2. Methods
An initial sample of 373 middle school students and 63 teachers was recruited from public middle schools in various districts of Shanghai. Of these, 15 students were excluded from the data analysis because of incomplete responses (attrition rate = 4.0%), which is considered acceptable, as Rubin and Babbie (2016) suggest that attrition rates under 20% are typically acceptable in social science research. The final student sample consisted of 358 participants (53.8% female; Mage = 13.38 years, SD = 2.20, range = 12–15 years; educational years = 7.80 years, SD = 1.70, range = 7 to 9 years). The teacher sample included 63 participants (57.8% female; Mage = 35.42 years, SD = 6.80, range = 28–52 years), consisting primarily of homeroom teachers and subject teachers who directly interacted with the participating students. All participants were ethnically Chinese and primarily came from middle- to lower-middle-class socioeconomic backgrounds. Participation was entirely voluntary, this classification was based on the socioeconomic characteristics of the school districts where participants were recruited, which are predominantly attended by families within these income brackets. Prior to data collection, the participating schools, teachers, students, and their parents were fully informed about the study’s objectives, procedures, confidentiality, and anonymity guarantees. Students, their parents, teachers, and schools provided informed consent to participate in the study. Data collection was conducted in group settings, with approximately 30–40 student participants per session, either in classrooms or conference rooms at the schools. The assessments for students included the Growth Mindset Inventory (GMI), the Perceived School Climate Scale (PSCS), and the Parental Autonomy Support Questionnaire (PASQ). Teachers completed the Growth Mindset Inventory (GMI) to assess their beliefs about intelligence and learning. Each questionnaire took approximately 5–10 min to complete, with the total session lasting around 25–30 min. Standardized instructions were provided to ensure consistency across sessions, with research assistants onsite to facilitate the process. The entire procedure was designed to ensure a calm and focused environment conducive to participant engagement. Additionally, Students’ academic achievement was assessed using district-level final exam scores on a 100-point scale. These scores were sourced directly from official transcripts provided by the participating schools, ensuring both the accuracy and completeness of the data. Teachers’ growth mindset data were collected using the GMI, and their responses were linked to the students they taught to explore the moderating effects of teachers’ growth mindset on the relationship between students’ growth mindset and academic performance. The research team collected all data after written consent was obtained from both the schools and the participants.
2.2. Instruments
2.2.1. Growth Mindset Scale
Both participating students’ and teachers’ growth mindsets were measured via the Chinese-adapted version of the growth mindset subscale, the Growth Mindset Inventory (GMI; Dweck 1999), which assesses individuals’ mindsets about the malleability of intelligence. This 4-item inventory uses a Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree) and includes statements such as “Even your basic intelligence level can be increased considerably.” A higher score indicates a greater degree of endorsement of the idea that intelligence is malleable. The GMI effectively mitigates the positive wording effect highlighted in recent literature (Yu and Kreijkes 2017), which can lead individuals to identify as incremental theorists due to social desirability bias, especially those familiar with the growth mindset concept included in the questionnaire (Song 2018). The GMI has been shown to be reliable in both Chinese and international contexts, with values of Cronbach’s α ranging from 0.76 to 0.83 (Wang et al. 2021; Zhang and He 2024). In this study, the GMI demonstrated similarly acceptable reliability, with Cronbach’s α = 0.82.
2.2.2. Perceived School Climate Support Scale
The Chinese-adapted version of the Delaware School Climate Survey for Students (DSC-S; Bear et al. 2011) was employed in this study to assess participants’ perceptions of school climate support. The DSC-S consists of 23 items designed to capture five core dimensions of school climate: (1) teacher–student relations (e.g., teachers care about their students), (2) student–student relations (e.g., students treat each other with respect), (3) fairness of rules (e.g., the school’s Code of Conduct is fair), (4) school safety (e.g., I feel safe in this school), and (5) liking of the school (e.g., I am proud of my school). The participants responded to these items on a four-point Likert scale ranging from “strongly disagree” (1) to “strongly agree” (4), which enabled a detailed analysis of their views on various aspects of the school climate. In a recent study with Serbian high school students, Đorđić et al. (2022) reported a Cronbach’s alpha of 0.74 for the DSCS-S, indicating acceptable internal consistency. Similarly, in the present study, the DSCS-S demonstrated acceptable reliability, with a Cronbach’s α of 0.85.
2.2.3. Perceived Parental Autonomy Support Scale
The Perceived Parental Autonomy Support Scale (P-PASS; Mageau et al. 2015) was used to assess participants’ perceptions of parental autonomy support. The scale consists of 24 items divided into six subdimensions: (1) choice within certain limits is measured by items such as “My parents encouraged me to make choices in line with my values”; (2) rationale for demands and limits is captured by statements such as “My parents explained the reasons behind their decisions and limits”; (3) for acknowledgment of feelings, an example item is “My parents were able to put themselves in my shoes and understand my feelings”; (4) threats of punishment are assessed with items such as “As soon as I didn’t do exactly what my parents wanted, they threatened to punish me”; (5) performance pressures are reflected by items such as “My parents insisted that I always be better than others”; and (6) guilt-inducing criticism is measured with statements such as “When my parents wanted me to act differently, they made me feel ashamed in order to make me change.” Participants rated each item on a 7-point Likert scale ranging from “Do not agree at all” (1) to “Very strongly agree” (7). A recent study by Inda Caro et al. (2022) supported the reliability of the P-PASS, obtaining Cronbach’s alpha values between 0.89 and 0.94. In the present study, the P-PASS demonstrated acceptable reliability, with a Cronbach’s alpha of 0.88.
2.2.4. Data Analysis
SPSS software (version 26.0) was used for the statistical analysis in this study. Before hypothesis testing, a Pearson correlation analysis was conducted to identify the bivariate correlations among the main study variables (i.e., students’ growth mindset, teachers’ growth mindset, perceived school climate support, perceived parental autonomy support, and academic achievement). The interpretation of the correlation coefficients followed Cohen’s (1988) guidelines, which define r ≥ 0.10 (or ≤−0.10) as indicating a small effect, r ≥ 0.30 (or ≤−0.30) as indicating a medium effect, and r ≥ 0.50 (or ≤−0.50) as indicating a large effect. In addition to the main variables, control variables such as age and sex were included to assess their potential influence on the results, ensuring a thorough examination of the relationships among the study variables.
To test the hypotheses with respect to the moderating effects of teacher mindset (H1), perceived school climate support (H2), and perceived parental autonomy support on the growth mindset−academic achievement relationship (H3) among students, hierarchical regression analyses were conducted. To test each of these three hypotheses, three steps were involved in the analyses. In Step 1 (Model 1), demographic variables such as age, gender, and education level were included to control for their potential influence on academic achievement. In Step 2 (Model 2), the direct effect of students’ growth mindset on academic achievement was examined. In Step 3 (Model 3), interaction terms between students’ growth mindset and each of the three contextual factors (i.e., teachers’ growth mindset, perceived school climate support, and perceived parental autonomy support) were added to assess their moderating effects on the relationship between growth mindset and academic achievement. Prior to running the regression, key assumptions were evaluated to ensure the reliability of the analyses. The normality of the residuals was assessed via skewness and kurtosis values, both of which were within acceptable ranges (skewness: −0.24 to 0.89; kurtosis: −0.31 to 1.10), confirming that there were no serious deviations from normality. Homoscedasticity was tested via the Breusch–Pagan test, which yielded nonsignificant results (p = 0.36), confirming that the variance of the residuals was consistent across the levels of the independent variables. Multicollinearity was assessed via variance inflation factor (VIF) statistics; all VIF values were below 2.0, ranging from 1.05 to 1.32, indicating no multicollinearity issues (Johnston et al. 2018). In addition to the regression analyses, simple slope plots were generated to visually represent the moderation effects, illustrating how academic achievement varied under high (M + 1SD) and low (M − 1SD) levels of the moderating variables. These plots followed standard practices in similar research, providing a clear visual representation of the moderation effects. All analyses were conducted with a significance threshold of p < 0.05.
CC BY 4.0. "Teachers' Growth Mindset, Perceived School Climate, and Perceived Parental Autonomy Support Moderate the Relationship Between Students' Growth Mindset and Academic Achievement" by Zhang K, He WJ, Journal of Intelligence (2025). https://doi.org/10.3390/jintelligence13010008. Licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Peer-reviewed and open access; full text from PubMed Central (PMC11766234). Headings and layout adapted for reading; figures and tables are in the original.