (2020) or the Australian elementary and secondary students studied by Martin and colleagues (2013).

And yet, neither are the results of the present study particularly disconfirming for potential academic benefits of arts education. Whereas the null results about the potential for transfer effects from music found in some experimental studies (e.g., Costa-Giomi, 2004; Mehr et al., 2013) suggest that there may be no benefit to music study beyond the important in and of itself goal of increased musical understanding, null results here indicate that students who spend some of their scarcest resource during high school—that is, time—on the formal study of music achieve postsecondary outcomes at least as good as their similar peers who did not choose to learn music during high school. In this way, the present study generally replicates the results found in a previous NCES longitudinal study with an earlier cohort of American students (Elpus, 2018). In other words, the present study provides some evidence that the results of the prior research yielding general sameness in postsecondary outcomes of music students and nonmusic students were not due to a unique cohort effect.

Both Elpus (2018) and Guhn et al. (2020) referred to the possible perception among school personnel and parents that election of music and arts courses in high school comes with a high “opportunity cost” in terms of burnishing a resume in anticipation of review by college admissions officers. In results that are replicated here, I previously suggested that there was likely “little to no ‘opportunity cost’ in the college admissions process to pursuing arts coursework” (Elpus, 2018, p. 119). In discussing their results showing greater academic benefits of music to students, Guhn and colleagues outright dismissed “the ‘opportunity cost’ hypothesis” as a reason that students should be steered away from music course work. And yet, the conflicting results found in the literature—some showing clear benefits of music study and others showing sameness—reveal an evidentiary base that remains in an early stage of development. Large-scale quasiexperimental studies using observational data, such as the present study, the Elpus study, and the Guhn et al. study, have not truly explicitly compared music students against those who deliberately spent elective time “doubling up” in math, science, or English. Although it is undoubtable that students who doubled up in these subjects are among the nonmusic students in the present study, in Elpus and in Guhn et al., the nonmusic group in all these studies also undeniably included students who did not elect to take music specifically because they were struggling academically. The present study makes some effort to address this situation through regression methods controlling for prior academic achievement. Still, the argument that music has no real opportunity cost against doubling up in math, science, or English would be bolstered by future research explicitly designed to compare music students directly to a comparison group comprising a matched set of students who were equally academically apt and who specifically chose to double up in a nonmusic subject. Such a comparison is possible with national longitudinal data currently available, and future research making this comparison seems warranted.

Conversely, the literature on the transition from K–12 to postsecondary schooling has also not yet explicitly analyzed outcomes for those students who enrolled in greater numbers of music courses. Many students in American high schools essentially double, triple, or quadruple up in music by pursuing more than the required number of courses in music performance; by adding nonperformance music electives such as music theory, music technology, or composition/songwriting to their high school course selection; or by pursuing ensemble performance across multiple modalities (Elpus & Abril, 2019). In the present study, I defined music students as those students who had taken at least one performance or nonperformance music course of any kind. This operationalization, although useful, involves a clear trade-off. The benefit of this operationalization is the clear “bright line” distinction between “music” and “nonmusic” students in that no student categorized as nonmusic received any school-based music instruction. This bright line yields a theoretically justifiable and useful counterfactual comparison between the groups. The trade-off’s loss is that the present analysis does not account for the intensity, breadth, or duration of music study undertaken by the music students. It is certainly possible that the transition-to-college outcomes examined here might vary among music students based on the “dosage” of music study. Future research examining the possibility of varied effects using a dosage analysis would be a welcome addition to this burgeoning line of literature.

Although the opportunity cost question may not yet be fully settled, the results of the present study are clear: In general, music students had the same rates of success in the transition to college as did nonmusic students. Importantly, music students were never at a disadvantage in the college application or admission process compared with their peers who were not in music. In some instances, certain music students found more success in the outcomes measured here. The main takeaway of this study—and the one with the most direct impact on American high school students—should be addressed to those adults charged with advising high schoolers on course selection and the college admissions process: No adult advising a high schooler should ever dissuade a student from pursuing music coursework under the misguided notion that the arts course will harm the student’s chances to get into a good college. Students who opted to spend their scarce elective high school time in music courses participated and succeeded in the college admissions process at the same rates as their nonmusic peers; students who opted to skip music courses were not advantaged in the college admissions process when compared to music students with similar academic and sociodemographic profiles. As crucial as these implications for high school course-taking and college admissions counseling are, this result is somewhat meaningless if knowledge of it is confined to the higher education audience for arts education research. This result needs to be widely shared with parents, guidance counselors, school administrators, and education policymakers to meaningfully impact students’ opportunities to study music in school.

Footnotes

Data Availability Statement

The data analyzed here are drawn from the restricted-use data set of the High School Longitudinal Study of 2009, a project of the U.S. Department of Education’s National Center for Education Statistics, a division of the Institute of Education Sciences (IES). A public-use version of the data set, lacking some of the variables used in this analysis, is freely available from . Restricted-use data are available to researchers under license from IES. Restricted-use data license holders may request statistical code to replicate all the analyses presented in this article from the author.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This analysis was supported in part by an award from the Research in the Arts program of the National Endowment for the Arts (NEA), Grant No. 1809950-38-18. The opinions expressed in this article are those of the author and do not represent the views of the NEA or the NEA Office of Research and Analysis. The NEA does not guarantee the accuracy or completeness of the information in this article and is not responsible for any consequence of its use. Additional funding support for this study was provided by the National Association for Music Education.

ORCID iD

Kenneth Elpus

Notes

Author Biography

Kenneth Elpus is associate professor and coordinator of music education at the University of Maryland, where he is principal investigator in the Music & Arts Education Data Lab (). His research interests include music in education policy, issues of demography and representation among music students and music teachers, and the social and academic consequents of music and arts education for K–12 students.

References

Catterall

J. S.

(1997). Involvement in the arts and success in secondary school (Vol. 1). Americans for the Arts.

Catterall

J. S.

(2009). Doing well and doing good by doing art. I-group Books.

Catterall

J. S.

Chapleau

Iwanaga

(1999). Involvement in the arts and human development: General involvement and intensive involvement in music and theatre arts. In Fiske

E. B.

(Ed.), Champions of change (pp. 1–18). Arts Education Partnership.

Catterall

J. S.

Dumais

S. A.

Hampden-Thompson

(2012). The arts and achievement in at-risk youth: Findings from four longitudinal studies (Research Report No. 55). National Endowment for the Arts.

Cohen

Sandy

S. V.

(2007). The social, emotional, and academic education of children: Theories, goals, methods, and assessments. In Bar-On

Maree

J. G.

Elias

M. J.

(Eds.), Educating people to be emotionally intelligent (pp. 63–77). Greenwood Publishing Group.

Costa-Giomi

(2004). Effects of three years of piano instruction on children’s academic achievement, school performance and self-esteem. Psychology of Music, 32(2), 139–152. https://doi.org/10.1177/0305735604041491

Deasy

R. J.

(Ed.). (2002). Critical links: Learning in the arts and student academic and social development. Arts Education Partnership.

Duprey

M. A.

Pratt

D. J.

Jewell

D. M.

Cominole

M. B.

Burns Fritch

Ritchie

E. A.

Rogers

J. E.

Wescott

J. D.

Wilson

D. H.

Christopher

E. M.

(2018). High School Longitudinal Study of 2009 (HSLS:09) base-year to second follow-up data file documentation. U.S. Department of Education.

Dwyer

M. C.

(2011). Reinvesting in arts education: Winning America’s future through creative schools. President’s Committee on the Arts and Humanities.

10.

Elpus

(2013). Is it the music or is it selection bias? A nationwide analysis of music and non-music students’ SAT scores. Journal of Research in Music Education, 61(2), 175–194. https://doi.org/10.1177/0022429413485601

11.

Elpus

(2017). Just ask me: Convergent validity of self-reported measures of music participation. Journal of Research in Music Education, 65(2), 129–138. https://doi.org/10.1177/0022429417713721

12.

Elpus

(2018). Estimating the effect of music and arts coursework on college admissions outcomes. Arts Education Policy Review, 119(3), 111–123. https://doi.org/10.1080/10632913.2016.1201030

13.

Elpus

Abril

C. R.

(2011). High school music ensemble students in the United States: A demographic profile. Journal of Research in Music Education, 59(2), 128–145. https://doi.org/10.1177/0022429411405207

14.

Elpus

Abril

C. R.

(2019). Who enrolls in high school music? A national profile of U.S. students, 2009–2013. Journal of Research in Music Education, 67(3), 323–338. https://doi.org/10.1177/0022429419862837

15.

Fitzpatrick

K. R.

(2006). The effect of instrumental music participation and socioeconomic status on Ohio fourth-, sixth-, and ninth-grade proficiency test performance. Journal of Research in Music Education, 54(1), 73–84. https://doi.org/10.1177/002242940605400106

16.

Fredricks

J. A.

Blumenfeld

P. C.

Friedel

Paris

A. H.

(2005). School engagement. In Moore

K. A.

Lippman

L. H.

(Eds.), What do children need to flourish? (pp. 305–321). Springer US.

17.

Fredricks

J. A.

Blumenfeld

P. C.

Paris

A. H.

(2004). School engagement: Potential of the concept, state of the evidence. Review of Educational Research, 74(1), 59–109. https://doi.org/10.3102/00346543074001059

18.

Fredricks

J. A.

Eccles

J. S.

(2005). Developmental benefits of extracurricular involvement: Do peer characteristics mediate the link between activities and youth outcomes? Journal of Youth and Adolescence, 34(6), 507–520. https://doi.org/10.1007/s10964-005-8933-5

19.

Gordon

R. L.

Fehd

H. M.

McCandliss

B. D.

(2015). Does music training enhance literacy skills? A meta-analysis. Frontiers in Psychology, 6, Article 1777. https://doi.org/10.3389/fpsyg.2015.01777

20.

Guhn

Emerson

S. D.

Gouzouasis

(2020). A population-level analysis of associations between school music participation and academic achievement. Journal of Educational Psychology, 112(2), 308–328. https://doi.org/10.1037/edu0000376

21.

Ilari

Helfter

Huynh

(2019). Associations between musical participation and young children’s prosocial behaviors. Journal of Research in Music Education, 67(4), 399–412. https://doi.org/10.1177/0022429419878169

22.

Johnson

C. M.

Memmott

J. E.

(2006). Examination of relationships between participation in school music programs of differing quality and standardized test results. Journal of Research in Music Education, 54(4), 293–307. https://doi.org/10.1177/002242940605400403

23.

Kirschner

Tomasello

(2010). Joint music making promotes prosocial behavior in 4-year-old children. Evolution and Human Behavior, 31(5), 354–364. https://doi.org/10.1016/j.evolhumbehav.2010.04.004

24.

Kraus

Chandrasekaran

(2010). Music training for the development of auditory skills. Nature Reviews Neuroscience, 11(8), 599–605. https://doi.org/10.1038/nrn2882

25.

Long

J. S.

Freese

(2014). Regression models for categorical dependent variables using Stata (3rd ed.). Stata Press.

26.

Martin

A. J.

Mansour

Anderson

Liem

G. A. D.

Sudmalis

(2013). The role of arts participation in students’ academic and nonacademic outcomes: A longitudinal study of school, home, and community factors. Journal of Educational Psychology, 105(3), 709–727. https://doi.org/10.1037/a0032795

27.

McCullagh

(1980). Regression models for ordinal data. Journal of the Royal Statistical Society: Series B (Methodological), 42(2), 109–127. https://doi.org/10.1111/j.2517-6161.1980.tb01109.x

28.

Mehr

S. A.

Schachner

Katz

R. C.

Spelke

E. S.

(2013). Two randomized trials provide no consistent evidence for nonmusical cognitive benefits of brief preschool music enrichment. PLoS ONE, 8(12), Article e82007. https://doi.org/10.1371/journal.pone.0082007.s002

29.

Moreno

Bialystok

Barac

Schellenberg

E. G.

Cepeda

N. J.

Chau

(2011). Short-term music training enhances verbal intelligence and executive function. Psychological Science, 22(11), 1425–1433. https://doi.org/10.1177/0956797611416999

30.

Nurmi

J.-E.

(2004). Socialization and self-development. In Lerner

R. M.

Sternberg

R. J.

(Eds.), Handbook of adolescent psychology (2nd ed., pp. 85–124). Wiley.

31.

Schellenberg

E. G.

Corrigall

K. A.

Dys

S. P.

Malti

(2015). Group music training and children’s prosocial skills. PLoS ONE, 10(10), Article e0141449. https://doi.org/10.1371/journal.pone.0141449

32.

Thomas

M. K.

Singh

Klopfenstein

(2015). Arts education and the high school dropout problem. Journal of Cultural Economics, 39, 327–389. https://doi.org/10.1007/s10824-014-9238-x

33.

Young

L. N.

Cordes

Winner

(2014). Arts involvement predicts academic achievement only when the child has a musical instrument. Educational Psychology, 34(7), 849–861. https://doi.org/10.1080/01443410.2013.785477