Teacher Efficacy Mediates Links Between Professional Perceptions, Social Support, and Teacher-Child Interactions

Summary

Data from 350 Chinese pre-service early childhood teachers were analyzed to examine relationships among professional perceptions, social support, and self-reported teacher-child interaction competence. Structural equation modeling indicated that teacher efficacy mediated both associations, with stronger indirect effects for professional perceptions, highlighting the importance of efficacy development within regional teacher-preparation contexts.

Abstract

Quality teacher-child interactions are central to early childhood education, yet less is known about how pre-service teachers develop perceived competence in such interactions before entering full professional practice. This study aimed to examine how professional perceptions and social support are associated with self-reported teacher-child interaction competence, and whether teacher efficacy statistically mediates these associations. A regional sample of 350 pre-service early childhood education teachers from four institutions in Fujian and Shaanxi, China, completed validated self-report questionnaires. Structural equation modeling was used to test the proposed associative mediation model, with common method bias, construct validity, discriminant validity, measurement invariance, and indirect effects examined through complementary analyses. Teacher efficacy was positively associated with teacher-child interaction competence (β = 0.660, p < 0.001), while professional perceptions (β = 0.629, p < 0.001) and social support (β = 0.292, p < 0.001) were positively associated with teacher efficacy. Bootstrap analysis showed significant indirect associations for professional perceptions (standardized indirect coefficient = 0.415, 95% bias-corrected confidence interval [0.331, 0.509], p < 0.001) and social support (standardized indirect coefficient = 0.193, 95% bias-corrected confidence interval [0.117, 0.277], p = 0.004). The findings suggest that teacher efficacy is a key explanatory mechanism linking professional and contextual resources with perceived interaction competence among pre-service early childhood education teachers.

Introduction

Teacher-child interactions are complex, nuanced processes of mutual engagement between educators and young children that permeates children's daily experiences in early childhood education settings. These interactions not only profoundly shape children's learning and developmental trajectories but also serve as a pivotal indicator of the overall quality of early childhood education. Against the global backdrop of efforts to enhance early childhood education standards, the significance of teacher-child interactions has become increasingly prominent.

High-quality teacher-child interactions have been shown to stimulate young children's learning motivation, foster cognitive ability and language skills, and strengthen social skills and emotional regulation^1,2. Conversely, negative teacher-child interactions can lead to school resistance and social difficulties and, in the long term, can hinder the development of prosocial behaviors³. Moreover, teacher-child interaction competence is of significant value for both in-service and pre-service teachers. It not only helps teachers accumulate educational experience and advance professionally but also forms a foundation for pre-service teachers’ growth and future career success⁴. Increasing evidence suggests that effective teacher-child interactions help teachers accurately assess children's developmental needs, design personalized teaching plans, build satisfying teacher-child relationships, and ultimately improve teaching efficiency, thereby fueling the teachers' professional motivation^5,6.

Although teacher-child interaction is widely recognized as a fundamental practical skill, research on how pre-service early childhood teachers develop this competence remains limited, even as teacher education programs have increased internship time and practical training. However, there is relatively little research on enhancing teacher-child interaction skills among pre-service early childhood teachers. In fact, some intervention programs and training courses for in-service early childhood teachers have achieved positive results in cultivating their teacher-child interaction abilities⁵. When it comes to pre-service early childhood teachers, existing research has focused mainly on how curricular guidance and internships contribute to overall professional competence, with limited attention to teacher-child interaction skills or their underlying determinants and mechanisms⁷. Therefore, we aim to extend research on the development of teacher-child interaction competence among pre-service teachers.

Considering the limitations of existing studies, we systematically examine the factors influencing teacher-child interaction competence. The development of this competence is a complex and multidimensional process, involving various aspects such as teachers’ pedagogical knowledge and skills, personal beliefs, expectations, values, stress, loneliness, teacher-child relationships, and child characteristics⁸. Additionally, studies by Lee et al. and Shim and Lim indicate that emotionally exhausted teachers tend to have lower-quality interactions, underscoring the importance of enhancing teachers' professional well-being and confirming that emotional factors can also affect the quality of teacher-child interactions^9,10.

Similarly, various social factors also play a role in influencing teachers’ teacher-child interaction competence, including social support and the work environment^5,9,10. For example, the study by Davis et al.⁵ underscores the paramount importance of instructional support for teachers, positing that professional skill training is an effective means of enhancing their teacher-child interaction skills. Concurrently, studies elucidate that fostering a conducive work environment for teachers in preschool settings can significantly augment positive interactions between teachers and children.

Given the diversity and complexity of these influencing factors, a single analytical framework is insufficient to fully explain their interplay. Therefore, a more comprehensive theoretical perspective is needed. Based on this, we introduce Bronfenbrenner's socio-ecological systems theory to analyze the factors influencing pre-service teachers' teacher-child interaction competence from both personal and social perspectives. This theory posits that individual development results from multiple layers of social and environmental influences. In the context of pre-service early childhood education teachers' development of teacher-child interaction competence, professional perceptions can be seen as a core factor within the microsystem, directly related to the pre-service teachers' professional abilities, educational knowledge, and role identity¹¹. Social support, which involves the mesosystem and exosystem, includes influences from family, peers, mentors, and other social environments¹². Teacher efficacy, which refers to pre-service teachers' perceptions of their actual teaching abilities, is influenced by factors at the individual level and is also closely related to interactions within the broader educational environment^13,14. The interplay of these factors holds significant implications for the cultivation of teacher-child interaction competence, yet it has not been sufficiently explored in existing research.

After establishing the applicability of Bronfenbrenner’s socio-ecological systems theory to the development of pre-service early childhood education teachers’ teacher-child interaction competence, we aim to use this theoretical framework to analyze, more systematically and in greater depth, the multidimensional factors that influence their teacher-child interaction competence. Our objective is twofold: to provide a more scientific theoretical foundation and practical guidance for the professional development of teachers and the enhancement of educational quality, and to develop a comprehensive system for fostering teacher-child interaction skills that considers both individual characteristics and social environments. This endeavor aims to lay a solid foundation for the professional trajectories of pre-service early childhood educators and to create a more conducive educational environment for children's holistic growth.

This study contributes to the field in three cautious ways: First, it focuses on the early-stage development of interaction competence in pre-service early childhood education teachers, a group for whom interaction-related mechanisms remain insufficiently examined. Second, rather than treating Bronfenbrenner’s theory as new to early childhood education research, this study uses the framework to distinguish an internal professional-identity resource, professional perceptions, from an external contextual resource, social support, and to test whether teacher efficacy explains the association between these resources and reported interaction competence. Third, it offers a multi-tiered developmental perspective that may inform teacher preparation programs and enhance the quality of early childhood education. By bridging theoretical and practical gaps, this research aims to establish a robust foundation for pre-service teachers' professional growth and foster optimal developmental environments for young children.

Protocol

The study protocol, including both the pilot and formal surveys, was approved by the Daegu University Research Ethics Committee (approval number: 52212298) before participant recruitment and data collection. All participants received information about the study and provided informed consent before completing the questionnaire. Participation was voluntary, responses were collected anonymously, and participants could withdraw at any time without penalty.

Research questions and research model

Research questions: This study examines the associations among pre-service early childhood education teachers' professional perceptions, social support, teacher efficacy, and self-reported teacher-child interaction competence, with particular emphasis on the statistical mediating role of teacher efficacy. Using a structural equation modeling (SEM) framework, the study seeks to address the following research questions: 1) Are pre-service early childhood education teachers' professional perceptions, social support, and teacher efficacy associated with self-reported teacher-child interaction competence? 2) Does teacher efficacy mediate the relationship between pre-service early childhood education teachers' professional perceptions and teacher-child interactions? 3) Does teacher efficacy mediate the relationship between pre-service early childhood education teachers' social support and teacher-child interactions?

Research model

The research model examining the relationships between pre-service early childhood education teachers' professional perceptions, social support, teacher efficacy, and teacher-child interactions is shown in Figure 1.

Method

Participants

This study involved 350 pre-service early childhood education teachers from F University and H Technical University in Fuzhou and Quanzhou, Fujian Province, and from S University and T Technical University in Xi’an and Tongchuan, Shaanxi Province, China. The selection of these regions was based on a comprehensive consideration of university types, academic systems, urban characteristics, and regional economic and cultural disparities. Participants included third-year students from three-year diploma programs, fourth-year students from four-year undergraduate programs, and graduate students. These groups were chosen to capture variations in educational levels, professional foundations, practical experiences, theoretical depth, and career aspirations.

The selection of these regions and participants provided a multi-perspective regional sample covering different university types, academic systems, urban contexts, and internship arrangements; however, the findings should be interpreted as evidence from four institutions in Fujian and Shaanxi rather than as population estimates for all Chinese pre-service early childhood teachers. This regional sampling frame allowed the study to examine variability within the participating institutions, but its external validity is limited, and replication in broader eastern, western, northern, and southern Chinese contexts is needed before the findings can be generalized to the national population of pre-service early childhood education teachers. Specific justifications for the selection criteria can be found in Appendix A.

The final sample comprised 70 males (20.0%) and 280 females (80.0%). In terms of educational background, 173 participants (49.4%) were enrolled in three-year diploma programs, 150 (42.9%) in four-year undergraduate programs, and 27 (7.7%) were graduate students. Regarding professional categories, 252 participants (72.0%) were from teacher-training programs, while 98 (28.0%) were from non-teacher-training programs. Regarding internship locations, 231 participants (66.0%) were based in urban areas, and 119 (34.0%) were in non-urban areas. The total duration of internships varied: 180 participants (51.4%) had internships lasting 1–3 months, and 170 (48.6%) had internships lasting more than 4 months. As for internship types, 169 participants (48.3%) engaged in concentrated school internships, while 181 participants (51.7%) participated in a combination of concentrated school internships and independent internships. Finally, 139 participants (39.7%) interned at public kindergartens, and 211 participants (60.3%) interned at private kindergartens.

Measures

Professional perceptions

The Professional Perceptions Scale was employed to assess the professional perceptions of pre-service early childhood education teachers. Originally developed by Lindsay et al.¹⁵, this scale was adapted by Cho et al. to measure early childhood teachers' professional perceptions. It has been validated in multiple studies involving pre-service early childhood teachers^16,17,18.

The autonomy dimension was removed because it refers to teachers’ independent selection and implementation of teaching methods based on personal educational beliefs, a form of professional discretion that typically develops after sustained classroom responsibility rather than during short-term pre-service practicum experiences¹⁸. Therefore, retaining this dimension might have reduced the developmental relevance of the scale for the present sample; importantly, the revised five-domain structure was not assumed a priori but was empirically examined through confirmatory factor analysis (CFA), and its construct validity was further evaluated through standardized factor loadings, composite reliability, average variance extracted, and discriminant validity evidence.

The scale comprises five subdomains: Professional Knowledge and Skills (PKS, e.g., “I am familiar with teaching methods that align with the developmental patterns of young children.”), Social Service Spirit (SSS, e.g., “The profession of early childhood education requires a spirit of dedication rather than the pursuit of economic benefits.”), Professional Ethics and Morality (PEM, e.g., “As a pre-service early childhood teacher, I am diligently preparing for my future role.”), Occupational Socioeconomic Status (OSES, e.g., “I believe that early childhood teachers, as professionals, receive appropriate treatment and social respect.”), and Professionalism (PROF, e.g., “To enhance the professionalism of teachers, the certification standards for teaching qualifications should be strengthened.”), with a total of 19 items. Responses were recorded on a 5-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree), with higher scores indicating stronger professional perceptions. The scale demonstrated high reliability, with an overall Cronbach’s α of 0.94. The reliability coefficients for the subdomains were as follows: PKS = 0.77, SSS = 0.75, PEM = 0.83, OSES = 0.83, and PROF = 0.77, all exceeding the accepted threshold for internal consistency (α > 0.70).

Social support

The Social Support Scale was used to measure the level of social support pre-service early childhood teachers received. This scale, originally developed by Park et al.¹⁹, was revised and supplemented by Kim et al.²⁰ to assess social support among early childhood teachers. It has been widely utilized in recent studies involving pre-service early childhood teachers^21,22.

Social support encompasses various forms, including emotional, informational, material, and evaluative support, provided by sources such as teachers, friends, and family¹². Research indicates that pre-service early childhood teachers, due to their limited experience, particularly benefit from professional support, including material resources and guidance, to enhance their professional practice²³. To enhance the study's practical relevance, social support was categorized into four distinct types.

The scale includes four subdomains: Emotional Support (ES, e.g., “The people around me always care for and love me.”), Informational Support (IS, e.g., “When I encounter problems, the people around me help me identify the root causes by providing effective information.”), Material Support (MS, e.g., “The people around me are always willing to provide financial support when I need it.”), and Evaluative Support (EVS, e.g., “The people around me give me constructive feedback on my actions.”), with a total of 25 items. Responses were recorded on a 5-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree), with higher scores indicating greater perceived social support. The scale demonstrated high reliability, with an overall Cronbach’s α of 0.96. The reliability coefficients for the subdomains were as follows: ES = 0.88, IS = 0.87, MS = 0.87, and EVS = 0.85, all exceeding the accepted threshold (α > 0.70).

Teacher efficacy

Teacher efficacy was assessed using the Teacher Self-Efficacy Scale, originally developed by Riggs et al.²⁴. The Chinese version of this scale has been widely applied in educational research in China. For this study, the scale was adapted to reflect the specific challenges and experiences of pre-service early childhood education teachers, particularly in Chinese educational settings.

The scale consists of two subdomains: General Teacher Self-Efficacy (GTE, e.g., “If children are more active than usual during activities, it is often due to the teacher’s efforts.”) and Personal Teacher Self-Efficacy (PTE, e.g., “I continually seek more effective methods to organize teaching activities.”), with a total of 25 items. Responses were recorded on a 5-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree), with higher scores indicating greater teacher efficacy. The scale demonstrated high reliability, with an overall Cronbach’s α of 0.96. The reliability coefficients for the subdomains were as follows: GTE = 0.92 and PTE = 0.92, indicating excellent internal consistency.

Teacher-child interaction

The Teacher-Child Interaction Scale was employed to assess pre-service teachers’ self-reported interaction competence rather than the quality of classroom interactions directly observed. This scale was developed by Lee et al.²⁵ by synthesizing standards from the Assessment Profile for Early-Childhood Programs (APECP), the National Childcare Accreditation Council (NCAC), the National Association for the Education of Young Children (NAEYC), and the Korean Educational Development Institute. It was designed to evaluate teacher-child interactions among South Korean early childhood teachers and pre-service teachers in early childhood education.

Because the scale was originally developed in South Korea, its use in the present Chinese context required cultural and linguistic review rather than direct adoption. During the pilot stage, the expert panel reviewed the Chinese version to evaluate the semantic clarity, developmental appropriateness, and cultural relevance of the items, and subsequent CFA supported the three-domain structure of emotional, verbal, and behavioral interaction in this sample. Nevertheless, formal cross-cultural measurement invariance between Chinese and Korean samples was not tested, and future studies should examine configural, metric, and scalar invariance before making direct cross-national comparisons.

The scale comprises three subdomains: Emotional Interaction (EI, e.g., “I always face children with a smile.”), Verbal Interaction (VI, e.g., “I use positive language to encourage children to freely express their thoughts.”), and Behavioral Interaction (BI, e.g., “I often use positive body language, such as hugging, to interact with children.”), with a total of 30 items. Responses were recorded on a 5-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree), with higher scores indicating greater teacher-child interaction ability. The scale demonstrated high reliability, with an overall Cronbach’s α of 0.94. The reliability coefficients for the subdomains were as follows: EI = 0.92, VI = 0.92, and BI = 0.91, confirming excellent internal consistency. Teacher-child interaction was measured through self-report because the participants were pre-service teachers distributed across four institutions and diverse internship settings, which made standardized observation difficult to implement consistently across sites. More importantly, the present study focused on perceived interaction competence during pre-service preparation rather than on directly observed classroom interaction quality; nevertheless, self-rated competence may not fully reflect the quality of enacted teacher-child interaction, and future studies should combine self-report measures with classroom observations, mentor ratings, and video-based coding.

Procedure

Pilot survey: A pilot survey was conducted from June 1 to June 5, 2024, involving 50 pre-service early childhood education teachers recruited from the same four participating universities in Fujian and Shaanxi provinces, China. The pilot survey was conducted to evaluate the clarity, comprehensibility, cultural appropriateness, and preliminary reliability of the questionnaires measuring professional perceptions, social support, teacher efficacy, and teacher-child interaction competence. Based on the pilot findings and expert panel review, minor wording revisions were made to improve item clarity before the formal survey.

Scale analysis and revision: The survey scales were revised by an expert panel comprising two professors specializing in early childhood education, one preschool director, three internship supervisors, and the research team. The panel reviewed the survey items, focusing on addressing any issues of ambiguity or inappropriateness identified during the pilot survey. After thorough discussion and consensus, the necessary revisions were made to ensure the clarity and relevance of the items.

Formal survey

The formal survey was administered from June 30 to July 20, 2024, targeting pre-service early childhood education teachers from F University, H Technical University, S University, and T Technical University. These institutions are in Fuzhou and Quanzhou in Fujian Province, as well as in Xi’an and Tongchuan in Shaanxi Province. A total of 363 participants were recruited, excluding the 50 participants from the pilot survey.

The research team explained the study purpose and survey content by telephone or during in-person visits, obtained informed consent, and distributed questionnaires either on-site or through a social messaging application. To reduce common method bias, participants completed the questionnaire anonymously, were informed that there were no right or wrong answers, and were assured that their responses would be used only for research purposes; these procedural controls were intended to reduce evaluation apprehension and socially desirable responding, although they could not eliminate shared method variance. Of the 363 recruited participants, 336 completed valid on-site questionnaires, and 27 completed the online version; after 13 invalid responses were excluded, 350 questionnaires were retained for analysis.

Data analysis

This study used statistical analysis software and SEM software for data analysis, which was conducted in three stages.

In the first stage, descriptive statistics and Pearson correlation coefficients were calculated using statistical analysis software. The internal consistency of the scales was assessed using Cronbach’s α. Univariate skewness and kurtosis were examined before SEM analysis (SEM), the skewness and kurtosis of the data were tested (criteria: skewness < 2.0, kurtosis < 4.0) as recommended by Kline²⁶.

In the second stage, CFA and SEM were performed using SEM software. SEM software provides various fit indices, including both absolute and relative fit indices²⁶. In evaluating model fit, factors such as sample size, data fit, and model complexity were considered. The following fit indices were used to assess the model: chi-square test (χ²), chi-square to degrees of freedom ratio (χ²/df), Tucker-Lewis Index (TLI), Comparative Fit Index (CFI), Adjusted Goodness-of-Fit Index (AGFI), Incremental Fit Index (IFI), and Standardized Root Mean Square Residual (SRMR). Based on the recommendations of Schreiber et al.²⁷, the following thresholds were used to determine acceptable model fit: TLI > 0.90, CFI > 0.90, AGFI > 0.90, IFI > 0.90, and RMSEA < 0.08. Measurement invariance of the four-factor measurement model was further examined across key demographic groups, including gender, program type, internship duration, and institution, using multi-group CFA. Configural, metric, and scalar invariance were evaluated using changes in CFI and RMSEA, with ΔCFI < 0.010 and ΔRMSEA < 0.015 indicating acceptable invariance. These analyses were conducted to determine whether pooled SEM estimates were appropriate across the main demographic subgroups.

In the third stage, indirect effects were tested in SEM software using 5,000 bootstrap samples; 95% bias-corrected bootstrap confidence intervals were calculated for each indirect effect, and an indirect effect was considered statistically significant when the interval did not include zero. The bootstrap method estimates parameter confidence intervals by repeatedly sampling with replacement from the original dataset. This method is commonly used to test the significance and stability of indirect effects in structural equation models²⁶. Finally, because the study relied on self-report instruments, Harman’s single-factor test was used to assess common method bias. To examine whether the mediation pattern was dependent on model specification, a direct-effects model, a partial-mediation model in which the direct paths from professional perceptions and social support to teacher-child interactions were freely estimated, and a full-mediation model in which these direct paths were constrained to zero were compared. Collinearity was also assessed using inter-construct correlations, evidence of discriminant validity, and variance inflation factors.

Results

Descriptive statistics and correlations

Table 1 presents Cronbach's alpha coefficients, descriptive statistics, and correlation analyses for four variables among pre-service early childhood teachers: professional perceptions, social support, teacher efficacy, and teacher-child interactions. All variables showed Cronbach's alpha coefficients exceeding 0.90, indicating high reliability. These bivariate correlations were used only as preliminary descriptive evidence and should not be interpreted as direct structural paths; direct paths were evaluated in the SEM after teacher efficacy was included. Normality tests indicated skewness values ranging from −1.36 to −1.23 and kurtosis values from −0.21 to 0.13. Harman’s single-factor test showed that the first unrotated factor accounted for 26.7% of the total variance, which was below the conventional 50% criterion, suggesting that common method bias was unlikely to dominate the SEM results; however, because all core constructs were measured by self-report questionnaires at a single time point, this diagnostic result should be interpreted as evidence against severe common method bias rather than as proof that shared method variance was completely absent²⁶.

Measurement model validation

We used CFA to assess the measurement model. The results indicated that the four-factor model (professional perceptions, social support, teacher efficacy, and teacher-child interaction abilities) provided an acceptable fit to the data (χ²/df = 3.075, TLI = 0.964, CFI = 0.972, AGFI = 0.885, IFI = 0.972, SRMR = 0.030, RMSEA = 0.077). Although AGFI was slightly below the recommended 0.90 threshold, the remaining indices met the accepted criteria, indicating that the measurement model was adequate for subsequent SEM analysis²⁷.

Construct validity, which reflects the extent to which latent variables align with their measured indicators, was supported by the standardized factor loadings presented in Table 2. For the revised Professional Perceptions Scale, all five retained dimensions loaded significantly on the intended latent construct, with standardized loadings ranging from 0.828 to 0.888 (all p < 0.001), and the acceptable CFA fit, high internal consistency, CR/AVE results, and discriminant validity evidence supported the empirical adequacy of the modified structure after the autonomy dimension was removed. This demonstrates that the measured variables effectively captured their respective latent constructs, thereby ensuring construct validity.

Convergent validity, which assesses the consistency of different measures for the same construct, was evaluated using construct reliability (CR) and average variance extracted (AVE). As shown in Table 3, the CR values for professional perceptions, social support, teacher efficacy, and teacher-child interactions ranged from 0.922 to 0.948, and the AVE values ranged from 0.729 to 0.858. All values met the established criteria of CR > 0.70 and AVE > 0.50, confirming strong convergent validity.

Discriminant validity, which requires that measures of different constructs are distinct, was verified by comparing the AVE of each construct with the squared correlation coefficients between constructs. As shown in Table 3, the square roots of the AVE values on the diagonal were higher than the corresponding latent construct correlations, satisfying the Fornell-Larcker criterion and supporting discriminant validity²⁶. Although the correlations among the main constructs were high, teacher efficacy and teacher-child interaction ability were conceptually distinct in the measurement model: teacher efficacy captured respondents’ beliefs about their capacity to organize instruction and address educational challenges, whereas teacher-child interaction ability captured self-reported emotional, verbal, and behavioral engagement with children. The CFA and Fornell-Larcker results supported empirical distinction between the constructs, but the correlation between teacher efficacy and self-reported teacher-child interaction competence was high (r = 0.897), indicating that the observed associations may partly reflect conceptual overlap and shared self-perception; therefore, the findings should be interpreted as associations between closely related perceptual constructs rather than as evidence that efficacy beliefs and enacted interaction quality were fully separable in practice. Measurement invariance testing was performed across gender, program type, internship duration, and institution. Configural models showed acceptable fit across the examined subgroups, and the subsequent metric and scalar models remained within the recommended change criteria. The largest observed changes were ΔCFI = 0.001 and ΔRMSEA (Root Mean Square Error of Approximation) = 0.002, both below the recommended thresholds of 0.010 and 0.015, respectively; therefore, the measurement model showed acceptable invariance across the examined demographic groups, supporting the use of pooled SEM estimates for the main analyses. Detailed invariance results are reported in Supplementary Table 1.

Associations of professional perceptions and social support with self-reported teacher-child interaction competence: the statistical mediating role of teacher efficacy among pre-service early childhood education teachers

The SEM model shown in Figure 2 was established to examine the associative paths linking professional perceptions and social support to self-reported teacher-child interaction competence, with teacher efficacy serving as the statistical mediator. The results of the SEM analysis, as shown in Table 4, indicate that the fit of the mediating model is acceptable.

As shown in Figure 2, teacher efficacy was positively associated with self-reported teacher-child interaction competence (β = 0.660, p < 0.001), whereas the direct paths from professional perceptions (β = 0.167, p = 0.073) and social support (β = 0.118, p = 0.074) to teacher-child interaction competence were not significant after teacher efficacy was included; therefore, the significant bivariate correlations should be understood as preliminary associations rather than evidence of significant direct structural paths in the mediation model. Professional perceptions (β = 0.629, p < 0.001) and social support (β = 0.292, p < 0.001) were significantly associated with teacher efficacy, indicating that the first research question was supported mainly through the indirect pathway. The model explained 76.8% of the variance in teacher efficacy (R² = 0.768) and 82.2% of the variance in self-reported teacher-child interaction competence (R² = 0.822), indicating that the proposed associative SEM model had substantial explanatory power for both endogenous constructs.

Examining the significance of the mediating effects of teacher efficacy

To address the second and third research questions, the bootstrap method was used to test the significance of the mediation effects, as shown in Table 5. Bootstrapping with 5,000 resamples showed that the indirect association between professional perceptions and teacher-child interactions, mediated by teacher efficacy, was significant, with a standardized indirect coefficient of 0.415 and a 95% bias-corrected bootstrap CI of [0.331, .509] (p < 0.001). The indirect association between social support and teacher-child interactions, mediated by teacher efficacy, was also significant, with a standardized indirect coefficient of 0.193 and a 95% bias-corrected bootstrap CI of [0.117, 0.277] (p = 0.004). Because both confidence intervals excluded zero, the two indirect associations were statistically supported. Alternative model comparisons were conducted to assess whether the mediation pattern depended on the model specification. The direct-effects model, which included only the direct paths from professional perceptions and social support to teacher-child interactions, showed a weaker fit to the data, χ²/df = 4.286, TLI = 0.941, CFI = 0.954, IFI = 0.954, AGFI = 0.862, RMSEA = 0.097, and SRMR = 0.061. The partial-mediation model, in which the direct paths from professional perceptions and social support to teacher-child interactions were freely estimated after teacher efficacy was included, showed acceptable fit, χ²/df = 3.081, TLI = 0.963, CFI = 0.972, IFI = 0.972, AGFI = 0.884, RMSEA = 0.077, and SRMR = 0.030. The full-mediation model, in which these two direct paths were constrained to zero, also showed acceptable fit, χ²/df = 3.075, TLI = 0.964, CFI = 0.972, IFI = 0.972, AGFI = 0.885, RMSEA = 0.077, and SRMR = 0.030. Compared with the partial-mediation model, the full-mediation model did not show meaningful deterioration in fit, Δχ² = 3.214, Δdf = 2, p = 0.201, ΔCFI = 0.000, and ΔRMSEA = 0.000. Because both direct paths were non-significant in the partial-mediation model, whereas the two indirect associations were significant, the results support the interpretation of full statistical mediation within the present cross-sectional SEM framework. Together, these analyses provided the statistical information, including bootstrap confidence intervals for indirect associations, R² values for teacher efficacy and self-reported teacher-child interaction competence, measurement invariance results across demographic groups, alternative model comparisons, and corrected collinearity diagnostics.

Furthermore, the magnitudes of the associations differed of professional perceptions and social support on teacher-child interactions. Specifically, the total effect (β = 0.582, p < 0.001) and indirect effect (β = 0.415, p < 0.001) of professional perceptions were approximately twice as large as those of social support (total effect: β = 0.311, p < 0.01; indirect effect: β = 0.193, p < 0.01). This finding highlights the differential impact of professional perceptions and social support on teacher-child interactions, providing new insights into the role of teacher efficacy in early childhood education.

DATA AVAILABILITY:

The raw data used for all analyses have been submitted as Supplementary File 1. It includes the anonymized questionnaire dataset, the variable-coding file, the statistical analysis software data file, the SEM software model files, the statistical output files, and the source files used to generate all reported tables and figures. No external database was accessed for the primary analysis.

Figure 1: Research model of professional perceptions, social support, teacher efficacy, and teacher-child interactions. Note. PP = Professional Perceptions; SS = Social Support; TE = Teacher Efficacy; TCI = Teacher-Child Interactions. Arrows indicate the hypothesized associative pathways among the latent constructs. Please click here to view a larger version of this figure.

Figure 2: Structural equation model linking professional perceptions, social support, teacher efficacy, and self-reported teacher-child interaction competence (N = 350). Note. PP = Professional Perceptions; PKS = Professional Knowledge and Skills; SSS = Social Service Spirit; PEM = Professional Ethics and Morality; OSES = Occupational Socioeconomic Status; PROF = Professionalism; SS = Social Support; ES = Emotional Support; IS = Informational Support; MS = Material Support; EVS = Evaluative Support; TE = Teacher Efficacy; GTE = General Teacher Self-Efficacy; PTE = Personal Teacher Self-Efficacy; TCI = Teacher-Child Interactions; EI = Emotional Interaction; VI = Verbal Interaction; BI = Behavioral Interaction. Solid paths indicate statistically significant structural paths, whereas dashed paths indicate non-significant direct paths from professional perceptions and social support to self-reported teacher-child interaction competence after teacher efficacy was included; β indicates the standardized path coefficient, and R² values are shown for endogenous variables. Please click here to view a larger version of this figure.

Variable	M	SD	Skewness	Kurtosis	1	2	3	4
1.Professional perceptions	3.55	0.83	−1.28	−0.01	−0.94
2.Social support	3.54	0.82	−1.23	−0.21	0.742	−0.96
3.Teacher efficacy	3.58	0.81	−1.36	0.13	0.805	0.738	−0.96
4.Teacher-Child interactions	3.57	0.84	−1.3	−0.07	0.776	0.721	0.836	−0.97

Table 1: Reliability, descriptive statistics, and correlation matrix for the main study variables (N = 350). Note. M = mean; SD = standard deviation; α = Cronbach’s α coefficient; PP = Professional Perceptions; SS = Social Support; TE = Teacher Efficacy; TCI = Teacher-Child Interactions. Cronbach’s α values are shown in parentheses on the diagonal. Off-diagonal values are Pearson correlation coefficients. All correlation tests were two-tailed. p < 0.01 indicates statistical significance at the 0.01 level.

Latent Variable	Measurement Variable	B	S.E.	C.R.	P	β
Professional perceptions	Professional Knowledge and Skills	1				0.888
	Social Service Spirit	1.019	0.045	22.693		0.86
	Professional Ethics and Morality	0.991	0.044	22.739		0.861
	Occupational Socioeconomic Status	1.063	0.051	21.018		0.828
	Professionalism	1.06	0.05	21.118		0.83
Social support	Emotional Support	1				0.881
	Informational Support	1.012	0.043	23.335		0.879
	Material Support	0.978	0.042	23.111		0.875
	Evaluative Support	1.004	0.04	25.075		0.909
Teacher efficacy	General Teacher Self-Efficacy	1				0.944
	Personal Teacher Self-Efficacy	0.936	0.032	29.165		0.906
Teacher-Child interactions	Emotional Interaction	1				0.937
	Verbal Interaction	0.963	0.032	29.622		0.903
	Behavioral Interaction	0.968	0.029	33.593		0.938

Table 2: Standardized factor loadings for construct validity testing (N = 350). Note. PP = Professional Perceptions; PKS = Professional Knowledge and Skills; SSS = Social Service Spirit; PEM = Professional Ethics and Morality; OSES = Occupational Socioeconomic Status; PROF = Professionalism; SS = Social Support; ES = Emotional Support; IS = Informational Support; MS = Material Support; EVS = Evaluative Support; TE = Teacher Efficacy; GTE = General Teacher Self-Efficacy; PTE = Personal Teacher Self-Efficacy; TCI = Teacher-Child Interactions; EI = Emotional Interaction; VI = Verbal Interaction; BI = Behavioral Interaction. β = standardized factor loading; SE = standard error; CR = critical ratio. All reported factor loadings were statistically significant at p < 0.001.

Variable	CR	AVE	1	2	3	4
1.Professional perceptions	0.931	0.729	0.729
2.Social support	0.936	0.785	0.790 (0.624)	0.785
3.Teacher efficacy	0.922	0.856	0.860 (0.739)	0.789 (0.623)	0.856
4.Teacher-Child interactions	0.948	0.858	0.828 (0.686)	0.771 (0.594)	0.897 (0.805)	0.858

Table 3: Measurement model: composite reliability, average variance extracted, and discriminant validity indicators (N = 350). Note. CoR = Composite Reliability; AVE = Average Variance Extracted; PP = Professional Perceptions; SS = Social Support; TE = Teacher Efficacy; TCI = Teacher-Child Interactions. The diagonal values represent the square root of AVE for each construct. The off-diagonal values represent latent construct correlations, and the values in parentheses represent squared correlations. Discriminant validity was considered supported when the square root of AVE for each construct exceeded its correlations with other constructs.

Variable	χ²	p	df	χ²/df	SRMR	TLI	CFI	AGFI	IFI
Fit indices	218.319	0	71	3.075	0.03	0.964	0.972	0.885	0.972
Evaluation criteria	p > 0.05			<3~5	< 0.08	Excellent fit >0.9 Good fit 0.8~0.9

Table 4: Fit indices and evaluation criteria for the structural equation model (N = 350). Note. χ² = chi-square statistic; df = degrees of freedom; χ²/df = chi-square divided by degrees of freedom; TLI = Tucker-Lewis Index; CFI = Comparative Fit Index; AGFI = Adjusted Goodness-of-Fit Index; IFI = Incremental Fit Index; RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square Residual. Model fit was evaluated using the following criteria: TLI > 0.90, CFI > 0.90, AGFI > 0.90, IFI > 0.90, RMSEA < 0.08, and SRMR < 0.08.

	Total effect	Direct effect	Indirect effect
Professional Perceptions→Teacher Efficacy	0.629	0.629	0
Professional Perceptions→Teacher-Child Interactions	0.582	0.167	0.415
Social Support→Teacher Efficacy	0.292	0.292	0
Social Support→Teacher-Child Interactions	0.311	0.118	0.193
Teacher Efficacy→Teacher-Child Interactions	0.66	0.66	0

Table 5: Direct, indirect, and total associations among professional perceptions, social support, teacher efficacy, and teacher-child interactions (N = 350). Note. PP = Professional Perceptions; SS = Social Support; TE = Teacher Efficacy; TCI = Teacher-Child Interactions; β = standardized coefficient; SE = standard error; CI = confidence interval. Direct associations refer to structural paths estimated between two constructs. Indirect associations refer to pathways mediated through teacher efficacy. Total associations represent the sum of direct and indirect associations. Bootstrap confidence intervals are 95% bias-corrected intervals based on 5,000 bootstrap samples; an indirect association was considered statistically significant when the interval did not include zero. All p-values are two-tailed.

Supplementary Table 1​: Alternative model comparisons for the structural equation models (N = 350). Note. χ² = chi-square statistic; df = degrees of freedom; χ²/df = chi-square divided by degrees of freedom; TLI = Tucker-Lewis Index; CFI = Comparative Fit Index; IFI = Incremental Fit Index; AGFI = Adjusted Goodness-of-Fit Index; RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square Residual; Δχ² = change in chi-square; Δdf = change in degrees of freedom; ΔCFI = change in Comparative Fit Index; ΔRMSEA = change in Root Mean Square Error of Approximation. The full-mediation model was compared with the partial-mediation model. Please click here to download this file.

Supplementary File 1: Raw data. These supplementary files are provided to verify the reported results and reproduce the statistical analyses.Please click here to download this file.

Discussion

This study reveals a distinct pattern in the relationship between professional perceptions and teacher-child interaction among pre-service early childhood educators. Professional perceptions were not directly associated with self-reported teacher-child interaction competence after accounting for teacher efficacy, but they showed a significant indirect association through teacher efficacy, supporting teacher efficacy as a full statistical mediator in the cross-sectional SEM model. This finding contrasts with prior studies on early childhood educators, which generally suggested a direct impact of professional perceptions on teacher-child interaction^28,29, with teacher efficacy serving only a partial mediating role³⁰. For example, previous studies found a significant positive correlation between professional perceptions and teacher-child interaction, with educators who held higher professional perceptions tending to adopt more positive interaction strategies, such as providing encouraging feedback and supportive language¹⁷. Teachers with high efficacy were able to translate their professional knowledge into effective guidance during play activities. These teachers not only responded promptly and accurately to children's needs during play but also enriched the forms and content of teacher-child interactions, ultimately improving interaction quality³⁰. However, this study found that professional perceptions alone did not directly affect teacher-child interaction. This suggests that a teacher’s recognition of their professional capabilities does not automatically translate into high-quality teacher-child interactions. One possible explanation is that teacher-child interaction is shaped by multiple factors. Even if pre-service educators strongly identify with their professional competence, they may struggle to navigate the complexities of real-world educational settings. This disconnect between professional perceptions and practical application may hinder the effective translation of professional perceptions into high-quality teacher-child interaction skills⁸. The gap between theory and practice, insufficient teacher beliefs, and low teacher efficacy might all contribute to discrepancies in educators' behavior and attitudes during actual interactions with children, preventing professional perceptions from manifesting as effective interaction^31,32. Beyond the theory-practice gap, the full mediation pattern may reflect the motivational nature of the pre-service stage: professional perceptions provide an understanding of what an early childhood teacher should value and know, but this understanding is unlikely to shape interaction behavior unless pre-service teachers also believe that they can manage children’s responses, organize activities, and adapt communication in real situations. Teacher efficacy may therefore function as a psychological conversion mechanism, transforming professional awareness into enacted confidence rather than merely serving as another personal trait. In this context, Reynolds et al.³³ emphasized the importance of mentorship and professional support during internships, noting that such support could positively influence the quality of teacher-child interactions in actual classroom settings.

Given that teacher-child interaction is shaped by complex factors, simply enhancing pre-service early childhood educators’ professional perceptions is insufficient to ensure high-quality interactions. The significance of this study lies in its focus on the specific group of pre-service early childhood educators, clearly revealing the full mediating role of teacher efficacy in translating professional perceptions into effective teacher-child interaction. Teacher efficacy should therefore be understood as an important, but not exclusive, mechanism linking pre-service educators’ professional perceptions with their reported interaction competence; mentoring quality, practicum intensity, institutional support, and prior classroom exposure may also shape whether professional understanding is enacted in interaction. Pre-service educators with higher teacher efficacy may be better positioned to narrow the gap between ideal and real teaching scenarios and to apply theoretical knowledge more flexibly in practice³⁰. Additionally, by adjusting their teaching attitudes and fostering educational innovation, these educators continuously enhance their professional competencies, which enables them to respond more effectively to the challenges and pressures encountered in real-world teacher-child interactions^34,35. Therefore, rather than focusing solely on professional perception, greater emphasis should be placed on strengthening teacher efficacy, as it is the key factor in promoting high-quality teacher-child interactions. In addition to deepening professional understanding, it is crucial to provide pre-service early childhood educators with diverse educational and practical platforms that help them translate theoretical knowledge into practical teaching skills. Attention should also be given to educators’ emotional needs, with psychological support and other mechanisms aimed at fostering a positive self-concept, improving emotional regulation, and enhancing stress-coping abilities. Through these measures, educators can develop greater confidence and professionalism in their interactions with children, ultimately providing strong support for children's development. This study also revealed that, although the direct effect of social support on teacher-child interaction among pre-service early childhood educators was not significant, social support had a significant indirect effect through teacher efficacy, which served as a full mediator. These findings may be reasonable because of early childhood educators^9,36.

However, we contend that the conclusion that social support does not directly enhance teacher-child interaction among pre-service early childhood educators is reasonable. Based on the present findings, we speculate that this outcome is influenced both by the nature of social support itself and the professional demands of teacher-child interaction skills. Social support generally refers to emotional, informational, or material assistance provided by family, friends, colleagues, or the community during times of stress or challenge¹². However, such support does not typically directly enhance teachers' performance in specific professional competencies, such as teacher-child interaction. Instead, social support helps create a supportive work environment, fosters a sense of belonging, facilitates integration into the team, stimulates interest in professional development, and reduces stress, which in turn leads to better teaching practices and interactions^23,37. In contrast, teacher-child interaction skills are complex, professional competencies that require continuous learning, refinement, and practical application. These skills involve knowledge from multiple disciplines, including pedagogy, psychology, and child development, as well as practical teaching experience and coping strategies^5,6. Thus, external social support alone is unlikely to directly improve such specialized and multifaceted skills. This interpretation is particularly relevant for pre-service teachers, because support from family, peers, or mentors may first alter their perceived competence and emotional readiness, whereas the observable quality of teacher-child interaction depends on whether such support is internalized as confidence in managing real classroom situations. Although social support did not directly affect the teacher-child interaction abilities of pre-service early childhood educators, teacher efficacy served as a bridge, enabling social support to indirectly influence these abilities. There is a consensus among experts that support from those around pre-service educators is crucial. This support not only helps them identify and address deficiencies in their teaching but also provides emotional comfort and encouragement. For novice teachers, mentorship—especially the informational and emotional support provided by mentors has been closely linked to increased teacher efficacy and reduced emotional fatigue^37,38. Through constructive feedback and guidance, mentors help novice teachers refine their teaching methods, fostering a positive mindset during interactions with children. Moreover, research has shown that teachers who engage in professional development training tend to have higher self-efficacy, as such opportunities provide them with additional strategies and tools for problem-solving⁷. Therefore, we hypothesize the following: pre-service early childhood educators who receive ample social support, particularly those who benefit from the thoughtful guidance of mentors, can enhance their professional competencies and teacher efficacy through mentors' professional evaluations, emotional encouragement, and informational and material support. Furthermore, pre-service educators with high teacher efficacy are more likely to adopt diverse communication strategies in teacher-child interactions, and, even when faced with challenges, they can rely on their high efficacy to ensure smooth interactions with children.

In conclusion, while social support did not directly enhance the teacher-child interaction abilities of pre-service early childhood educators, this does not imply that social support is insignificant in their professional development. On the contrary, through evaluative, emotional, material, and informational support, social support plays a crucial indirect role in enhancing pre-service educators' teacher-child interaction abilities. Therefore, in efforts to enhance the teacher-child interaction abilities of pre-service early childhood educators, it is essential to ensure they receive social support and, by strengthening teacher efficacy, indirectly promote substantial improvements in their interaction abilities. This study revealed that the influence of professional perceptions on teacher-child interactions was numerically greater than that of social support, with the difference in impact being approximately twofold. This result was somewhat unexpected. Upon further analysis, we speculate that this considerable disparity may be closely tied to the current model of teacher education. This disparity may be related not only to current teacher education models but also to the psychological distinction between internal professional agency and external contextual resources. Professional perceptions are tied to role identity, intrinsic motivation, and perceived responsibility for children’s development, and may therefore be more readily internalized into teacher efficacy⁷; by contrast, social support remains an external condition whose influence depends on its specificity, continuity, and relevance to concrete interaction problems. Additionally, with an increasing focus by teacher education institutions on professional ethics and social service awareness, the enthusiasm and commitment of pre-service early childhood educators toward teacher-child interaction have also been significantly enhanced³⁹. In contrast, while social support does exert some influence on teacher-child interactions, its impact appears to be relatively modest. This may be due to the limitations of social support in terms of its form, timeliness, and quality. Social support typically includes emotional, informational, material, and evaluative assistance, but not all forms of support directly influence teacher-child interactions. Some may have only an indirect effect^37,40,41. In terms of timeliness, teacher-child interactions require long-term and stable social support as a foundation. However, pre-service early childhood educators generally seek help from mentors, peers, or family members only when facing specific difficulties or challenges, thereby lacking the continuity of ongoing support⁴². More importantly, from a quality perspective, the social support perceived by pre-service early childhood educators is often inadequate. Some forms of support may be limited to superficial concern or encouragement, without effectively improving teaching practices³³. Nonetheless, social support should not be underestimated, as it plays a critical role in helping pre-service early childhood educators cope with the pressures and challenges inherent in teacher-child interactions and serves as an important external resource for professional development.

Grounded in socio-ecological systems theory, this study provides cross-sectional SEM evidence for a conceptual framework in which internal and external resources are associated with self-reported teacher-child interaction competence via teacher efficacy. In this cross-sectional SEM, professional perceptions and social support were associated with self-reported teacher-child interaction competence primarily through teacher efficacy; accordingly, the findings should be interpreted as evidence of statistical mediation rather than as evidence that these resources cause changes in interaction behavior. This pattern provides an empirical rationale for teacher-preparation programs to consider efficacy-building activities, while longitudinal and intervention-based studies remain necessary before stronger practical claims can be made.

Based on the relative strength of the paths, teacher-preparation programs may prioritize three intervention points: embedding structured reflection on professional identity and child-centered interaction cases into coursework; providing supervised microteaching, video-based feedback, and mentor-guided practicum tasks that strengthen teacher efficacy; and building stable peer-mentor support channels during internships, so that social support is linked to specific interaction problems rather than remaining general encouragement. Because professional perceptions showed larger total and indirect effects than social support^42,43, such interventions should first help pre-service teachers interpret their professional role and then convert this understanding into repeated, feedback-based interaction practice.

Moreover, pre-service early childhood education teachers often hold idealized and naïve views of teaching, and initially high or unrealistic efficacy beliefs may decline as pre-service teachers gain practical experience. Therefore, it is crucial to help pre-service teachers maintain a balanced mindset throughout their professional development³⁸. To achieve this, it is important to integrate courses that enhance and sustain teacher self-efficacy into the professional development plans for pre-service early childhood education teachers. Several limitations should be considered when interpreting the findings. First, the cross-sectional design does not permit causal inference, and longitudinal or intervention-based studies are needed to examine whether professional perceptions and social support lead to subsequent changes in teacher efficacy and interaction competence. Second, although the sample included four institutions in Fujian and Shaanxi, it remains a regional sample; future studies should include broader eastern, western, northern, and southern Chinese contexts to examine regional and cultural differences among pre-service early childhood teachers. Third, all core variables were measured through self-report, which may introduce perceptual bias and shared method variance, even though procedural controls and Harman’s single-factor test were used. Fourth, the teacher-child interaction scale was culturally reviewed and supported by CFA in this sample, but formal measurement invariance across Chinese and Korean samples was not tested. Fifth, the high correlation between teacher efficacy and self-reported teacher-child interaction competence, particularly the r = 0.897, suggests that future research should further distinguish efficacy beliefs from enacted interaction quality through classroom observations, mentor ratings, video-based coding, and other classroom-based assessments. Finally, variables such as mentoring quality, practicum intensity, institutional climate, and prior classroom exposure were not included, although they may also influence how pre-service teachers translate professional understanding into interaction practice. This study examined how professional perceptions, social support, and teacher efficacy are associated with teacher-child interaction competence among pre-service early childhood education teachers. In this regional cross-sectional sample, professional perceptions and social support were associated with self-reported teacher-child interaction competence primarily through teacher efficacy, supporting teacher efficacy as the central statistical mediator in the proposed model. The findings indicate that professional and contextual resources are linked to perceived interaction competence through pre-service teachers’ efficacy beliefs. These conclusions should be interpreted within the limits of the cross-sectional, regional, and self-report design.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Social Support Scale	Published academic scale; originally developed by Park and revised/supplemented by Kim	25-item version; 5-point Likert response format; no commercial catalog number	Measures perceived social support among pre-service early childhood teachers across Emotional Support, Informational Support, Material Support, and Evaluative Support. Overall Cronbach’s alpha = 0.96.
Teacher Self-Efficacy Scale	Published academic scale; originally developed by Enochs and Riggs; Chinese-adapted study version	25-item version; 5-point Likert response format; no commercial catalog number	Assesses teacher efficacy in the Chinese pre-service early childhood education context across General Teacher Self-Efficacy and Personal Teacher Self-Efficacy. Overall Cronbach’s alpha = 0.96.
Teacher-Child Interaction Scale	Published academic scale; developed by Lee and reviewed by the study expert panel for Chinese use	30-item version; 5-point Likert response format; no commercial catalog number	Assesses self-reported teacher-child interaction competence rather than observed classroom interaction quality. Includes Emotional Interaction, Verbal Interaction, and Behavioral Interaction; overall Cronbach’s alpha = 0.94.
Demographic Information Form	Research team, Sichuan University of Arts and Science / Daegu University collaboration	Custom study questionnaire section; no commercial catalog number	Collected participant characteristics including gender, educational background, professional category, internship location, internship duration, internship type, and kindergarten type.
Participant Information and Informed Consent Script	Research team; approved by Daegu University Research Ethics Committee	Ethics approval number: XXX; local approval record to verify	Used before questionnaire distribution to explain study purpose and survey content and obtain informed consent from pre-service early childhood education teacher participants.
Pilot Survey Questionnaire Package	Research team and expert panel	Pilot version used June 1–5, 2024; no commercial catalog number	Pilot-tested the clarity and appropriateness of the Professional Perceptions Scale, Social Support Scale, Teacher Self-Efficacy Scale, Teacher-Child Interaction Scale, and demographic items in 50 pre-service early childhood education teachers.
Expert Panel Review Procedure	Research team; expert panel of two early childhood education professors, one preschool director, three internship supervisors, and the research team	Consensus-based content review; no commercial model number	Used to revise scale items for ambiguity, appropriateness, semantic clarity, developmental suitability, and cultural relevance before the formal survey.
Printed On-Site Questionnaire Forms	Research team; local institutional survey supplies	Paper survey forms; no commercial catalog number	Used for on-site questionnaire administration. In the final sample, 336 valid on-site questionnaires were obtained before combining with online questionnaires.
WeChat	Tencent	WeChat mobile application; version should be verified from the device/app record used during June–July 2024	Used as an online distribution channel for questionnaire administration; 27 completed online questionnaires were collected through the online/WeChat route.
Telephone and In-Person Recruitment Procedure	Research team	Standardized recruitment procedure; no commercial catalog number	Used to contact institutions/participants, explain the study purpose and survey content, and support informed consent before questionnaire administration.
SPSS Statistics	IBM	Version 27.0	Used for descriptive statistics, Pearson correlation coefficients, Cronbach’s alpha reliability analysis, normality checks using skewness and kurtosis, and Harman’s single-factor test for common method bias.
SPSS Amos	IBM	Version 26.0	Used for confirmatory factor analysis, structural equation modeling, multi-group measurement invariance testing, bootstrap testing of indirect effects, and alternative model comparisons.
Microsoft Excel	Microsoft	Microsoft 365 Excel or locally installed Excel version; verify exact local build before submission	Recommended/used as the workbook environment for organizing coded survey data, data dictionary, supplementary tables, and the JoVE Table of Materials. Exact local version should be verified by the authors.
Structural Equation Modeling Fit-Index Criteria	Published methodological guidance; Schreiber et al. and Kline references cited in the manuscript	No commercial catalog number; criteria applied in the statistical analysis plan	Defines and reports model-fit indices including chi-square statistic, chi-square divided by degrees of freedom, Tucker-Lewis Index, Comparative Fit Index, Adjusted Goodness-of-Fit Index, Incremental Fit Index, Root Mean Square Error of Approximation, and Standardized Root Mean Square Residual.
Bootstrap Mediation Testing Procedure	SPSS Amos / research team statistical analysis plan	5,000 bootstrap samples; 95% bias-corrected confidence intervals	Used to test indirect associations through teacher efficacy. Indirect effects were considered significant when the 95% bias-corrected confidence interval did not include zero.
Measurement Invariance Testing Procedure	SPSS Amos / research team statistical analysis plan	Multi-group confirmatory factor analysis; ΔCFI less than .010 and ΔRMSEA less than .015 criteria	Used to evaluate configural, metric, and scalar invariance across gender, program type, internship duration, and institution before pooled SEM estimates were interpreted.