Predicting singing ability from parents’ parental support and openness in high school students: Using a self-adjusted Goldsmiths Musical Sophistication Index Inventory

Abstract

Cognitive investment theory states that personality can predict individuals’ behavior patterns on one specific activity, and further predict a related ability. Besides, personality itself is predicted by environmental factors throughout individuals’ development. The present research explored if and how parental support predicts singing ability through openness and musical training, using a sample of high school students in mainland China. A measurement tool for acquiring the musical training and singing ability of targeted sample was lacking. Therefore, an adjustment study was conducted, which adjusted the Goldsmiths Musical Sophistication Index–traditional Chinese version adjusted for Taiwanese (GMSI-TC), to obtain a suitable simplified Chinese version for the high school students (GMSI-SC-H). Thus, this research comprised two studies: Study 1 was an adjustment study, and Study 2 was a relation exploration study. Results of Study 1 showed that the GMSI-SC-H possessed acceptable reliability and validity on the general musical sophistication factor, Musical Training, Perceptual Abilities, and Singing Abilities sub-scales. The results of Study 2 showed that two mediating paths from parental support to singing ability were significant. Therefore, the conclusion was that parental support can predict singing ability through openness and musical training in a group of high school students. This research expands our understanding of links between personality and life outcomes.

Keywords

high school students musical sophistication inventory inventory adjustment parental support openness musical training singing ability

Singing is a very prevalent form of musical activity in music culture, occurring in different genres and guises (Hutchins et al., 2014). It has captured the attention of professionals in many areas, for example, some musical psychologists or music educators use singing ability as a dependent variable to explore the factors that affect it (Barrett et al., 2020; Welch et al., 2010).

Some psychological factors can predict singing ability. According to cognitive investment theory proposed by von Stumm et al. (2011), people who have a high level of openness (a Big Five dimension that includes curiosity, artistic interest, and imagination) tend to spend more time and energy on musical training, hence improving their singing ability. The personality trait openness can also be shaped by environmental factors in individuals’ development, such as the parental support individuals receive in formative years (Chen & Liu, 2018); therefore, parental support may predict individuals’ singing ability indirectly, through openness and musical training. In this research, the relations among parental support, openness, musical training, and singing ability were explored in a sample of high school students in mainland China.

Musical training/instruction and singing ability

Of all the possible factors that may affect singing ability, musical instruction or more specifically singing instruction has been the causal factor that has been focused on most frequently. A national singing program Sing Up (including long-term singing instruction) conducted in the United Kingdom showed that children who were in the program obtained significantly more improvement on singing ability compared with their non-Sing Up peers (Welch et al., 2009, 2010). Research conducted in 11 Australian primary schools also revealed that implementing generalist teacher-led music programs in early education settings can positively affect young children’s singing skills (Barrett et al., 2020). Svec (2018) has done a meta-analysis to summarize the effects of singing instructions on singing ability of children based on 34 studies, obtaining an identical conclusion.

Openness and musical ability/singing ability

Some researchers have linked openness with various musical abilities. Greenberg et al. (2015) found that aesthetic interest (a sub-trait of openness) positively predicts all five sub-scale scores on the Goldsmiths Musical Sophistication Index (GMSI), including Singing Abilities. Thomas et al. (2015) proposed a mediating model in which musical discrimination ability is predicted by openness fully through musical training. The theoretical foundation was cognitive investment theory, which indicates that the investing traits of individuals can result in more spending of time, energy, cognitive resources on certain activities, which then causes the improvement of some abilities (von Stumm et al., 2011). Specifically, if individuals have higher openness, it means that they are more curious and enthusiastic about learning various arts including music; thus, they should actively participate in musical training activities, during which their engagement should be more devoted. These investments of time and cognitive resources in music should ultimately improve their musical ability. Swaminathan and Schellenberg (2018) explored the predictability of openness and musical training to musical competence (defined as music perceptual abilities). A similar result was obtained, which indicates that musical training partially mediates openness and musical competence. Jankovic and Bogaerts (2021) did research on high school music students and discovered that openness significantly predicts school music performance.

Figure 1.

The Hypothetical Mediation Model.

Parental support and singing ability

No research has been done to specifically explore the relation between parental support and singing ability. Nevertheless, some research has shown that parental support can significantly predict children’ ability development and academic performances (Christian, 2020; Peng et al., 2022), such as instrumental musical performances (Oliveira et al., 2021). This may be because parents offer enough encouragement (emotional support), appropriate suggestions (informational support), and so on, which makes children engage in learning activities (such as musical training) more wholeheartedly, which then improves their abilities. Therefore, parental support may predict singing ability through musical training.

Besides, some research has associated parental support with openness. Chen and Liu (2018) found that the emotional support parents offer to their teenage children, as a part of positive parenting, can positively predict children’s openness. Wang et al. (2006) did an investigation on teenagers, and reached an identical conclusion. Using a longitudinal approach, Schofield et al. (2012) found that teenagers’ positive personalities could be predicted by positive parenting, including sufficient emotional support. Therefore, combined with the cognitive investment theory, it is likely that parental support can predict singing ability through openness.

The present research

The necessity of inventory adjustment

In this investigation, the relations among four variables were explored in a sample of Chinese high school students. For measuring musical training and singing ability on a large scale efficiently, an appropriate self-assessing inventory is needed. The GMSI, an inventory that was designed to measure various musical activity patterns or abilities, has been adjusted in several languages including traditional Chinese (adjusted with Taiwanese sample; Lin et al., 2019), but directly using the traditional Chinese version (abbreviated as GMSI-TC) on high school students in the mainland of China was unsuitable for two reasons. First, there were differences in character-shapes and writing expression. Traditional and simplified Chinese are two writing systems for the same language; Taiwan region uses the former system and the mainland mainly uses the latter. These two systems are similar but still have differences in some character shapes and expressions of some concepts. The second reason relates to the particularity of the targeted group of the present study. The participants here were high school students in mainland China. Due to exam stress and less emphasis on music education, they have limited time and resources to invest in music-related activities. As a result, directly using GMSI-TC here can cause a severe floor effect on the Active Engagement and Musical Training dimensions. Therefore, it was necessary to adjust in a new version of the GMSI specifically for high school students in mainland China (GMSI Simplified Chinese Version for High School Students, abbreviated as GMSI-SC-H), based on the GMSI-TC.

Four hypothetical paths from parental support to singing ability

Parental support–openness–musical training–singing ability

Individuals’ personalities are influenced in a complex way by genetic and environmental factors (Johnson & Krueger, 2004), and how much teenagers’ parents offer them in emotional support throughout their development can shape their openness (Chen & Liu, 2018). In addition, there are more kinds of supports that parents offer to their children besides emotional support, such as informational, appraisal, and instrumental support (Munoz-Laboy et al., 2014). Informational support indicates parents give children appropriate suggestions or information to help them solve problems, appraisal support means parents offer suitable evaluations for children’s performances, and instrumental support indicates parents offer practical help or materials for their kids. All of them may improve openness in different ways. For example, parents who give proper suggestions and appraisals in children’s exploring and learning activities, may facilitate their success, making them more confident, which may improve their general motivation to explore and learn new things (openness). Or parents may offer children environments with abundant resources (books, musical instruments, etc.), facilitating children exploring and learning new knowledge and skills, enlightening their general curiosity to new things. Also, as Thomas et al. (2015) state, openness is a typical cognitive investing trait. People who are more curious about new skills and knowledge, or are more interested in the arts, tend to spend more resources on musical training, which can consequently increase individuals’ musical discriminating ability. Similarly, advanced singing ability also requires plenty of practice, which comprises a great deal of cognitive processing. Thus, it is reasonable to say that openness as an investing trait can enhance the level of formal musical training, which, in turn, improves singing ability. Therefore, in the present research, it was hypothesized that parental support significantly predicts high school teenagers’ singing ability, through openness and musical training in sequence.

Parental support–openness–singing ability

Beyond formal training, individuals with high openness should engage in more music-related activities which are discernible, such as listening to concerts, or more obscure ones like experiencing inner music (Beaty et al., 2013). These informal activities probably also have positive impacts on singing ability. Swaminathan and Schellenberg (2018) also found that a direct path exists between openness and musical competence. Therefore, a direct path should exist between openness and singing ability, representing the mediating effects of other musical activities except formal training.

Parental support–musical training–singing ability

Due to the lack of professional background, most parents cannot offer appropriate informational or appraisal support for their children on musical training specifically. But they may offer emotional and instrumental support to enhance high school students’ singing ability through musical training, for instance, by actively encouraging children to learn and practice, or offering money for children to take courses. Therefore, it is possible that parental support predicts singing ability through musical training.

Parental support–singing ability

Parents may offer a good material environment for their children to freely explore and implicitly learn (not through musical training) about music, which may lead to an improvement in their singing ability. Therefore, it was also hypothesized that a direct effect may exist between parental support and singing ability.

Thus, this research comprises two studies. Study 1 involved adapting GMSI-TC into a simplified Chinese version for high school students in mainland China (GMSI-SC-H). Study 2 involved testing the hypothetical mediation model (Figure 1) of four variables.

Study 1

Method

Character-shape conversion, words substitution, and alteration

The aims of this work include: (1) keeping the factor structure of GMSI and GMSI-TC; (2) keeping the number of scoring items unchanged (38 items), and ensuring that the meaning of each item is in accordance with the items in GMSI and GMSI-TC; and (3) altering the ways of description of some items to ensure suitability, on the basis of unchanging meanings.

Traditional and simplified Chinese are two similar character systems of the same language. Most pairs of characters or words in traditional and simplified Chinese corresponding by meanings are identical or only differ from each other in character shapes, Therefore, for most of the inventory, directly converting the traditional character-shape to simplified character-shape is valid. Whereas for some concepts that appeared in recent decades (such as the internet), pairs of characters or words direct to the same concepts can be very different from each other which can cause confusion in understanding. Words of this kind in the inventory were substituted.

Due to the aims and features stated above, this process included three steps: character-shape conversion, word substitution, and alteration. The character shapes conversion step involved converting the character shapes of traditional Chinese words that only differ from simplified Chinese words on character shapes, for example, converting “練習” to “练习” (both meaning “practice”). The word substitution step involved substituting traditional Chinese words that are very different from simplified Chinese words, for example, substituting “部落格” with “博客” (both meaning “blog”). The alteration step was for altering some items’ descriptions to fit the participants of the present research, for example, the item “I have had 0/0.5/1/2/3–5/6–9/10 or more years of formal training on a musical instrument (including voice) during my lifetime,” was altered to “0/1 month/3 months/0.5 years/1 year/2–3 years/4 or more years of formal training . . .”, to avoid a possible floor effect due to the lack of emphasis and resources on musical education in mainland China. The full GMSI-SC-H, and GMSI-TC are included in Supplementary Materials online.

Participants

A total of 1,035 students from high school A in a county, and high school B in a city of mainland China participated in this study. The average age was about 16 years old. A total of 808 participants from high school A completed the GMSI-SC-H, and 740 of them were screened as valid (validity rate = 91.6%, 57.7% were females). This sample was used for obtaining Cronbach’s α and the factor structure (confirmatory factor analysis—CFA). A subset of 194 of these 740 students completed the GMSI-SC-H again 3 months later for evaluating test–retest reliability, in which 183 of them were screened as valid (validity rate = 94.3%, 53.0% were females). A total of 227 Participants from high school B completed the GMSI-SC-H, the musical sub-scale of High School Students Multiple Intelligence Self-Rating Scale, and 202 of them were screened as valid (validity rate = 89.0%, 45.5% were females). This sample was used to obtain convergent and divergent validity.

Procedure

Inventory completions were conducted three times. Participants and their parents gave informed consent verbally before completing. The first investigation was conducted in high school A. A paper version of GMSI-SC-H was completed by the students under the instructions and supervision of the researcher and teachers. Before completing, students were told that the inventory includes questions involving self-evaluation of musical abilities. They needed to compare themselves with schoolmates and peers. For example, one student gave herself a rating of 7 (highest in a 7-point Likert-type scale) on singing intonation, meaning that she thought she is one of the best on this among her peers. The process took about 15 min. The data screening standards were: (1) Participants who missed more than two items (> 2) (excluded items of basic information) were excluded; (2) Participants who consecutively filled the same choices more than ten items (>10) were excluded.

In the second investigation 3 months later, five classes of students from the first investigation were re-recruited. Participants received the identical instruction and completed the GMSI-SC-H which took about 15 min. The data screening standards were identical to the first investigation on the first two clauses; furthermore, participants who completed the second measurement validly only, were also excluded.

The third investigation was conducted in high school B. Five classes of students were recruited. Students completed the GMSI-SC-H and the musical intelligence sub-scale of High School Students Multiple Intelligence Self-Rating Scale (Yuan, 2010) under instruction. They were also told to compare themselves with peers while doing the ability evaluation questions. The whole process took about 20 min. The data screening standards were consistent with the first investigation on the first two clauses, besides that participants who answered the detecting items incorrectly were excluded.

Measurements

GMSI-SC-H inventory

The GMSI-SC-H Inventory involves judgments on a 7-point Likert-type scale. It contains five dimensions, namely, Active Engagement (nine items), Emotions (six items), Musical Training (seven items), Perceptual Abilities (nine items), and Singing Abilities (seven items). The General Musical Sophistication index is calculated by summing the scores on all items.

High school students multiple intelligence self-rating scale, musical intelligence sub-scale

This inventory was developed according to Gardner’s Multiple Intelligence Theory. It contains 8 sub-scales including a Musical Intelligence dimension (10 items for each dimension), on a 6-point Likert-type scale (Yuan, 2010).

Analysis

The missing item scores were supplemented with the mean scores of the test items in the same sub-scale. Most of the psychometric indices were calculated with SPSS 25.0. The CFA was realized with AMOS 21.0.

Results

Reliability

The internal consistency of inventory was indicated by Cronbach’s α. As Table 1 shows, all the five dimensions and the General Musical Sophistication factor possessed adequate to very good internal reliability, ranging from .737 to .920. The test–retest reliability indicators of five sub-scales were generally lower than internal consistency, ranging from .640 to .806.

Table 1.

Cronbach’s α and Test–Retest Reliability for General Factor and Each Sub-Scale.

Dimension	Cronbach’s α (n = 740)	Test–retest reliability (n = 183)
Active Engagement	.825	.679
Emotion	.774	.640
Musical Training	.737	.806
Perceptual Abilities	.805	.704
Singing Abilities	.788	.741
Musical Sophistication	.920	.799

Validity

For exploring the best model fit of the GMSI, Müllensiefen et al. (2014) proposed four hypothetical models as candidates. Model 1 is a hierarchical model which indicates that a general factor affects five group factors, which influence their own group items. Model 2 is a bi-factor model, in which each item is partially influenced by a general factor and its own group factor. In Model 3, the general factor does not exist; all items are affected by their own group factors only and all the group factors are independent of each other. Model 4 assumes that the general factor does not exist either, but five group factors are intercorrelated with each other. Model 2 is illustrated in Figure 2.

Figure 2.

Model 2: Bi-Factor Model.

As Table 2 shows, following Müllensiefen et al. (2014) and Lin et al. (2019), the fitness indices used here were χ², Bayesian information criterion (BIC), Tucker–Lewis index (TLI), comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR). For a large sample, the χ² value is usually significant although the model is fit (Cao & Zhang, 2017). Therefore, it is inappropriate to take χ² as an index of absolute fit, but using it to indicate relative fit is acceptable. On relative fit indices, Model 2 obtained the lowest χ² and BIC in four candidates, revealing that Model 2 was the fittest. On absolute indices, the RMSEA and SRMR of Model 2 were the only ones below the strict adequate threshold of .05. The CFI of Model 2 was the only one beyond .90. These three values indicated that Model 2 had a good absolute fit. Model 2’ s TLI was the highest, but did not exceed the recommended cut-off value of .90. When a CFA’s number of factors is more than 5 and the degrees of freedom is very large, the TLI value can hardly exceed .90. Therefore, it is recommended that a TLI exceeding .850 can also reveal the adequacy of model fit (Marsh et al., 2010). Accordingly, the factor structure that had been shown to be the best fit in studies of Müllensiefen et al. (2014) and Lin et al. (2019) still held the best in the present study.

Table 2.

The CFA Results of Four Hypothetical Models.

Name	Model 1	Model 2	Model 3	Model 4
Name	Hierarchical	Bi-factor	Simple factor	Correlated factor
χ²	2,040.051	1,485.376	3,426.483	1,984.767
df	660	627	665	655
BIC	2,575.190	2,238.534	3,928.589	2,552.939
TLI	.842	.896	.686	.846
CFI	.851	.908	.703	.857
RMSEA	.053	.043	.075	.052
SRMR	.0680	.0440	.2104	.0653

Note. N = 740. CFA = confirmatory factor analysis; BIC = Bayesian information criterion; TLI = Tucker–Lewis index; CFI = comparative fit index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual.

The correlations between musical intelligence and all GMSI dimensions are displayed in Table 3. All dimensions of GMSI were significantly correlated with the Musical Intelligence score at the .01 level.

Table 3.

Correlations Between GMSI and Musical Intelligence Sub-Scale of High School Students Multiple Intelligence Self-Rating Scale.

Dimension or Scale	Active Engagement	Emotion	Musical Training	Perceptual Ability	Singing Ability	Musical Sophistication
Musical Intelligence	.685**	.716**	.584**	.753**	.736**	.891**

Note. N = 202. GMSI = Goldsmiths Musical Sophistication Index.

p < .01.

Discussion

In this study, an adjusted version of GMSI was created, named the GMSI-SC-H. Indices of reliability and validity were acquired. Firstly, based on a sample of 740 students, Cronbach’s α indicated that GMSI-SC-H has a qualified internal consistency for each dimension and the general factor. Whereas on test–retest reliability, according to a sample of 183 students, the results revealed that the general musical sophistication factor score, the Musical Training, Perceptual Abilities, and Singing Abilities had qualified test–retest reliability for use (> .7), but the Musical Engagement and Emotion are not stable enough on time due to their low test–retest reliability (< .7). Thus, care should be taken while using these two sub-scales. Compared with the adaptation research of Lin et al. (2019), the test–retest indicators here were obviously lower (.640–.806 vs .76–.93). This was probably due to the longer interval between two tests (3 months vs 3 weeks, longer interval usually causes lower test–retest reliability) and the lower ages of participants (15–18 years old teenagers vs 13–65 years old). Compared with older participants, teenagers are less educated and less mature in intelligence, their self-cognition is more unstable. Also, it is more likely that teenagers’ behavior modes change substantially during a short period. Thus, stable response may be less likely for high school students (Zhang et al., 2022).

For the validity part, the bi-factor model was successfully replicated here, which proved that the GMSI-SC-H maintains a good factor structure. According to a sample of 202 students, the GMSI scores were significantly correlated to the musical intelligence scores, reaching .891 (p < .01). Musical Intelligence was defined by Gardener as music perception and production abilities, and sensitivity to sounds (Stanford, 2003), which is similar to musical sophistication (referring to many sorts of musical abilities and activity patterns). Therefore, the high correlation between them can be considered as good evidence of high convergent validity. As for each dimension of the GMSI-SC-H, Musical Intelligence was most strongly correlated with Singing Abilities and Perceptual Abilities among the five sub-scales (.736 and .753, respectively). This can also be regarded as good evidence for convergent validity in that they indicate the capacities and skills aspect of musical sophistication. On the contrary, Musical Training and Active Engagement had the weakest correlations with Musical Intelligence (.584 and .685, respectively). This can be used to verify divergent validity because Musical Training and Active Engagement are designed to measure music-related behavior (Müllensiefen et al., 2014).

Study 2

Method

Participants

A total of 636 students from high school C in a county of mainland China participated in the investigation 552 were screened as a valid sample, 60.5% were females, and the average age was 16.391 ± 1.010 years old.

Procedure

Fourteen classes of students were recruited. Participants and their parents gave informed consent verbally before completion. All students completed the investigation through an internet survey platform (wjx.cn) in computer classrooms. The entire investigation included Parental Support sub-scale of Child and Adolescent Social Support Scale (CASSS), Big Five Inventory–2 (BFI-2), and GMSI-SC-H. Students were also told that the GMSI-SC-H includes questions of self-evaluation on musical abilities which needed them to compare themselves with their peers. The whole process took about 30 min. Data screening standards were (1) participants who completed the investigation in less than 508 s (on average less than 4 s for each item) were discarded and (2) participants who answered the detecting items incorrectly were discarded.

Measurements

GMSI-SC-H Inventory

This is the adjusted inventory from Study 1. According to the conclusion of Study 1, Musical Training and Singing Abilities dimensions are reliable and valid. Calculating with the sample of Study 2, the Cronbach’s α of these two dimensions were .745 and .757, respectively.

BFI-2 Chinese version

The original Big Five Inventory–2 was developed by Soto and John (2017), and the Chinese version was adjusted by Zhang et al. (2022). All items are judged on a 5-point Likert-type scale and there are 12 for each domain. In the adjustment work, a sample of high school students (N = 315) was used. The results showed that the BFI-2 Chinese version is reliable and valid on the domain level, α (openness) = .791 (using the sample from Study 2).

Parental support sub-scale of CASSS Chinese version

The original Child and Adolescent Social Support Scale was developed by Malecki and Demaray (2003), and the Chinese version was adjusted by Luo et al. (2017). CASSS contains 12 items for each sub-scale, for measuring five sources of support: Parents, Teachers, Schoolmates, Friends, and School. Each source of support includes four types of support, categorized based on the ways that support is given: Emotional, Informational, Appraisal, and Instrumental. All items are judged on a 4-point Likert-type scale. The overall general support factor and the parental support sub-scale had good reliability and criteria validity, α (parental support) = .916 (using the sample from Study 2).

Analysis

First, common method bias was tested utilizing Harman’s single-factor test (Ye et al., 2017). Second, descriptive data and simple correlations were acquired. Third, a hierarchical multiple regression analysis was performed in three steps.

Fourth, a mediating analysis was conducted for testing the mediating hypotheses. Two methods were utilized, they were the classical Causal Steps Approach proposed by Baron and Kenny (Wen & Ye, 2014) and the prevalent Bias-Corrected Bootstrap Approach with maximum-likelihood estimation and 5,000 bootstrap samples.

Finally, after the mediating analysis, some paths may appear insignificant. If this is the case, the insignificant paths should be removed and new path coefficients should be calculated, and the fitness of the new model is also tested.

All the analyses except the last were performed by SPSS 25.0. The mediating analysis was performed by the plug-in package PROCESS 3.5 of SPSS, the last analysis was performed by AMOS 21.0.

Results

Common method bias test

Eight factors’ eigenvalues exceeded 1.0, and the factor with the highest eigenvalue accounted for 20.26% of the total variance, which was far less than the threshold of 40%, Therefore, common method bias was insignificant, which meant using identical methods did not distort the results.

Simple correlations among variables

As Table 4 shows, r between each pair of variables was significant at .01 level except the r between neuroticism and musical training. This result laid a good foundation for further explorations.

Table 4.

Simple Correlations Between Explored Variables.

Variables	1	2	3	4	5	6	7	8
1. Parental support	1.000
2. Extraversion	.334**	1.000
3. Agreeableness	.315**	.376**	1.000
4. Conscientiousness	.327**	.364**	.579**	1.000
5. Neuroticism	−.279**	−.442**	−.416**	−.511**	1.000
6. Openness	.256**	.474**	.383**	.449**	−.275**	1.000
7. Musical Training	.142**	.271**	.174**	.192**	−.069	.379**	1.000
8. Singing Ability	.160**	.289**	.215**	.197**	−.127**	.466**	.457**	1.000
M	44.658	37.413	44.402	41.578	35.295	41.871	16.402	28.397
SD	10.857	7.151	5.293	6.504	7.679	6.310	6.122	5.762

Note. N = 552.

p < .01.

Hierarchical multiple regression in three steps

As Table 5 shows, parental support and gender were entered in Step 1 of analysis, both factors significantly predicted singing ability. In Step 2, five domains of BFI-2 were added into the regression. Openness significantly predicted singing ability in Step 2, whereas the other four domains were insignificant. According to large-scale research of retesting many personality trait-life outcomes links, controlling for overlap between personality traits substantially decreased the strength of many associations (Soto, 2021). Thus, for obtaining a more reliable conclusion, the other four dimensions of the Big Five needed to be considered as covariates. The gender factor remained significant, but the previously significant factor of parental support was now insignificant, which indicated that openness possibly acted as a mediating variable between parental support and singing ability. In Step 3, musical training was entered into the regression. Gender, openness, and musical training significantly predict singing ability, and openness was the strongest predictor. Females and males were coded as 2 and 1 respectively, indicating that girls’ singing ability was significantly higher than boys.

Table 5.

Results of Hierarchical Multiple Regression.

	Standard partial regression coefficients
Variables	Step 1	Step 2	Step 3
Gender	.153***	.173***	.123**
Parental support	.176***	.046	.033
Extraversion		.088	.046
Agreeableness		.018	.022
Conscientiousness		−.040	−.055
Neuroticism		.007	−.019
Openness		.439***	.343***
Musical Training			.298***
Adjusted R²	.045	.246	.318

Note. N = 552.

p < .01; ***p < .001.

For the three independent variables that were included in the mediation model, the overall adjusted R² was 30.6%, exceeding the threshold of 30.0%, revealing that it is a reasonably predictive model (Cao & Zhang, 2017).

Mediating analysis

Two methods were used as mentioned for testing the mediating model: The result of the Causal Steps Approach showed that the total effect from parental support to singing ability was significant (simple regression coefficient b = .160, p < .001). As for the effect of each path illustrated in Figure 3, the path coefficient from parental support to openness was the standardized simple regression coefficient from the former to the latter. While musical training or singing ability was regarded as the dependent variable, the path coefficient was the standardized partial regression coefficient from independent variables to their corresponding dependent variable. The long path (parental support–openness–musical training–singing ability) and the short path (parental support–openness–singing ability) were significant, but the direct effect from parental support to singing ability and the other short path (parental support–musical training–singing ability) were insignificant. The e1, e2, and e3 represent the residual variances of the regressions. Identical results were obtained using Bias-Corrected Bootstrapping as shown in Table 6.

Figure 3.

The Mediation Model.

Table 6.

Results of Bias-Corrected Bootstrapping With 95% Confidence Level.

Paths	Completely standardized indirect effect	Standard error of indirect effect estimation	Lower limit of confidence interval	Upper limit of confidence interval
PS–OPEN–SA	.086	.017	.055	.122
PS–MT–SA	.016	.015	−.013	.046
PS–OPEN–MT–SA	.031	.007	.019	.045

Note. N = 552, confidence level = .95, bootstrap sample = 5,000. PS = parental support; OPEN = openness; SA = singing ability; MT = musical training.

Model fitness test after removing insignificant paths

Two insignificant paths were removed and the fitness of the remaining model was tested. As Table 7 shows, all absolute fit indices indicated that the significant paths model possessed good fitness. Also, as Figure 4 shows, there was little change in the path coefficients compared with the saturated model, revealing the stability of the mediation model.

Table 7.

Fitness Indices of Mediating Model After Removing Insignificant Paths.

Indices	χ²	df	p	RMSEA	GFI	AGFI	CFI	NFI	TLI	RFI
Values	1.998	2	.368	.000	.998	.991	1.000	.994	1.000	.982

Note. N = 552. RMSEA = root mean square error of approximation; GFI = goodness of fit index; AGFI = adjusted goodness of fit index; CFI = comparative fit index; NFI = normed fit index; TLI = Tucker–Lewis index; RFI: relative fit index.

Figure 4.

The Mediation Model After Removing Insignificant Paths.

Discussion

Significant path “parental support–openness–musical training–singing ability.”

Chen and Liu (2018) underlined that as a vital part of positive parenting, abundant emotional support parents offer to children can promote the well development of openness. When children are frustrated, parents give comfort and encouragement, making their children believe that they are loved and accepted. Or parents offer their children a safe base and encourage them to explore the world. With all these emotional supports, children can develop a healthy sense of control, security, and trust in circumstances (Barlow et al., 2009). Therefore, they will be more willing to explore the external world, more willing to spend time and energy on learning new things such as various arts (a vital aspect of openness). Besides, parental support here is not limited to emotional, but also includes informational, appraisal, and instrumental. On the informational aspect, if parents offer children good suggestions, encouraging them to overcome difficulties, their self-efficacy will improve, then a willingness of learning new things including various arts is reinforced. Appraisal support indicates that parents give children appropriate appraisals, which can make children learn from experience and simultaneously maintain their general curiosity and motivation of exploration. Lastly, instrumental means practical or material support, the abundant material environment families offer to children is good for the appearance of exploring behaviors, therefore enhancing the development of openness. Accordingly, all these supports can probably more or less promote the development of openness.

Openness, as an important personality domain, influences many aspects of individuals’ behaviors and life outcomes. The cognitive investment theory proposed by von Stumm et al. (2011) states that investing personality traits result in tendencies to invest more time and cognitive resources on certain activities, which in turn improve some abilities. Specifically, individuals who possess a high level of openness are more curious about new knowledge and skills, and they have stronger motivations to pursue aesthetics and arts, Therefore, they may spend more time on musical training spontaneously, and be fully engaged during the learning process, including using plentiful cognitive resources to deeply process musical stimuli, which lays a good foundation for output activities like singing. They probably would also participate in singing training specifically. All these investments would ultimately improve their singing ability.

Significant path “parental support–openness–singing ability.”

As discussed, parental support predicts openness significantly. For the significant direct effect from openness to singing ability, the explanation was that openness does not only enhance singing ability through formal musical training, but also positively influences singing ability through other various music-related activities. Obvious ones include attending concerts, spontaneously listening to music, obscure ones include actively capturing and processing musical stimuli in the surrounding environment, and actively imagining music, while no musical stimulus exists (inner music; Beaty et al., 2013). Individuals with a high level of openness are also more possible to do the activities mentioned above, therefore enhancing the development of their singing ability.

Insignificant path “parental support- music training- singing ability.”

With a sample of high school students in mainland China, the present research showed that parental support does not significantly predict musical training. This is likely due to two reasons. First, most parents lack professional knowledge and experience in musical training, so they cannot offer good suggestions (informational), or accurate comments (appraisal), The support they can give is encouragement (emotional) and enough material (instrumental), which means that compared with the abundant supports parents can offer on cultivating openness through different activities, the supports parents can offer on musical training directly were far less, which results in the insignificant effect from parental support to musical training. Second, the participants in the present research were high school students in mainland China, where musical education has not generally gained much emphasis, especially in less developed regions. Music is not a subject of college entrance exams, and most Chinese students do not plan to be professional musicians. Thus, most parents do not pay much attention to musical education. Even though they send their children to learn some music, it is only for letting their children have leisure-time hobbies. Therefore, most parents do not offer their children much support on musical training specifically. For example, they seldom encourage their children to learn music, or actively suggest their children on musical training.

Insignificant direct effect from parental support to singing ability

For the insignificant direct effect from parental support to singing ability, there are also two similar possible reasons: First, for the direct effect, the most possible support parents can give is instrumental, but not the other three types—offering a material environment (e.g., musical recordings or magazines), letting children improve their singing ability through spontaneous exploration and attempts (not formal training), which means that the support that parents can offer is even less. Thus, the direct effect between parental support and singing ability was also insignificant. Second, most parents in mainland China do not attach importance to musical education, thus they do not offer their children much support for the improvement of singing ability.

General discussion

In the development research of GMSI (Müllensiefen et al., 2014) and the adjustment research of GMSI-TC (Lin et al., 2019), they utilized a Melodic Memory Test (Harrison et al., 2016) and a Beat Perception Test (Iversen & Patel, 2008) to measure all participants, and correlated them with the results of self-reported inventory, as the evidence for convergent and divergent validity. Study 1 did not include objective listening tests. Therefore, future research should follow the procedure of using objective listening tests to test the validity of GMSI-SC-H. An available approach is using software to present musical stimuli and collect listening test data, and offering earphones to the participants and ensuring quietness of the environment. The selected objective listening tests must correspond to the participants’ capacity, in case of ceiling or floor effect.

The links between personality and life outcomes is an important domain of personality psychology. The present research used high school students as participants, associating openness with singing ability, expanding our understanding of the links between personality and life outcomes. Personality influences individuals’ specific behaviors, therefore influencing various life outcomes (Ozer & Benet, 2006). Theoretically, exploring personality-life outcomes links can help researchers to understand the essences and structures of personality more deeply. Practically, associating personality with life outcomes is the first step toward revealing the causal mechanism between personality and life outcomes. If the causation mechanism is clear, educators can know what kind of personality pattern needs to be cultivated to obtain an ideal life outcome. It has been shown that the majority of research on links between personality and life outcomes has good replicability and generality (Soto, 2019; Soto, 2021), which further confirmed the value of this domain of research.

After discovering that openness significantly predicts singing ability, it is also meaningful to find the predicting variables of openness for laying the foundation of causation research and intervention practice. Previous research has associated some environmental factors with openness, including parents’ education levels (Jonassaint et al., 2011) and parenting style (Chen & Liu, 2018; Reti et al., 2002). In Chen and Liu’s research, they emphasized parental support as an important part of positive parenting for cultivating openness. The present research separated parental support from positive parenting and revealed that parental support can significantly predict openness (directly) and singing ability (indirectly). Study 2 used a cross-sectional approach to explore the relations among four variables, thus causation could not be confirmed. For obtaining definite causal relations, a longitudinal approach needs to be used for further research.

Conclusion

The conclusion of Study 1 was that the GMSI-SC-H, adjusted on the basis of GMSI-TC has acceptable psychometric properties on the general Musical Sophistication factor, the Musical Training, Perceptual Abilities, and Singing Abilities sub-scales. Therefore, they are reliable and valid, whereas for the other two sub-scales, Active Engagement and Emotion, care should be taken while using them, due to their unqualified stability over time. The conclusion of Study 2 was that in the group of high school students, parental support significantly predicts singing ability through two paths, the long path is through openness and musical training in sequence, the short path is through openness.

Supplemental Material

sj-docx-1-pom-10.1177_03057356221141928 – Supplemental material for Predicting singing ability from parents’ parental support and openness in high school students: Using a self-adjusted Goldsmiths Musical Sophistication Index Inventory

Supplemental material, sj-docx-1-pom-10.1177_03057356221141928 for Predicting singing ability from parents’ parental support and openness in high school students: Using a self-adjusted Goldsmiths Musical Sophistication Index Inventory by Ye Yuan, Xiao Jie Yu and Xiao Wei Yuan in Psychology of Music

Footnotes

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Ye Yuan

Xiao Jie Yu

Supplemental material

Supplemental material for this article is available online.

References

Barlow

D. H.

Durand

V. M.

Stewart

S. H.

(2009). Abnormal psychology: An integrative approach. Nelson Education.

Barrett

M. S.

Zhukov

Brown

J. E.

Welch

G. F.

(2020). Evaluating the impact of a generalist teacher-led music program on early childhood school children’s singing skills and attitudes to music. Psychology of Music, 48(1), 120–136.

Beaty

R. E.

Burgin

C. J.

Nusbaum

E. C.

Kwapil

T. R.

Hodges

D. A.

Silvia

P. J.

(2013). Music to the inner ears: Exploring individual differences in musical imagery. Consciousness and Cognition, 22(4), 1163–1173.

Cao

Y. W.

Zhang

Y. P.

(2017). Multivariate statistical methods in psychology and education. Peking University Press.

Chen

H. B.

Liu

(2018). The relations between family socioeconomic status and teenagers’ wisdom: The mediating roles of positive parenting style and openness personality. Psychological Development and Education, 34(5), 558–566.

Christian

(2020). Parenting style and parental support on learners’ academic achievement. Journal of Sociology and Social Anthropology, 11, 198–205.

Greenberg

D. M.

Müllensiefen

Lamb

M. E.

Rentfrow

P. J.

(2015). Personality predicts musical sophistication. Journal of Research in Personality, 58, 154–158.

Harrison

P. M. C.

Musil

J. J.

Müllensiefen

(2016). Modelling melodic discrimination tests: Descriptive and explanatory approaches. Journal of New Music Research, 45, 265–280.

Hutchins

Larrouy-Maestri

Peretz

(2014). Singing ability is rooted in vocal-motor control of pitch. Attention, Perception, & Psychophysics, 76(8), 2522–2530.

10.

Iversen

J. R.

Patel

A. D.

(2008). The Beat Alignment Test (BAT): Surveying beat processing abilities in the general population. In Miyazaki

Hiraga

Adachi

Nakajima

Tsuzaki

(Eds.), Proceedings of the 9th international symposium on computer music modelling and retrieval (pp. 311–318). Queen Mary University of London.

11.

Jankovic

Bogaerts

(2021). Predicting success in the performing arts: Ballet and music. Psychology of Music, 49(4), 945–957.

12.

Johnson

Krueger

R. F.

(2004). Genetic and environmental structure of adjectives describing the domains of the Big Five Model of personality: A nationwide U.S. twin study. Journal of Research in Personality, 38(5), 448–472.

13.

Jonassaint

C. R.

Siegler

I. C.

Barefoot

J. C.

Edwards

C. L.

Williams

R. B.

(2011). Low life course socioeconomic status (SES) is associated with negative NEO PI-R personality patterns. International Journal of Behavioral Medicine, 18(1), 13–21.

14.

Lin

Kopiez

Müllensiefen

Wolf

(2019). The Chinese version of the Gold-MSI: Adaptation and validation of an inventory for the measurement of musical sophistication in a Taiwanese sample. Musicae Scientiae, 25(2), 226–251.

15.

Luo

X. F.

Chen

Q. S.

S. K.

(2017). Child and adolescent social support scale: Validation and preliminary application. Chinese Journal of Clinical Psychology, 25(4), 671–674.

16.

Malecki

Demaray

(2003). What type of support do they need? Investigating student adjustment as related to emotional, informational, appraisal, and instrumental support. School Psychology Quarterly, 18, 231–252.

17.

Marsh

H. W.

Lüdtke

Muthén

Asparouhov

Morin

A. J. S.

Trautwein

Nagengast

(2010). A new look at the Big Five factor structure through exploratory structural equation modeling. Psychological Assessment, 22(3), 471–491.

18.

Müllensiefen

Gingras

Musil

Stewart

(2014). The musicality of non-musicians: An index for assessing musical sophistication in the general population. PLOS ONE, 9(2), Article e89642.

19.

Munoz-Laboy

Severson

Perry

Guilamo-Ramos

(2014). Differential impact of types of social support in the mental health of formerly incarcerated Latino men. American Journal of Men’s Health, 8(3), 226–239.

20.

Oliveira

McPherson

Mota-Ribeiro

Oliveira-Silva

(2021). Musical achievement during a lockdown: The parental support miracle. Research Studies in Music Education, 9, 1–16.

21.

Ozer

Benet

(2006). Personality and the prediction of consequential outcomes. Annual Review of Psychology, 57, 401–421.

22.

Peng

X. J.

Sun

X. N.

(2022). Influence mechanism of teacher support and parent support on the academic achievement of secondary vocational students. Frontiers in Psychology, 13, Article 863740.

23.

Reti

I. M.

Samuels

J. F.

Eaton

W. W.

Bienvenu Iii

O. J.

Costa

P. T.

Jr. Nestadt

(2002). Influences of parenting on normal personality traits. Psychiatry Research, 111(1), 55–64.

24.

Schofield

T. J.

Conger

R. D.

Donnellan

M. B.

Jochem

Widaman

Conger

K. J.

(2012). Parent personality and positive parenting as predictors of positive adolescent personality development over time. Merrill-Palmer Quarterly, 58(2), 255–283.

25.

Soto

C. J.

(2019). How replicable are links between personality traits and consequential life outcomes? The life outcomes of personality replication project. Psychological Science, 30(5), 711–727.

26.

Soto

C. J.

(2021). Do links between personality and life outcomes generalize? Testing the robustness of trait–outcome associations across gender, age, ethnicity, and analytic approaches. Social Psychological and Personality Science, 12, 118–130.

27.

Soto

C. J.

John

O. P.

(2017). The next Big Five Inventory (BFI-2): Developing and assessing a hierarchical model with 15 facets to enhance bandwidth, fidelity, and predictive power. Journal of Personality and Social Psychology, 113(1), 117–143.

28.

Stanford

(2003). Multiple intelligence for every classroom. Intervention in School and Clinic, 39(2), 80–85.

29.

Svec

C. L.

(2018). The effects of instruction on the singing ability of children ages 5 to 11: A meta-analysis. Psychology of Music, 46(3), 326–339.

30.

Swaminathan

Schellenberg

(2018). Musical competence is predicted by music training, cognitive abilities, and personality. Scientific Reports, 8, 9223.

31.

Thomas

Silvia

Nusbaum

Beaty

Hodges

(2015). Openness to experience and auditory discrimination ability in music: An investment approach. Psychology of Music, 44(4), 792–801.

32.

von Stumm

Chamorro-Premuzic

Ackerman

P. L

. (2011). Handbook of individual differences. Wiley-Blackwell.

33.

Wang

Z. H.

Luo

H. L.

Zhang

J. X.

(2006). The relationship between parental rearing and adolescent personality traits. Chinese Journal of Clinical Psychology, 14(3), 315–317.

34.

Welch

G. F.

Himonides

Papageorgi

Saunders

Rinta

Stewart

Preti

Lani

Hill

(2009). The National Singing Programme for primary schools in England: An initial baseline study. Music Education Research, 11(1), 1–22.

35.

Welch

G. F.

Himonides

Saunders

Papageorgi

(2010). Researching the impact of the National Singing Programme “Sing Up” in England: Main findings from the first three years (2007–2010). Children’s singing development, self-concept and sense of social inclusion. International Music Education Research Centre, Institute of Education.

36.

Wen

Z. L.

B. J.

(2014). Analyses of mediating effects: The development of methods and models. Advances in Psychological Science, 22(5), 731–745.

37.

B. J.

Fang

X. T.

Yang

Zheng

Liu

L. L.

Guo

S. Y.

(2017). The effects of difficulties in emotional regulation on college students’ mobile phone addiction: The chain mediating effect of facial negative physical self and social avoidance and distress. Psychological Development and Education, 33(2), 249–256.

38.

Yuan

Z. K.

(2010). The initial establishment of high school students multiple intelligence self-rating scale. New Curriculum Guidance, 19, 1–4.

39.

Zhang

Y. M.

Luo

Yin

Chen

Soto

C. J.

John

O. P.

(2022). The Big Five Inventory-2 in China: A comprehensive psychometric evaluation in four diverse samples. Assessment, 29, 1262–1284.

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

0.03 MB