Skip to main content
Frontiers in Psychology logoLink to Frontiers in Psychology
. 2026 Apr 1;17:1767641. doi: 10.3389/fpsyg.2026.1767641

Music teaching self-efficacy among generalist teachers in preschool and primary education: a meta-analysis from Türkiye

Cagri Sen 1,*
PMCID: PMC13079340  PMID: 41993831

Abstract

Background

Music teaching self-efficacy has been widely examined internationally; however, comprehensive meta-analytic syntheses within context-specific teacher education systems remain limited.

Aim

This study synthesized quantitative evidence on the music teaching self-efficacy of preschool and classroom (primary) teachers and teacher candidates in Türkiye.

Methods

A systematic review identified 22 eligible studies (published between 2009 and 2025). Using a random-effects model in Comprehensive Meta-Analysis (CMA), pooled mean estimates, subgroup comparisons, 95% prediction intervals, moderator analyses (instrument type, publication type, publication year), and leave-one-out sensitivity diagnostics were performed to address four research questions concerning gender, overall self-efficacy level, educational level (preschool vs. primary), and professional status (pre-service vs. in-service).

Results

Female participants reported slightly higher self-efficacy than males, but the difference was small and not statistically significant. The pooled mean self-efficacy score across Türkiye was 59.63 on a 0–100 scale (95% prediction interval: 41.76–77.50), indicating moderate but highly variable confidence levels. Subgroup differences by educational level and professional status were not statistically significant. Instrument type significantly contributed to between-study heterogeneity, whereas publication type and year did not. Leave-one-out diagnostics indicated that the pooled estimate was robust. Publication bias analyses did not reveal meaningful asymmetry.

Conclusion

Music teaching self-efficacy in Türkiye appears moderate but structurally uneven, shaped by contextual and measurement-related factors rather than by demographic characteristics alone. The findings underscore the importance of strengthening practice-oriented music education components in both pre-service and in-service teacher preparation and provide a foundation for future longitudinal, intervention-based, and cross-cultural research.

Keywords: generalist teachers, meta-analysis, music teaching self-efficacy, preschool education, primary education

1. Introduction

Music education is one of the fundamental learning areas that contributes to children’s cognitive, social, and emotional development during early childhood and primary school years. Yet, across many educational systems worldwide, music instruction at the preschool and primary school levels is typically undertaken by generalist teachers rather than specialists in music (de Vries, 2013; Jeanneret, 1997; Şen, 2024; Vannatta-Hall, 2010). In this context, generalist teachers refer to educators trained in broad early childhood or primary education programs who are responsible for teaching multiple subjects, including music, without holding a specialized degree in music education.

According to Jorgensen (2008), expertise in music education is distributed inversely; the broadest base of expertise is required at the preschool and primary levels, yet it is precisely at these levels that the least amount of musical expertise is found. Although this situation does not, in itself, constitute a competence deficit, factors such as limited music education content in undergraduate programs, insufficient course hours, the prioritization of other subjects, and restricted institutional resources may narrow the scope of practice in music education (Chung, 2021; de Vries, 2013; Garvis and Pendergast, 2011; Hallam et al., 2009; Kim and Choy, 2008; Temmerman, 1997; Valdebenito and Almonaci-Fierro, 2022; Vannatta-Hall, 2010). In this context, considering the various insecurities that lead many teachers and prospective teachers to avoid music teaching (Barnes and Shinn-Taylor, 1988; Bresler, 1993; de Vries, 2013; Jeanneret, 1997), generalist teachers’ perceptions of self-efficacy in relation to music teaching become a central factor shaping the quality of music education. Indeed, a study carried out in the UK showed that primary school teachers’ confidence levels regarding music teaching are low, largely due to limited professional development opportunities (Holden and Button, 2006).

Bandura (1997) theory of self-efficacy emphasizes that an individual’s belief in their ability to perform specific tasks effectively under varying circumstances is more decisive than the mere possession of those skills when performing a task. In the context of teaching, self-efficacy is regarded as an important psychological construct that shapes teachers’ willingness to cope with classroom challenges, attempt new strategies, and support student learning (Tschannen-Moran and Hoy, 2001).

This belief becomes even more influential in disciplines that combine both cognitive and performative elements, such as music teaching. In the specific context of music education, music teaching self-efficacy is conceptualized in this study as generalist teachers’ task-specific beliefs about their capability to plan, implement, and manage music-related instructional activities effectively under typical classroom conditions, consistent with Bandura (1997) domain-specific conceptualization of self-efficacy. These beliefs encompass confidence in leading singing and rhythm activities, using basic classroom instruments, organizing music-based learning experiences, fostering student participation and creativity, and integrating music into broader curricular objectives (Afacan, 2008; Koca, 2016; Yıldız, 2017).

In the studies included in this meta-analysis, music teaching self-efficacy was operationalized primarily through self-report scales assessing perceived competence in core instructional tasks. These instruments typically encompass combinations of pedagogical confidence in music instruction, perceived competence in fundamental musical skills (e.g., rhythm, singing, and basic instrument use), and the ability to facilitate classroom musical engagement. Although these scales share a common focus on task-specific instructional confidence (see Appendix 2 for the instruments), their item content and dimensional structures differ across studies. Such variations in operationalization may contribute to between-study heterogeneity in pooled estimates.

Studies conducted in different countries have shown that the self-efficacy levels of generalist teachers and prospective teachers in relation to music teaching are influenced by multidimensional factors such as subject knowledge, personal musical experience, access to professional support, and opportunities for classroom practice (Burak, 2019; de Vries, 2011; Garvis and Pendergast, 2011; Hallam et al., 2009; Holden and Button, 2006; Jeanneret, 1997; Kim and Choy, 2008; Taş and Atılgan, 2022). These findings suggest that self-efficacy in music teaching cannot be reduced to individual competence alone but must be understood within broader structural and contextual frameworks.

As in many countries, music education in Türkiye is organized within generalist teacher training programs, and preschool and classroom teachers are likewise responsible for providing music instruction. In the Turkish context, the term “generalist” similarly refers to teachers and prospective teachers trained in preschool or classroom/primary education programs who teach music despite not specializing in music education. Undergraduate preparation in these programs often includes limited theoretical coursework and short-term practical components (Açılmış and Kayıran, 2021; Burak, 2019; Taş and Atılgan, 2022; Yapalı et al., 2025), and the limited time allocated to music lessons further reduces practical opportunities (Koca, 2016; Taş and Atılgan, 2022). Such conditions may affect teachers’ confidence in carrying out basic musical tasks and integrating music meaningfully into classroom practice. Therefore, the Turkish case provides a relevant context for examining how generalist teacher training models relate to music teaching self-efficacy and for situating these findings within international discussions.

While studies on the self-efficacy of generalist teachers and prospective teachers in music teaching have increased in Türkiye in recent years, the use of different sample types, measurement tools, study designs, and reporting formats (Açılmış and Kayıran, 2021; Akpınar and Kaya, 2025; Burak, 2019; Gülle and Akay, 2019; Koca, 2016; Şeker and Çilingir, 2022) makes it difficult to evaluate the findings holistically. A recent meta-analysis in this field (Ertekin Kaya, 2024) identified inconsistencies in gender-related results; however, its focus was limited to studies reporting gender differences. Moreover, that study combined specialist music teachers and generalist teachers within the same analytical framework. Given that music teaching self-efficacy is closely linked to subject-matter knowledge and musical experience, analyzing specialist and non-specialist groups together may limit contextual interpretation. In addition, because the scope of the present meta-analysis extends beyond gender comparisons, several studies included in the current dataset were not part of the earlier synthesis. For this reason, the present study focuses exclusively on generalist teachers and prospective teachers.

Based on the literature, differences in music teaching self-efficacy may reasonably be expected across gender, educational level (preschool vs. primary), and professional status (pre-service vs. in-service). Gender-related findings in prior research have been inconsistent, warranting systematic synthesis. Differences between preschool and primary contexts may reflect variations in curricular emphasis, pedagogical expectations, and classroom structures. Similarly, professional status may shape self-efficacy through differences in mastery experiences, classroom exposure, and institutional responsibilities. Examining these variables comparatively may contribute to a more nuanced understanding of structural and contextual influences on music teaching self-efficacy.

The aim of this research is to systematically synthesize scattered quantitative findings regarding the music teaching self-efficacy levels of preschool and classroom teachers and prospective teachers, who are responsible for music teaching despite being outside the field of music, within the Turkish context. By synthesizing these findings through meta-analysis, the study seeks to identify general trends and patterns. The results are expected to offer evidence-based contributions to teacher education programs and policy initiatives aimed at supporting generalist teachers’ self-efficacy in music teaching. Furthermore, the Turkish case is intended to provide a contextual perspective for international discussions on the integration of music into preschool and primary education programs, given that the fundamental elements of music are largely universal and that preschool and primary school teachers in many countries begin their careers with similarly limited musical preparation.

In line with this objective, the following research questions were addressed:

RQ1: Do self-efficacy levels in music teaching differ by gender?

RQ2: What is the self-efficacy level of preschool and classroom teachers/candidate teachers in music teaching?

RQ3: Is there a significant difference in music teaching self-efficacy between preschool teachers/candidate teachers and classroom teachers/candidate teachers?

RQ4: Is there a significant difference between the self-efficacy levels of teachers and candidate teachers in music teaching?

2. Methodology

2.1. Research model

This study is a meta-analysis that combines and analyses quantitative findings concerning self-efficacy in music teaching among preschool and classroom teachers and candidate teachers in Türkiye. Meta-analysis is a quantitative synthesis method that aims to reveal an overall effect size by statistically combining findings from independent studies (Borenstein et al., 2021).

The 22 studies that met the inclusion criteria for this research covered the period between 2009 and 2025. The analysis and reporting process was carried out in accordance with the PRISMA 2020 guidelines (PRISMA statement, 2020). Microsoft Excel was used to code the studies identified in the search, and the Comprehensive Meta-Analysis (CMA) software was employed to conduct the meta-analytic calculations. The total sample size of the studies included in the meta-analysis consisted of 5,594 participants. Effect sizes and weighted average estimates were calculated based on group means and standard deviations reported across studies. The significance level for all statistical analyses was set at p < 0.05.

In this meta-analysis, the research questions were structured according to the PICOS (Participants, Intervention, Comparisons, Outcomes, Study Design) framework (Higgins and Thomas, 2019). The participants (P) consisted of preschool and primary school teachers and teacher candidates in Türkiye. The intervention (I) was defined as music teaching and the related course or instructional activity. The comparisons (C) included subgroups such as gender, field of study (preschool vs. primary education) and professional status (in-service teachers vs. teacher candidates), depending on the specific research question. The primary outcome (O) was the level of self-efficacy regarding music teaching. The study designs (S) comprised cross-sectional or pretest–posttest studies that reported quantitative findings on music teaching self-efficacy and provided the statistical information required for inclusion in the meta-analysis.

2.2. Inclusion and exclusion criteria

As part of this research, studies reporting quantitative findings on the self-efficacy of preschool and classroom teachers and prospective teachers in music teaching in Türkiye were systematically reviewed. The search period began in 2008, corresponding to the earliest identified scale development study in this field (Afacan, 2008). All accessible quantitative studies (peer-reviewed journal articles, master’s/doctoral theses, and conference proceedings) published between 2008 and 2025 were screened for eligibility.

To ensure conceptual consistency, only studies that explicitly defined and operationalized music teaching self-efficacy as a task-specific belief related to instructional competence in music were considered. Studies focusing solely on general teaching self-efficacy or unrelated constructs were excluded.

Therefore, the search was initiated from 2008 onward; all accessible and suitable quantitative studies (peer-reviewed articles, master’s/doctoral theses, and conference proceedings) published after 2008 were included in the review. Consequently, studies conducted with Turkish samples between 2009 and 2025 that presented quantitative findings on self-efficacy in music teaching were included in the meta-analysis based on their compliance with the specified criteria.

The inclusion criteria were as follows:

  1. The full text had to be accessible in Turkish or English.

  2. The sample had to consist of preschool or classroom teachers/teacher candidates in Türkiye.

  3. A valid measurement tool had to be used to assess self-efficacy in music teaching.

  4. The study had to report sufficient statistical data (e.g., sample size, mean, standard deviation) to enable meta-analytic calculations.

In addition, studies were screened for minimum methodological reporting standards, including clear description of the sample, measurement instrument, and statistical procedures.

The exclusion criteria were as follows:

  1. Studies based exclusively on qualitative data, case reports, or theoretical discussions.

  2. Studies lacking sufficient statistical data for meta-analysis.

  3. Studies employing non-validated or ad hoc measurement tools without reported reliability evidence.

  4. Duplicate publications based on the same dataset; in such cases, only the peer-reviewed journal version was retained to preserve independence of effect sizes.

2.3. Literature search and study selection

International electronic databases and national indexes (Web of Science, Scopus, ERIC, Google Scholar, ProQuest, YÖK National Thesis Center, DergiPark) were systematically searched. The final search across all databases was completed on 18 September 2025. Searches were conducted using the following Boolean string: (“music” OR “müzik”) AND (“self-efficacy” OR “öz-yeterlik” OR “özyeterlik”) AND (“preschool” OR “pre-school” OR “early childhood” OR “okul öncesi”) AND (“primary school” OR “elementary school” OR “sınıf”) AND (“Turkey” OR “Türkiye”). Minor adaptations were made to accommodate database-specific syntax requirements, while preserving the conceptual structure of the search. In international databases, country filters were applied where available to limit results to studies conducted in Türkiye. The reference lists of potentially eligible studies were also reviewed manually. The literature search identified a total of 38 studies, 16 of which did not meet the inclusion criteria and were therefore excluded from the analysis. The remaining 22 studies were included in the meta-analysis. The study selection process is summarized in the flow diagram prepared in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines and presented in Figure 1.

Figure 1.

Flowchart titled “Identifying studies through electronic databases” shows four stages: identification, screening, eligibility, and included. It details study counts and exclusion criteria, with final inclusion of twenty-two studies for meta-analysis.

PRISMA flow diagram.

2.4. Data set and sample characteristics

The data set included in the meta-analysis comprised a total of 22 studies. However, these studies were analyzed in different ways depending on the research questions. For example, 15 studies were included in the analysis for RQ1 (gender-related differences) because seven studies did not provide data on the gender variable. Conversely, for RQ2 (general music teaching self-efficacy), data from all studies (n = 22) were used, and each study was represented by a single value in the meta-analysis.

Most of the studies included in the meta-analysis consisted of a single sample, but some reported multiple subgroups (e.g., preschool vs. primary education; in-service teachers vs. prospective teachers). For the comparative research questions (RQ3 and RQ4), the relevant subgroups were treated as separate samples and entered into the meta-analysis accordingly. Therefore, for RQ3 (comparison by field of teaching), 25 separate samples from 22 studies were used; for RQ4 (comparison of teachers and prospective teachers), 24 separate samples from 22 studies were analyzed. This methodological approach is based on the principle of preserving independent data and preventing within-sample dependency in meta-analyses (Borenstein et al., 2021). The full list of studies included in the meta-analysis is presented in Appendix 1.

2.5. Data coding process

Separate data files were created in Microsoft Excel for each research question, and each study was systematically coded based on the following variables:

  • Descriptive variables: Author(s), year of publication, full title of the study, publication type (article, thesis, presentation), publication language (Turkish/English), education level (preschool/classroom education), sample type (teacher/preservice teacher), gender distribution (if applicable), name of the scale used, number of items in the scale, and scoring range.

  • Statistical variables: Mean (M), standard deviation (SD), sample size (N), and total score; separate M, SD, and n values for each group in two-group comparisons.

Throughout the data coding process, all data were systematically reviewed multiple times to prevent erroneous or duplicate entries.

Almost all of the studies included in the meta-analysis were based on a cross-sectional design, and only two studies reported a pretest-posttest design (Kaya, 2022; Umuzdaş and Işıldak, 2018). In these studies, only pretest data, which were natural self-efficacy scores, were included in the meta-analysis because post-intervention posttest scores were considered to reduce comparability with other cross-sectional studies and artificially alter the combined effect size. This methodological decision was made to prevent differences in study designs from skewing the meta-analytic results.

2.6. Data transformation and matching procedures

2.6.1. Normalization of scales

Because different measurement instruments were used to assess music teaching self-efficacy in the studies included in the meta-analysis (the scales used are listed in Appendix 2), the score ranges of these instruments varied. Therefore, in the analyses for RQ2, RQ3, and RQ4, all scale scores were normalized to a 0–100 range to eliminate inter-scale score differences. Normalization was performed by rescaling the mean (M) and standard deviation (SD) values reported in each study according to that instrument’s minimum and maximum score ranges (Aksu et al., 2019). The formulas used for this transformation are shown in Figure 2.

Figure 2.

Two mathematical formulas for normalization shown. First, normalized mean equals reported mean minus minimum, divided by maximum minus minimum, times one hundred. Second, normalized standard deviation equals reported standard deviation divided by maximum minus minimum, multiplied by one hundred.

Normalization formulas.

This approach ensured that all mean and standard deviation values were expressed on the same scale (0–100), thereby reducing variations arising from differing scale types and enabling more accurate comparisons across studies. It recommends placing data measured with different instruments but representing the same underlying construct on a common metric. Indeed, Borenstein et al. (2021) emphasize that removing discrepancies in measurement units prior to conducting a meta-analysis both prevents information loss and allows all studies to be assessed on a single, shared scale.

In the analyses related to RQ1, no normalization was applied because the effect size was calculated using the standardized mean difference (Hedges’s g; Borenstein et al., 2021).

2.6.2. Combining subgroups

In several studies included in the analysis, aggregated statistics (e.g., overall mean or standard deviation) required to address the research questions were not directly reported. In such instances, combined mean and standard deviation values were calculated using data from subgroups. The mean (M) and standard deviation (SD) reported for each subgroup were weighted according to their sample sizes to obtain the combined values (Higgins and Thomas, 2019). This procedure ensures that the total variance is represented in a statistically comprehensive manner while treating the subgroups as independent. The formulas used in the calculations are provided in Figure 3.

Figure 3.

Two formulas are shown for combining means and standard deviations. The combined mean formula: M_sub_combined = (n1 M1 + n2 M2) divided by (n1 + n2). The combined standard deviation formula: SD_sub_combined equals the square root of [(n1 minus 1) SD1 squared plus (n2 minus 1) SD2 squared plus n1 n2 divided by (n1 plus n2) times (M1 minus M2) squared] divided by (n1 plus n2 minus 1).

Combined formulas. M1 and M2 represent the mean score for each subgroup, and n1 and n2 represent the sample size for each subgroup.

All calculations were carried out in Microsoft Excel, and the resulting values are labeled as “combined” in the tables.

2.7. Analysis strategy

Meta-analytic calculations were conducted using the Comprehensive Meta-Analysis (CMA) software. Analysis types were structured differently depending on the nature of the research questions.

In comparative analyses, differences between groups were assessed based on group means. Accordingly, effect sizes for RQ1 were calculated using Hedges’s g, while in the analyses for RQ3 and RQ4, subgroup meta-analysis was performed using mean scores normalized to the 0–100 scale. Hedges’s g was selected to minimize small-sample bias associated with sample size (Borenstein et al., 2021). Effect sizes were interpreted using Cohen (1988) classification: small (≈0.20), medium (≈0.50), and large (≈0.80).

For RQ2, which involved one-group means, normalized scores obtained from different studies were combined using the weighted mean method, taking account of sample sizes and variance values.

All analyses were performed separately for each research question. The group variables examined in RQ3 and RQ4 were treated as moderator variables, and subgroup analyses were conducted accordingly. In addition, to explore potential sources of heterogeneity in RQ2, moderator analyses were conducted for publication type and instrument type using subgroup comparisons, and publication year was examined using meta-regression (Borenstein et al., 2021).

In addition, 95% prediction intervals (PI) were calculated to estimate the dispersion of true mean scores across populations (Borenstein et al., 2017; IntHout et al., 2016). Finally, sensitivity analyses were performed using a leave-one-out procedure to evaluate the influence of individual studies on the pooled mean estimates (Higgins and Thomas, 2019).

2.8. Heterogeneity analysis

Heterogeneity across studies was assessed using Cochran’s Q test, I2 (I-squared), and τ2 (Tau-squared). The Q statistic tests the null hypothesis that all studies share a common effect size. A significant Q value (p < 0.05) indicates that true differences exist between studies that cannot be explained solely by sampling error (Borenstein et al., 2021). The I2 statistic reflects the proportion of total variance attributable to genuine heterogeneity. According to the classification proposed by Higgins et al. (2003), threshold values of 25%, 50, and 75% indicate low, moderate, and high heterogeneity, respectively. Although τ2 does not directly indicate the level of heterogeneity, it represents the core variance component of the random-effects model by capturing the variance of the true effect size differences between studies (Borenstein et al., 2021). Summary values for heterogeneity are presented in Table 1.

Table 1.

Heterogeneity statistics.

Research question Number of studies/Samples Q p I2 (%) τ2 Model
RQ1 15 studies 41.10 <0.001 65.9 0.04 Random
RQ2 22 studies 2197.07 <0.001 99.0 70.19 Random
RQ3 22 studies, 25 independent samples 2222.95 <0.001 98.9 70.22 Random
RQ4 22 studies, 24 independent samples 2270.80 <0.001 99.0 77.54 Random

The analysis results indicate high levels of heterogeneity, particularly for RQ2, RQ3, and RQ4, and a moderate–high level for RQ1. According to Dinçer (2021), it is common for studies in the field of educational sciences to display a heterogeneous structure. For this reason, the analyses were conducted using a random-effects model (Higgins and Thomas, 2019). Subgroup analyses were carried out within the framework of the Mixed-Effects Model approach. In this model, the studies within each subgroup were combined using a random-effects model, and differences between groups were tested using the Q_between statistic (Borenstein et al., 2021).

2.9. Publication bias

Publication bias, a key factor that can undermine the reliability of meta-analytic findings, may arise when only studies reporting statistically significant or positive results are published, or when certain studies are included following a limited literature search. Such circumstances can lead to an overestimation of the overall effect size (Dinçer, 2021). To assess this possibility, the datasets included in the meta-analysis were examined using both visual and statistical methods.

First, the relationship between effect sizes and standard errors was visualized using a funnel plot, and the symmetry of the plot was evaluated. Marked asymmetry is considered a visual indicator of publication bias (Borenstein et al., 2021).

To statistically assess the likelihood of publication bias, the Egger’s Regression Test and the Begg–Mazumdar Rank Correlation Test were applied in addition to visual inspection (Begg and Mazumdar, 1994; Egger et al., 1997). Significant p values (p < 0.05) in either of the tests are interpreted as evidence of publication bias (Borenstein et al., 2021).

To evaluate the potential impact of missing studies on effect size, the Duval & Tweedie Trim-and-Fill method was employed, and adjusted effect sizes were calculated (Duval and Tweedie, 2000).

3. Results

This section presents the findings relating to the four research questions addressed in the meta-analysis.

3.1. Music teaching self-efficacy by gender (RQ1)

The meta-analysis concerning the gender variable was conducted using data from 15 independent studies. Moderate to high heterogeneity was observed among these studies (see Table 1). Accordingly, the analyses were performed using a random-effects model.

The individual effect sizes, confidence intervals, and study weights of the 15 studies included in the analysis are provided in Table 2; the visual distribution of these findings is illustrated in the forest plot in Figure 4. In the visual, the squares represent the weights of the studies, while the diamond indicates the overall average effect size. As shown, although levels of music teaching self-efficacy by gender vary slightly across samples, the overall trend favors women. Most effect sizes are positive, indicating a difference in favor of female participants.

Table 2.

Individual study effect sizes for gender differences in music teaching self-efficacy.

Study Hedges’s g 95% CI Weight (%)
Açılmış and Kayıran (2021) 0.17 [−0.01, 0.36] 9.2
Akpınar and Kaya (2025) 0.02 [−0.40, 0.44] 5.4
Altındaş (2023) 0.51 [−0.06, 1.07] 3.8
Çelik (2018) −0.01 [−0.26, 0.24] 8.2
Çelik and Yetim (2017) 0.27 [−0.04, 0.57] 7.2
Çevik (2011) 0.39 [0.05, 0.73] 6.6
Demiralay (2023) 0.16 [−0.18, 0.50] 6.6
Kavaklı (2022) 0.25 [0.04, 0.45] 8.9
Koca (2013a) −0.72 [−1.29, −0.15] 3.8
Koca (2013b) 0.17 [−0.36, 0.69] 4.3
Koca (2016) 0.76 [0.26, 1.25] 4.5
Saylam Kırcıoğlu (2009) −0.05 [−0.36, 0.25] 7.2
Şeker and Çilingir (2022) −0.31 [−0.51, −0.10] 8.9
Topoğlu (2014) 0.13 [−0.12, 0.38] 8.2
Yücesan (2023) 0.26 [−0.04, 0.57] 7.1

Figure 4.

Forest plot displaying Hedges’ g effect sizes with ninety-five percent confidence intervals for fifteen studies and an overall random effects summary. Each study, listed by author and year, is represented by a black square proportional to sample size, with horizontal lines depicting confidence intervals centered around varying values on the x-axis. The summary effect is shown as a diamond at the plot’s bottom. The vertical dashed line represents a zero effect.

Forest plot for gender differences in music teaching self-efficacy.

To summarize the general trend regarding the gender variable, the findings from both the fixed-effects and random-effects models are presented in Table 3. However, due to the above-moderate level of heterogeneity, interpretations are based on the random-effects model.

Table 3.

Overall meta-analysis results for gender differences in music teaching self-efficacy.

Model k Hedges’s g SE 95% CI Z p
Fixed-effects model 15 0.11 0.04 [0.03, 0.18] 2.77 0.006
Random-effects model 15 0.13 0.07 [−0.01, 0.26] 1.81 0.07

As shown in Table 3, the random-effects model indicates that female teachers and prospective teachers exhibit higher levels of self-efficacy in music teaching than male counterparts; however, this difference is not statistically significant (g = 0.13, p = 0.07). Although the positive effect size suggests a general trend favoring women, the magnitude of this difference falls below the conventional small-effect (≈0.20) benchmark according to Cohen (1988) classification.

3.1.1. Publication bias analysis (RQ1)

Both visual and statistical procedures were employed to assess potential publication bias in the studies included in the meta-analysis for the gender variable. The funnel plot for RQ1 demonstrated a broadly symmetrical distribution (see Figure 5), indicating no visual evidence of systematic publication bias among the studies.

Figure 5.

Funnel plot displaying standard error on the vertical axis and Hedges's g on the horizontal axis, with individual study estimates represented by circles and a triangular region illustrating potential publication bias.

Funnel plot for gender differences in music teaching self-efficacy.

The statistical analysis results also support this conclusion. The Egger’s Regression Test (Intercept = 0.86; 95%CI [−2.09, 3.80]; p = 0.54) and the Begg–Mazumdar Rank Correlation Test (τ = 0.09, p = 0.66) indicated no significant evidence of asymmetry in the distribution.

The Duval & Tweedie Trim-and-Fill analysis, applied to account for the potential impact of publication bias, estimated that three studies may be missing to the left of the mean and produced an adjusted effect size of g = 0.05 (95%CI [−0.09, 0.19]). This result suggests that the possible missing studies did not meaningfully influence the overall effect.

Taken together, these findings indicate that no systematic publication bias was detected in the analyses conducted for RQ1, and the overall effect can be considered robust.

3.2. Music teaching self-efficacy levels in Türkiye (RQ2)

To determine the music teaching self-efficacy levels of preschool and classroom teachers and prospective teachers in Türkiye, a one-group mean meta-analysis was conducted using data from 22 independent studies. Prior to the analysis, scores obtained from different scales were normalized to a 0–100 range to create a common metric.

High heterogeneity was observed among the studies (see Table 1); therefore, analyses were carried out using a random-effects model. The mean music teaching self-efficacy scores and confidence intervals for the 22 studies included in the meta-analysis are presented in Table 4.

Table 4.

Individual study means for overall music teaching self-efficacy.

Study Mean (0–100) 95% CI
Açılmış and Kayıran (2021) 58.61 [57.24, 59.98]
Akpınar and Kaya (2025) 79.50 [78.15, 80.85]
Altındaş (2023) 56.00 [53.64, 58.37]
Burak (2019) 48.04 [46.56, 49.52]
Çelik (2018) 59.68 [58.35, 61.01]
Çelik and Yetim (2017) 59.90 [58.32, 61.48]
Çevik (2011) 61.25 [58.99, 63.51]
Demiralay (2023) 55.49 [53.25, 57.73]
Gülle and Akay (2019) 59.18 [58.16, 60.20]
Kavaklı (2022) 62.45 [61.26, 63.64]
Kaya (2022) 45.00 [41.45, 48.56]
Koca (2013a) 65.00 [63.05, 66.96]
Koca (2013b) 58.25 [56.74, 59.76]
Koca (2016) 79.50 [77.66, 81.34]
Saylam Kırcıoğlu (2009) 43.05 [40.95, 45.15]
Şeker and Çilingir (2022) 54.68 [53.45, 55.91]
Sonakın (2022) 62.00 [59.06, 64.94]
Taş and Atılgan (2022) 53.77 [52.33, 55.21]
Topoğlu (2014) 66.34 [64.24, 68.44]
Umuzdaş and Işıldak (2018) 63.11 [60.90, 65.33]
Yegül (2014) 53.95 [52.47, 55.43]
Yücesan (2023) 66.61 [65.01, 68.21]

To summarize the level of music teaching self-efficacy across Türkiye, the results from both the fixed-effects and random-effects models are presented in Table 5. However, due to the high level of heterogeneity, the interpretations are based on the random-effects model.

Table 5.

Overall meta-analysis results for music teaching self-efficacy.

Model k Mean (0–100) SE 95%CI
Fixed-effects model 22 60.08 0.17 [59.74, 60.42]
Random-effects model 22 59.63 1.80 [56.10, 63.15]

As shown in Table 5, the random-effects model indicates that the pooled mean for music teaching self-efficacy of preschool and classroom teachers, as well as prospective teachers across Türkiye, is 59.63 (95%CI [56.10, 63.15]). This value, being above 50 –the midpoint of the 0–100 scale– suggests an above-midpoint level of general self-efficacy. Given the substantial heterogeneity observed across studies (Q = 2197.07, p < 0.001; I2 = 99.0%; τ2 = 70.19), a 95% prediction interval (PI) was calculated to estimate the distribution of true mean scores across different populations within the same context. The prediction interval ranged from 41.76 to 77.50. This wide interval indicates that, despite the above-midpoint pooled estimate, true mean self-efficacy levels vary considerably across samples and educational contexts within Türkiye.

To explore potential sources of the substantial heterogeneity observed in the overall analysis, moderator analyses were conducted. Publication type (article vs. thesis) yielded a marginal between-group difference (Q_between(1) = 3.81, p = 0.051). Although articles (M = 61.84) tended to report slightly higher mean self-efficacy levels than theses (M = 54.87), this difference did not reach conventional levels of statistical significance.

Instrument type (scale name) revealed a statistically significant between-group difference (Q_between(4) = 668.94, p < 0.001), indicating that pooled mean estimates varied significantly across measurement instruments. Mean self-efficacy scores differed substantially across scales, suggesting that measurement characteristics contributed to between-study variability.

Publication year was examined using meta-regression, and the regression coefficient was not statistically significant (β = 0.27, SE = 0.40, p = 0.50), indicating that music teaching self-efficacy levels did not demonstrate a systematic linear change over time within the included studies.

A leave-one-out sensitivity analysis further indicated that the pooled mean estimate was robust. When individual studies were removed sequentially, the pooled mean ranged between 58.69 and 60.42, suggesting that no single study exerted a disproportionate influence on the overall estimate.

3.2.1. Publication bias analysis (RQ2)

Both visual and statistical methods were used to assess potential publication bias in the studies included in the meta-analysis on music teaching self-efficacy in Türkiye. The funnel plot for RQ2 displayed a broadly symmetrical distribution (see Figure 6), suggesting no visual evidence of systematic publication bias across the studies.

Figure 6.

Funnel plot displaying standard error versus mean, with data points clustered symmetrically around a central mean value and precision increasing as standard error decreases, indicating potential publication bias assessment.

Funnel plot for the overall music teaching self-efficacy.

Statistical analysis results also support this finding. The Egger’s Regression Test (Intercept = −2.04; 95%CI [−18.99, 14.91]; p = 0.80) and the Begg–Mazumdar Rank Correlation Test (τ = 0.004, p = 0.98) provide no significant evidence of an asymmetric distribution.

The Duval & Tweedie Trim-and-Fill analysis, applied to account for potential publication bias, did not identify any missing studies to the left or right of the mean (trimmed = 0). Consequently, the adjusted values and observed values were identical. The point estimate and confidence intervals remained unchanged in both the fixed-effects and random-effects models (59.63, 95%CI [56.10, 63.15]).

Taken together, these findings indicate that no systematic publication bias was detected in the meta-analysis conducted under RQ2, and the resulting composite mean can be considered reliable.

3.3. Comparison of music teaching self-efficacy between preschool and primary school teachers/candidates (RQ3)

A subgroup analysis was conducted to determine whether there was a significant difference between the music teaching self-efficacy levels of preschool teachers/candidates and classroom teachers/candidates. The analysis was carried out using normalized mean scores (on a 0–100 scale) based on independent samples from each group. Owing to the high level of heterogeneity across studies (see Table 1), a Mixed-Effects Model approach was preferred (Borenstein et al., 2021).

As shown in Table 6, the mean music teaching self-efficacy levels of both groups lie within the moderate to upper range. The composite mean for the preschool teaching group was 62.01 (95%CI [55.44, 68.59]), while that of the classroom teaching group was 56.93 (95%CI [54.32, 59.55]). These results suggest that the preschool teaching group tended to have higher self-efficacy levels than the classroom teaching group.

Table 6.

Subgroup analysis results for preschool and primary school teachers/candidates.

Group k Mean (0–100) SE 95%CI I2 (%) Q_between p
Preschool teachers/candidates 11 62.01 3.35 [55.44, 68.59] 99.2 1.98 0.16
Primary school teachers/candidates 14 56.93 1.33 [54.32, 59.55] 97.3

The subgroup comparison presented in Table 6 yielded Q_between(1) = 1.98, p = 0.16, indicating no statistically significant difference in music teaching self-efficacy levels between the preschool and primary school groups. However, substantial heterogeneity was observed within both subgroups (I2 = 99.2 and 97.3, respectively). To account for this variability, 95% prediction intervals (PI) were calculated. For preschool teachers, the prediction interval ranged from 35.86 to 88.17, whereas for primary school teachers it ranged from 45.85 to 68.01. These wide and overlapping intervals indicate that true mean self-efficacy levels vary considerably across samples within each subgroup, and that the observed difference between preschool and classroom teacher groups should be interpreted with caution.

3.3.1. Publication bias analysis (RQ3)

Both visual and statistical analyses were conducted to assess potential publication bias among the studies included in the meta-analysis on preschool and classroom teachers/prospective teachers. The funnel plot for RQ3 showed no visual evidence of systematic publication bias across the studies (see Figure 7).

Figure 7.

Funnel plot illustrating standard error on the y-axis and mean on the x-axis, with data points scattered mostly around the mean of 60, displaying a symmetrical distribution within funnel-shaped confidence boundaries.

Funnel plot for preschool and primary school teacher/candidate comparison.

Statistical test results also support this conclusion. The Egger’s Regression Test (Intercept = −5.58; 95%CI [−20.47, 9.32]; p = 0.45) and the Begg & Mazumdar Rank Correlation Test (τ = −0.04; p = 0.78) do not provide significant evidence for the presence of an asymmetric distribution.

The Duval & Tweedie Trim-and-Fill method, applied to assess the potential influence of missing studies, estimated that five studies might lie to the right of the mean and produced an adjusted point estimate of 62.02 (95%CI [58.64, 65.40]). This adjustment did not lead to a meaningful change in the overall result.

When all these findings are considered together, it can be said that no systematic publication bias was detected in the subgroup meta-analysis conducted under RQ3, and the resulting combined mean is reliable.

3.4. Comparison of music teaching self-efficacy between teachers and teacher candidates (RQ4)

A subgroup analysis was conducted to determine whether there was a significant difference between the music teaching self-efficacy levels of teachers and teacher candidates. The analysis was conducted using normalized mean scores (0–100 scale) based on independent samples from each group; given the high heterogeneity across studies (see Table 1), the Mixed-Effects Model was preferred (Borenstein et al., 2021).

According to the results presented in Table 7, both groups exhibited moderate to above-midpoint levels of music teaching self-efficacy. The combined mean for prospective teachers was 60.49 (95%CI [56.91, 64.08]), while the combined mean for teachers was 58.26 (95%CI [50.78, 65.74]). These findings suggest that the music teaching self-efficacy levels of prospective teachers tend to be slightly higher than those of teachers.

Table 7.

Subgroup analysis results for teachers and teacher candidates.

Group k Mean (0–100) SE 95%CI I2 (%) Q_between p
Teacher candidates 15 60.49 1.83 [56.91, 64.08] 98.6 0.28 0.60
Teachers 9 58.26 3.82 [50.78, 65.74] 99.4

The statistical analysis results for the intergroup comparison presented in Table 7 yielded Q_between(1) = 0.28, p = 0.60, indicating no statistically significant difference in music teaching self-efficacy between teachers and preservice teachers. However, substantial heterogeneity was observed within both subgroups (I2 = 98.6 and 99.4, respectively). To account for this variability, 95% prediction intervals (PI) were calculated. For preservice teachers, the prediction interval ranged from 44.83 to 76.16, whereas for in-service teachers it ranged from 29.84 to 86.69. These wide and overlapping intervals indicate that true mean self-efficacy levels vary considerably across samples within each subgroup, and that the observed difference between teachers and preservice teachers should be interpreted with caution.

This may be due to factors such as sample diversity, the measurement tools used, implementation contexts, and music teaching experience. Consequently, the findings should be interpreted with caution in light of these contextual variations.

3.4.1. Publication bias analysis (RQ4)

Both visual and statistical analyses were conducted to assess potential publication bias among studies included in the meta-analysis on teachers and preservice teachers. The funnel plot for RQ4 displays a generally symmetric distribution, suggesting no visual evidence of systematic publication bias (see Figure 8).

Figure 8.

Funnel plot displaying standard error on the vertical axis and mean on the horizontal axis, with data points represented by circles clustered around the central mean, showing symmetrical distribution within funnel-shaped confidence limits.

Funnel plot for teacher and teacher candidate subgroup comparison.

Statistical test results are consistent with these visual findings. The Egger’s Regression Test (Intercept = −1.79; 95%CI [−16.24, 12.65]; p = 0.80) and the Begg & Mazumdar Rank Correlation Test (τ = 0.00; p = 1.00) do not provide significant evidence for the presence of an asymmetric distribution.

The Duval & Tweedie Trim-and-Fill method, applied to identify potentially missing studies, did not detect any studies missing either to the left or right of the mean (trimmed = 0). Therefore, the adjusted and observed values were identical. The point estimate and confidence intervals in both the fixed-effects and random-effects models remained unchanged (59.65, 95%CI [56.24, 63.06]).

Taken together, these findings indicate that no systematic publication bias was identified in the subgroup meta-analysis conducted under RQ4, and the resulting combined mean can be considered reliable.

4. Discussion, conclusion and recommendations

The findings obtained in this meta-analysis provide a comprehensive and systematic assessment of the music teaching self-efficacy levels of preschool and classroom teachers and prospective teachers in Türkiye. The results highlight both the general levels of self-efficacy and the differences associated with demographic and professional variables.

Regarding RQ1, the findings relating to gender indicated that female participants had slightly higher music teaching self-efficacy levels than male participants, although this difference was not statistically significant.

Zelenak (2024) notes in his meta-analysis that the literature presents inconsistent findings on gender-related self-efficacy levels. Some studies report higher self-efficacy for music teaching in favor of men (Koca, 2013a; Şeker and Çilingir, 2022), while others report higher levels in favor of women (Çevik, 2011; Koca, 2016). Conversely, several studies with different sample groups have found that gender does not have a significant effect on music teaching self-efficacy (Açılmış and Kayıran, 2021; Akpınar and Kaya, 2025; Burak, 2019; Çelik, 2018; Çelik and Yetim, 2017; Gülle and Akay, 2019; Kavaklı, 2022; Koca, 2013b; Topoğlu, 2014).

The present meta-analytic findings align more closely with the latter pattern, suggesting that gender, in itself, is not a robust determinant of music teaching self-efficacy. From a social-cognitive perspective (Bandura, 1997), self-efficacy beliefs are primarily shaped by mastery experiences, vicarious experiences, social persuasion, and physiological states rather than by demographic characteristics alone. Studies in music education similarly indicate that contextual variables such as prior musical background, practical teaching experience, and access to institutional support significantly influence teachers’ self-efficacy beliefs (de Vries, 2013). Therefore, any observed gender differences may reflect indirect influences rather than gender as a causal factor per se.

In the Turkish context, the absence of a significant gender effect may be related to relatively standardized teacher education pathways and curriculum structures across genders, potentially reducing systematic variation in music teaching preparation. Moreover, variations in sample composition, measurement instruments, and reporting practices across primary studies may partly account for the contradictory gender-related findings reported in the literature. Taken together, the small and statistically non-significant pooled effect indicates that gender-related differences, where present, are likely to be context-dependent and mediated by structural and experiential factors rather than constituting a stable pattern across studies.

For RQ2, the combined data from 22 studies conducted across Türkiye indicate that the average music teaching self-efficacy level (M = 59.63) of 5,594 teachers and prospective teachers was above the midpoint of the 0–100 scale. This finding suggests that participants held a generally positive, yet not particularly strong, perception of their competence in music teaching. From a social-cognitive perspective (Bandura, 1997), an individual’s belief in their ability to perform a specific task depends not only on their knowledge and skills but also on their confidence in applying these competencies effectively. In this sense, the pooled estimate may reflect a moderate level of confidence that enables basic instructional functioning but may not consistently translate into high levels of pedagogical innovation or sustained musical engagement in classrooms.

However, the substantial heterogeneity observed across studies and the wide prediction interval (41.76–77.50) indicate that music teaching self-efficacy in Türkiye is not uniformly distributed but varies considerably across institutional and educational contexts. While the pooled mean reflects an above-midpoint central tendency, the breadth of the interval demonstrates that true mean levels differ markedly across samples. Rather than reflecting a uniform national pattern, the results indicate marked heterogeneity in music teaching self-efficacy across educational settings.

The moderator analyses provide further insight into this variability. Instrument type emerged as a significant moderator, indicating that differences in measurement tools contributed substantially to between-study variability. This finding is theoretically and methodologically important, as it suggests that inconsistencies reported in the literature may partly stem from differences in operational definitions and scale structures rather than from genuine population-level disparities. In contrast, publication type showed only a marginal effect, and publication year did not predict systematic changes in self-efficacy over time. The absence of a significant temporal trend implies that, despite ongoing reforms in teacher education, perceived competence in music teaching has remained relatively stable over the years covered by the included studies.

Importantly, the leave-one-out sensitivity analysis demonstrated that the pooled estimate was robust, as the removal of any single study did not materially alter the overall mean. This stability strengthens confidence in the central tendency identified in the meta-analysis, despite the high heterogeneity.

These results are broadly consistent with individual studies conducted in Türkiye, which report moderate levels of music teaching self-efficacy (Açılmış and Kayıran, 2021; Çelik, 2018; Çevik, 2011; Taş and Atılgan, 2022), while others document relatively higher levels (Akpınar and Kaya, 2025; Koca, 2016). This pattern aligns with the substantial variability observed in the meta-analysis and suggests that self-efficacy perceptions may differ depending on sample characteristics, the measurement tools employed, and the educational contexts in which the studies were conducted. Indeed, the mean self-efficacy scores reported in the individual studies ranged from 43.05 to 79.50, indicating that the composite mean represents a central tendency within a broad distribution rather than an extreme value.

International findings echo similar patterns. Lowe et al. (2017) reported that preservice teachers felt only moderately competent in music teaching, and Valdebenito and Almonaci-Fierro (2022) identified persistent insecurities in teachers’ perceived competence across contexts. Such parallels suggest that moderate but uneven self-efficacy may be a common feature of generalist-led music education systems rather than a uniquely Turkish phenomenon.

Taken together, the findings indicate that strengthening music teaching self-efficacy requires more than incremental curricular adjustments. Previous research suggests that inadequate music education, limited practice-based content, restricted prior musical experience, the prioritization of other subjects, and insufficient access to professional development opportunities may weaken teachers’ self-efficacy beliefs (de Vries, 2011, 2013; Hallam et al., 2009; Lowe et al., 2017). Future research should therefore examine the mechanisms through which mastery experiences, structured practicum opportunities, and sustained engagement with music instruction influence self-efficacy development over time. Longitudinal and intervention-based designs could clarify whether targeted, practice-oriented teacher education components produce durable gains in music teaching self-efficacy. Additionally, given the significant instrument effect observed, future studies should assess measurement equivalence across scales and populations to ensure that apparent differences reflect substantive variation rather than methodological artifacts.

According to the subgroup analysis conducted under RQ3, preschool teachers and preservice teachers reported higher mean self-efficacy levels in music teaching (M = 62.01) than classroom teachers and prospective teachers (M = 56.93). However, this difference was not statistically significant (Q_between = 1.98, p = 0.16). Although both groups demonstrated above-midpoint levels of self-efficacy, the prediction intervals (PI) calculated for each subgroup (35.85–88.17 for preschool; 45.85–68.01 for primary school) revealed substantial within-group variability and considerable overlap. Although the preschool group showed a higher pooled mean, the substantial heterogeneity and overlapping prediction intervals indicate that this descriptive difference cannot be interpreted as a consistent structural advantage.

Nevertheless, previous literature has proposed contextual explanations that may help interpret this trend. In preschool education, music teaching is often embedded in play-based activities, and the emphasis on singing, movement, and rhythm may provide more frequent mastery experiences, which are central sources of self-efficacy development (Bandura, 1997). Kavaklı (2022) likewise found that preschool teachers reported significantly higher self-efficacy levels than classroom teachers, and Burak (2019) associated this difference with the scope and quality of applied music courses in preschool teacher education programs. Additionally, the perception that musical tasks in early childhood settings are developmentally appropriate and attainable may further strengthen confidence.

Conversely, music instruction in classroom teaching programs may be more theoretically oriented or constrained by curriculum pressures associated with academic subjects, potentially limiting opportunities for sustained practice-based reinforcement. At the same time, the high heterogeneity observed within both subgroups suggests that contextual, institutional, and programmatic factors vary substantially across settings. Therefore, the apparent advantage of preschool teachers should be interpreted as context-dependent rather than categorical.

Future research could examine whether the intensity of practicum experiences, the proportion of applied music coursework, or the integration of music into daily classroom routines moderates differences between preschool and primary education contexts. Such analyses would help determine whether structural program features, rather than educational level per se, account for variations in music teaching self-efficacy. In practical terms, the findings underscore the importance of designing music education components within teacher preparation programs that are more targeted and sensitive to the distinct pedagogical realities of both preschool and classroom teaching contexts.

According to the subgroup analysis conducted under RQ4, preservice teachers reported slightly higher mean self-efficacy levels in music teaching (M = 60.49) than in-service teachers (M = 58.26). However, this difference was not statistically significant (Q_between = 0.28, p = 0.60). Both groups demonstrated above-midpoint levels of self-efficacy, yet the prediction intervals (44.83–76.16 for preservice teachers; 29.84–86.69 for in-service teachers) revealed substantial within-group variability and considerable overlap. Although preservice teachers showed a marginally higher pooled mean, the heterogeneity and overlapping intervals indicate that this descriptive difference cannot be attributed solely to professional status.

Valdebenito and Almonaci-Fierro (2022) emphasize that self-efficacy in music teaching develops through teacher education, professional experience, and opportunities for practice. From this perspective, professional experience would ordinarily be expected to strengthen self-efficacy. However, the present findings suggest that this expectation may not uniformly apply in music education contexts.

From a social-cognitive standpoint (Bandura, 1997), self-efficacy develops primarily through mastery experiences. Garvis (2013) observed that while self-efficacy increases with experience in core academic subjects, this trajectory may not hold for music. Similarly, Barnes and Shinn-Taylor (1988) reported that generalist teachers frequently perceive music as their weakest instructional area. This pattern may reflect the marginalized position of music within many school curricula, where limited instructional time and reduced institutional emphasis constrain opportunities for sustained practice and professional growth (Valdebenito and Almonaci-Fierro, 2022).

At the same time, previous research has highlighted the role of pedagogically rich and practice-based music education experiences in shaping preservice teachers’ self-efficacy beliefs (Jeanneret, 1997; Kim and Choy, 2008; Umuzdaş and Işıldak, 2018; Vannatta-Hall, 2010). Conversely, limited practical opportunities and insufficient pedagogical content during teacher education may constrain self-efficacy development (Lowe et al., 2017; Temmerman, 1997). Such programmatic differences may help explain why preservice teachers sometimes report confidence levels comparable to—or slightly higher than—those of practicing teachers, particularly in contexts where ongoing professional reinforcement in music is limited.

Taken together, these findings support theoretical accounts suggesting that self-efficacy is shaped not merely by years of professional experience but by the quality of music education received, the practice-oriented structure of curricula, and the frequency of engagement in musical activities (Holden and Button, 2006; Valdebenito and Almonaci-Fierro, 2022). In this sense, years of teaching experience alone may not translate into enhanced music teaching self-efficacy unless accompanied by continued engagement, targeted professional development, and institutional support. It is therefore plausible that professional experience exerts an indirect rather than a direct influence on music teaching self-efficacy, mediated by the frequency and quality of musical teaching opportunities.

Future research could examine whether ongoing participation in music-related professional development, frequency of classroom music implementation, or mentoring support moderates differences between preservice and in-service teachers. Longitudinal designs would be particularly valuable in determining whether self-efficacy trajectories in music diverge from those observed in other subject areas over the course of a teaching career.

Practically, the findings underscore the importance of sustaining structured and practice-oriented music education components not only during preservice preparation but also throughout teachers’ professional careers. Access to regular and high-quality in-service professional development may be critical for maintaining continued engagement with music teaching (Chung, 2021). As Bresler (1993) emphasizes, effective implementation of music education depends not only on individual teacher competence but also on institutional support within schools. Furthermore, meta-analytic evidence from music performance contexts suggests that structured interventions targeting self-efficacy can meaningfully strengthen efficacy beliefs (Zelenak, 2024), indicating that systematic and sustained support mechanisms may be essential for both preservice and in-service teachers.

This meta-analysis synthesizes fragmented findings on music teaching self-efficacy among generalist preschool and classroom teachers and preservice teachers in Türkiye, providing an integrative overview of overall levels and subgroup patterns. While demographic and professional variables appear to relate to self-efficacy, their effects are generally small and context-dependent. The findings therefore highlight the importance of structural, curricular, and institutional factors in shaping music teaching self-efficacy.

Future research could extend this work by examining potential mediating and moderating mechanisms, such as frequency of music implementation, mentoring support, workload, or school climate. Longitudinal designs would help clarify how music teaching self-efficacy evolves over the course of a teaching career. In addition, cross-national meta-analyses conducted across diverse educational systems could provide further insight into the contextual and cultural conditions under which music teaching self-efficacy develops.

5. Limitations

This meta-analysis has several limitations. First, the included studies relied exclusively on self-report instruments, which capture perceived competence rather than observed instructional performance. Consequently, reported self-efficacy levels may not fully correspond to objective teaching quality.

Second, substantial heterogeneity was observed across analyses. Although moderator and sensitivity analyses were conducted, unmeasured contextual variables (e.g., regional differences, institutional conditions, or prior musical experience) may have contributed to variability.

Third, the included studies employed different measurement instruments with varying dimensional structures, which may limit direct comparability despite score normalization procedures. Similar concerns regarding measurement specificity and construct interpretation have been noted in the broader literature on teacher self-efficacy (Tschannen-Moran and Hoy, 2001).

Finally, the cross-sectional nature of the primary studies precludes causal interpretation. The findings should therefore be interpreted as descriptive patterns rather than evidence of directional effects.

For these reasons, the findings should be interpreted cautiously and understood as reflecting perceived rather than objectively verified teaching competence.

Acknowledgments

Parts of this work were presented as an oral presentation at the 2nd International Black Sea Educational Research Congress (October 31, 2025). An abstract version was published in the conference abstract book (E-ISBN: 978-625-96460-1-5). No full paper has been published from this presentation. The submitted manuscript substantially extends the conference abstract/presentation.

Funding Statement

The author(s) declared that financial support was not received for this work and/or its publication.

Footnotes

Edited by: Neil Dagnall, Manchester Metropolitan University, United Kingdom

Reviewed by: Wen Li, Universiti Sains Malaysia, Malaysia

Leonidas Gavrilas, University of Ioannina, Greece

Data availability statement

Publicly available datasets were analyzed in this article. These data were extracted from the published articles listed in the reference list and Appendix 1. Further inquiries can directed to the corresponding author.

Author contributions

CS: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was used in the creation of this manuscript. All stages of the study were carried out independently by the author. An AI-based tool (ChatGPT) provided limited technical assistance only in the form of grammatical refinement, the formatting of some formulae for Word/Excel, and the formatting of the forest plot figure for RQ1. These contributions do not affect the scientific originality, data integrity, or validity of the findings. Responsibility for content creation, statistical analysis, data interpretation, and conclusions lies solely with the author.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg.2026.1767641/full#supplementary-material

supplementary_file_1.docx (679.3KB, docx)

References

  1. Açılmış H., Kayıran B. K. (2021). An analysis of affective factors of in-service and preservice primary school teachers about music education of the article. Opus Int. J. Soc. Res. 18, 4746–4773. doi: 10.26466/opus.890948 [DOI] [Google Scholar]
  2. Afacan Ş. (2008). The music education self efficacy scale. Ahi Evran Univ. J. Kirsehir Educ. Fac. 9, 1–11. [Google Scholar]
  3. Akpınar E., Kaya İ. (2025). Examination of preschool teachers music education self-efficacy beliefs. Kirikkale Univ. J. Soc. Sci. 15, 261–279. [Google Scholar]
  4. Aksu G., Güzeller C. O., Eser M. T. (2019). The effect of the normalization method used in different sample sizes on the success of artificial neural network model. Int. J. Assess. Tools Educ. 6, 170–192. doi: 10.21449/ijate.479404 [DOI] [Google Scholar]
  5. Altındaş D. (2023). Investigation of preschool teachers self-sufficiency and metaphoric perceptions regarding musical activities. [master’s thesis]. Trabzon: Trabzon University. [Google Scholar]
  6. Bandura A. (1997). Self-Efficacy: The Exercise of Control. New York: Freeman. [Google Scholar]
  7. Barnes L. R., Shinn-Taylor C. (1988). Teacher competency and the primary school curriculum: a survey of five schools in north-East England. Br. Educ. Res. J. 14, 283–295. [Google Scholar]
  8. Begg C. B., Mazumdar M. (1994). Operating characteristics of a rank correlation test for publication bias. Biometrics 50:1088. [PubMed] [Google Scholar]
  9. Borenstein M., Hedges L. V., Higgins J. P. T., Rothstein H. R. (2021). Introduction to Meta-Analysis. Hoboken: Wiley. [Google Scholar]
  10. Borenstein M., Higgins J. P. T., Hedges L. V., Rothstein H. R. (2017). Basics of meta-analysis: I2 is not an absolute measure of heterogeneity. Res. Synth. Methods 8, 5–18. doi: 10.1002/jrsm.1230 [DOI] [PubMed] [Google Scholar]
  11. Bresler L. (1993). Music in a double-bind: instruction by non-specialists in elementary schools. Bull. Counc. Res. Music. Educ. 95, 1–13. [Google Scholar]
  12. Burak S. (2019). Self-efficacy of pre-school and primary school pre-service teachers in musical ability and music teaching. Int. J. Music. Educ. 37, 257–271. doi: 10.1177/0255761419833083 [DOI] [Google Scholar]
  13. Çelik Y. (2018). The Investigation Between the Attitudes of Prospective Class Teachers on Teaching Music, Their Self-Efficiency Beliefs and Self Efficiencies About Their Music Abilities. [master’s thesis. Amasya: Amasya University. [Google Scholar]
  14. Çelik Y., Yetim H. (2017). The investigation of the relationship between the class of teacher’s nominated teachers ‘relationships on music teaching and musical teaching and self-satisfaction beliefs. Electron. J. Educ. Sci. 6, 150–170. [Google Scholar]
  15. Çevik D. B. (2011). Examining elementary education pre-service teachers’ self-efficacy beliefs. Ahi Evran University J. Kırşehir Educ. Faculty 12, 145–168. [Google Scholar]
  16. Chung D. F. M. Y. (2021). The impact of music pedagogy education on early childhood teachers’ self-efficacy in teaching music: the study of a music teacher education program in Hong Kong. Pacific Early Childhood Educ. Res. Assoc. 15, 63–86. doi: 10.17206/apjrece.2021.15.2.63 [DOI] [Google Scholar]
  17. Cohen J. (1988). Statistical Power Analysis for the Behavioral Sciences. New York: Erlbaum. [Google Scholar]
  18. de Vries P. (2011). The first year of teaching in primary school: where is the place of music? Int. J. Educ. Arts 12, 1–25. [Google Scholar]
  19. de Vries P. (2013). Generalist teachers’ self-efficacy in primary school music teaching. Music. Educ. Res. 15, 375–391. doi: 10.1080/14613808.2013.829427 [DOI] [Google Scholar]
  20. Demiralay H. (2023). Examining classroom teachers and teacher candidates’ attitudes towards the use of technology in music education and their self-efficacy levels towards music teaching. [master’s thesis]. Antalya: Akdeniz University. [Google Scholar]
  21. Dinçer S. (2021). Eğitim Bilimlerinde Uygulamalı Meta-Analiz [Applied Meta-Analysis in Educational Sciences]. Ankara: Pegem. [Google Scholar]
  22. Duval S., Tweedie R. (2000). Trim and fill: a simple funnel-plot–based method of testing and adjusting for publication bias in meta-analysis. Biometrics 56, 455–463. doi: 10.1111/j.0006-341X.2000.00455.x, [DOI] [PubMed] [Google Scholar]
  23. Egger M., Smith G. D., Schneider M., Minder C. (1997). Bias in meta-analysis detected by a simple, graphical test. BMJ 315, 629–634. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Ertekin Kaya M. N. (2024). Müzik öğretimi özyeterlik düzeyinde cinsiyet farklılıkları: Bir meta analiz çalışması [Gender differences in self-efficacy levels in music teaching: A meta-analysis study]. Ulusal Eğitim Akademisi Dergisi 8, 182–192. doi: 10.32960/uead.1536350 [DOI] [Google Scholar]
  25. Garvis S. (2013). Beginning generalist teacher self-efficacy for music compared with maths and English. Br. J. Music Educ. 30, 85–101. doi: 10.1017/S0265051712000411 [DOI] [Google Scholar]
  26. Garvis S., Pendergast D. (2011). An investigation of early childhood teacher self-efficacy beliefs in the teaching of arts education. Int. J. Educ. Arts 12, 1–16. [Google Scholar]
  27. Gülle A., Akay C. (2019). Analyzing the pre-service primary school teachers’ music education self-efficacy and their written-visual metaphors. J. Int. Soc. Res. 12, 1043–1057. [Google Scholar]
  28. Hallam S., Burnard P., Robertson A., Saleh C., Davies V., Rogers L., et al. (2009). Trainee primary-school teachers’ perceptions of their effectiveness in teaching music. Music. Educ. Res. 11, 221–240. doi: 10.1080/14613800902924508 [DOI] [Google Scholar]
  29. Higgins J., Thomas J. (2019). Cochrane Handbook for Systematic Reviews of Interventions. Newark: Wiley-Blackwell. [Google Scholar]
  30. Higgins J., Thompson S. G., Deeks J. J., Altman D. G. (2003). Measuring inconsistency in meta-analyses. BMJ 327, 557–560. doi: 10.1136/bmj.327.7414.557, [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Holden H., Button S. (2006). The teaching of music in the primary school by the non-music specialist. Br. J. Music Educ. 23, 23–38. doi: 10.1017/S0265051705006728 [DOI] [Google Scholar]
  32. IntHout J., Ioannidis J. P. A., Rovers M. M., Goeman J. J. (2016). Plea for routinely presenting prediction intervals in meta-analysis. BMJ Open 6:e010247. doi: 10.1136/bmjopen-2015-010247, [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Jeanneret N. (1997). Model for developing preservice primary teachers’ confidence to teach music. Bull. Counc. Res. Music. Educ. 133, 37–44. [Google Scholar]
  34. Jorgensen E. R. (2008). The Art of Teaching Music. Bloomington & Indianapolis: Indiana University Press. [Google Scholar]
  35. Kavaklı H. (2022). Comparison of Music Self-Efficacy Levels of Primary School Teacher Candidates and Preschool Teacher Candidates. [master’s thesis]. Balıkesir: Balıkesir University. [Google Scholar]
  36. Kaya Ö. (2022). The Effect of Cooperative Learning Method on Pre-school Teacher Candidates in Self-Efficacy, Success and Music Literacy in Music Lessons. [doctoral dissertation]. Malatya: Inönü University. [Google Scholar]
  37. Kim J., Choy D. (2008). Learning to toot your own horn: preservice teachers integrating music into a childhood classroom. J. Res. Child. Educ. 22, 405–423. doi: 10.1080/02568540809594636 [DOI] [Google Scholar]
  38. Koca Ş. (2013a). An investigation of music teaching self efficacy levels of preschool teachers. Fine Arts 8, 164–175. [Google Scholar]
  39. Koca Ş. (2013b). An investigation of music teaching self-efficacy levels of prospective preschool teachers. Educ. Res. Rev. 8, 897–900. [Google Scholar]
  40. Koca Ş. (2016). Self-efficacy perceptions of pre-service preschool teachers towards music activities. Mehmet Akif Ersoy Univ. J. Educ. Fac. 1, 375–386. [Google Scholar]
  41. Lowe G. M., Lummis G. W., Morris J. E. (2017). Pre-service primary teachers’ experiences and self-efficacy to teach music: are they ready? Issues Educ. Res. 27, 314–329. [Google Scholar]
  42. PRISMA statement (2020). PRISMA statement. Available online at: https://www.prisma-statement.org [Accessed October 8, 2025].
  43. Saylam Kırcıoğlu Ç. (2009). Determination of primary school teachers’ music-teaching knowledge level, self-efficacy perceptions, and attitudes. [master’s thesis]. Çanakkale: Çanakkale Onsekiz Mart University. [Google Scholar]
  44. Şeker S. S., Çilingir V. (2022). Examining the relationship between primary school teachers’ attitudes towards music education and their self-efficacies regarding music education. J. Yüzüncü Yıl Univ. Soc. Sci. Inst. 57, 74–85. doi: 10.53568/yyusbed.1140270 [DOI] [Google Scholar]
  45. Şen E. (2024). Developing the music listening skills test (MLST) for pre-school and elementary school teacher candidates. Int. J. Music. Educ. 44, 222–236. doi: 10.1177/02557614241268010 [DOI] [Google Scholar]
  46. Sonakın S. (2022). Investigation of pre-school teachers’ self-efficiencies towards music teaching. [master’s thesis]. Istanbul: Marmara University. [Google Scholar]
  47. Taş S., Atılgan D. S. (2022). The relationship between classroom teachers’ self-efficacy in teaching music with their music teaching skills and musical development. Educ. Sci. 47, 95–110. doi: 10.15390/EB.2022.10944 [DOI] [Google Scholar]
  48. Temmerman N. (1997). An investigation of undergraduate music education curriculum content in primary teacher education programmes in Australia. Int. J. Music. Educ. 30, 26–34. [Google Scholar]
  49. Topoğlu O. (2014). Investigating the classroom teacher candidates’ self-efficacy believes towards music teaching in terms of miscellaneous variables. Int. J. Hum. Sci. 11, 730–743. doi: 10.14687/ijhs.v11i2.3020 [DOI] [Google Scholar]
  50. Tschannen-Moran M., Hoy A. W. (2001). Teacher efficacy: capturing an elusive construct. Teach. Teach. Educ. 17, 783–805. doi: 10.1016/S0742-051X(01)00036-1 [DOI] [Google Scholar]
  51. Umuzdaş S., Işıldak C. K. (2018). The effect of music lessons on self- sufficiency levels of classroom teacher candidates. Van Yüzüncü Yıl University J. Educ. 15, 1316–1331. [Google Scholar]
  52. Valdebenito K., Almonaci-Fierro A. (2022). Teacher self-efficacy in music teaching: systematic literature review 2011-2021. J. Curric. Teach. 11:317. doi: 10.5430/jct.v11n8p317 [DOI] [Google Scholar]
  53. Vannatta-Hall J. E. (2010). Music Education in Early Childhood Teacher Education: The Impact of a Music Methods Course on Pre-Service Teachers’ Perceived Confidence and Competence to Teach Music. [doctoral dissertation]. Urbana-Champaign: University of Illinois. [Google Scholar]
  54. Yapalı Y., Çakır İlhan A., Sever S. (2025). Problems experienced in the field of vocational specialization for early childhood music education in Turkey and a solution proposal. Educ. Sci. 50, 21–44. doi: 10.15390/EB.2025.13289 [DOI] [Google Scholar]
  55. Yegül B. U., (2014). Examination of music education self-sufficiency perceptions of teacher candidates. 9th International Balkan Education and Science Congress, Edirne. [Google Scholar]
  56. Yıldız G. (2017). Development of music education self-efficacy scale for preschool teacher candidates. Mehmet Akif Ersoy University J. Soc. Sci. Institute 9, 396– 407. doi: 10.20875/makusobed.298242 [DOI] [Google Scholar]
  57. Yücesan E. (2023). Pre-service pre-school teachers’ music education self-efficacy levels: A case study. Kastamonu Education Journal 31, 60–69. doi: 10.24106/kefdergi.1246438 [DOI] [Google Scholar]
  58. Zelenak M. S. (2024). Self-efficacy and music performance: a meta-analysis. Psychol. Music 52, 649–667. doi: 10.1177/03057356231222432 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplementary_file_1.docx (679.3KB, docx)

Data Availability Statement

Publicly available datasets were analyzed in this article. These data were extracted from the published articles listed in the reference list and Appendix 1. Further inquiries can directed to the corresponding author.


Articles from Frontiers in Psychology are provided here courtesy of Frontiers Media SA

RESOURCES