Mindset and achievement in general chemistry: insights from Turkish undergraduate students

Abstract

Chemistry-specific mindsets have been shown to influence student engagement, self-efficacy, goal orientation, and academic achievement in chemistry. These beliefs are increasingly recognized as context-dependent, shaped through interactions within specific social and educational environments. The present study investigated the extent to which Turkish undergraduates have a growth chemistry mindset and how these beliefs relate to academic performance in chemistry across engineering and natural sciences majors. A total of 817 second-semester general chemistry students participated, completing surveys after the second midterm and prior to the final exam. The Chemistry Mindset Instrument (CheMI; Santos D. L., Barbera J. and Mooring S. R., (2022), Chem. Educ. Res. Pract., 23(3), 742–757) was used to assess students’ mindsets, while grade records provided measures of exam performance. Confirmatory factor analysis (CFA) supported the unidimensional structure of the CheMI, consistent with previous findings, providing evidence for its validity in this cultural and educational context. Structural equation modeling (SEM) examined whether chemistry mindset mediated the relationship between midterm and final exam performance across majors. Additional regression analyses explored how mindset influenced the midterm–final performance relationship among students with low, medium, and high levels of achievement. SEM results indicated that second midterm performance did not significantly predict mindset for either major. However, mindset significantly predicted final exam performance for engineering students, whereas this relationship was nonsignificant for natural sciences students. Furthermore, mindset emerged as a significant predictor only among the highest-performing students, suggesting that it may play a differentiating role at the upper end of achievement. These findings underscore the importance of considering mindset in instructional design and provide insights for targeted strategies to enhance student success in chemistry.

---

Introduction

The investigation into why individuals in identical circumstances pursue varying goals has prompted researchers to develop more comprehensive implicit theories. The implicit theories have provided a clear portrayal for adaptive functioning through identification of a missing piece in the patterns of affect-behavior: cognition (Elliott and Dweck, 1988). How individuals implicitly conceptualize the nature of their ability, in other words their theory of intelligence, has been found to be an important predictor of goal orientation (Nicholls, 1984). Individuals with an incremental theory about their intelligence believe that intelligence is fixed, whereas for those holding an entity theory, intelligence is perceived as malleable (Dweck and Leggett, 1988; Dweck et al., 1995).

The concept of mindset has increasingly been employed in the literature as a preferred term instead of theories of intelligence to describe individuals’ beliefs regarding the extent to which intelligence is malleable. Consequently, the terms growth mindset and fixed mindset have become more prevalent in recent literature, while remaining conceptually grounded in the original frameworks of the incremental and entity theories of intelligence, respectively (Lüftenegger and Chen, 2017). Since mindset promotes persistence among students in science majors (Molden and Dweck, 2006), it has potential to help solve an important issue, the high withdrawal rate in introductory chemistry compared to other courses (McKinney et al., 2019). Accordingly, researchers in chemistry education have directed their attention to mindset beliefs. Their efforts have shown that a growth mindset is positively associated with higher academic performance in chemistry courses (Fink et al., 2018; Wang et al., 2021; Santos et al., 2022), greater chemistry self-efficacy (Santos et al., 2022; Naibert et al., 2024; Pulukuri et al., 2025), and a stronger inclination to adopt mastery-approach goals (Santos et al., 2022; Naibert et al., 2024), while being linked to a weaker preference for mastery-avoidance (Santos et al., 2022) and performance-avoidance goals (Naibert et al., 2024). In addition, growth mindset has been shown to indirectly influence key student outcomes by enhancing self-efficacy and shaping achievement goal orientations. For example, it predicts higher exam performance through increased self-efficacy (Pulukuri et al., 2025) and impacts both summative and formative assessments via goal orientation pathways (Naibert et al., 2024). Additionally, perceptions of students held by instructors with a fixed mindset have been shown to be indirectly related to lower course grades through increasing students' sense of academic misfit (Kattoum et al., 2024).

Researchers have acknowledged that an individual might hold both incremental and entity theories of intelligence; while one theory might be more central but either theory might be accessible under specific circumstances (Dweck and Leggett, 1988; Dweck et al., 1995). That is, mindset beliefs are context-dependent (Little et al., 2016; Dweck, 2017; Limeri et al., 2020), which aligns with their emergence in concrete situations through interaction with the (social) context (de Ruiter and Thomaes, 2023). Current literature on the chemistry mindset at the undergraduate level has been dominated by research in US universities. To date, to the best of our knowledge, there has been no research on the degree to which undergraduate students in Turkey believe that intelligence is malleable and how their mindset predicts their academic performance in chemistry. Therefore, we aim to contribute to understanding the circumstances that stimulate different theories to provide more insight into implicit theories of intelligence (Santos and Mooring, 2022) through investigation of students’ mindset and its relationship to performance in a culturally different context. In addition, we seek to provide validity and reliability evidence associated with the Chemistry Mindset Instrument (CheMI, Santos et al., 2022) in this new context. The potential to detect individual differences across cultures is important to corroborate the generalizability and validity of a personality construct (Heine and Buchtel, 2009).

Theoretical framework: mindset

Individuals endorsing a fixed mindset perceive intelligence as an inherent, stable trait that cannot be altered through effort. In contrast, those who hold a growth mindset believe that intelligence can be developed over time through dedication and hard work (Dweck and Leggett, 1988; Dweck et al., 1995). These beliefs are closely associated with distinct goal orientations. Specifically, individuals with a fixed mindset tend to adopt performance-oriented goals, wherein achievement is evaluated based on external standards, such as grades. Conversely, individuals with a growth mindset are more likely to pursue mastery-oriented goals, which emphasize personal improvement and learning based on intrapersonal criteria (Dweck and Leggett, 1988; Dupeyrat and Mariné, 2005; Dinger and Dickhäuser, 2013). Moreover, mindset beliefs shape individuals’ emotional and behavioral responses to academic challenges. Those with a fixed mindset may interpret challenges as indicative of a lack of ability and a threat to self-esteem, often resulting in heightened anxiety (Diener and Dweck, 1980; Dweck et al., 1995). In contrast, individuals with a growth mindset are more likely to view challenges as opportunities for development, which can foster increased effort, greater satisfaction, and positive emotions such as pride and enjoyment (Diener and Dweck, 1978; Dweck and Leggett, 1988; Dweck et al., 1995). These differing interpretations of experience, shaped by underlying mindset beliefs, ultimately lead to divergent behavioral patterns. Individuals with a fixed mindset often avoid challenges, exhibit low persistence, procrastinate, and evade self-evaluation. In contrast, those with a growth mindset demonstrate higher levels of persistence, actively seek out challenging tasks, and embrace feedback as a means to enhance learning (Dweck and Leggett, 1988; Molden and Dweck, 2006; Burnette et al., 2013).

Mindset beliefs can be changed since they are relatively malleable (Dweck et al., 1995). Parents’ judgmental stance and praise for intelligence could stimulate the development of fixed mindset beliefs whereas growth mindset could be strengthened through praising and emphasizing effort and strategies (Dweck, et al., 1995; Mueller and Dweck, 1998). Interventions focusing on incremental theories of intelligence (Fink et al., 2018; Wang et al., 2021; Khandelwal et al., 2024; Nguyen, et al., 2024) have been also found to be effective in reinforcing growth mindset beliefs, which in turn increase the accessibility of these beliefs for individuals.

Sensitivity of mindset beliefs to context was raised by Dweck and colleagues (Dweck and Leggett, 1988; Dweck et al., 1995; Dweck, 2017) although they did not explicate the specific cultural factors that influence the mindset belief nor the degree of sensitivity to those factors. This work advocated that individuals may hold both incremental and entity theories of intelligence, with situational factors determining which becomes salient (Dweck and Leggett, 1988; Dweck et al., 1995). Subsequent work delved into the complexities of the context dependency of mindset (Little et al., 2016; Limeri et al., 2020; Kroeper et al., 2022; de Ruiter and Thomaes, 2023; Muradoglu et al., 2023; Asbury et al., 2025). To construct the foundation for context-dependent beliefs, interviews were used instead of domain general surveys to elicit students’ mindset beliefs (Little et al., 2016). Analysis of interview data indicated that prompts involving family storytelling frequently elicited fixed mindset responses, while questions focused on improvement and experiences of pride in physics classes more commonly evoked growth mindset responses. Benefiting from interviews to examine the factors that influence students’ mindset beliefs, Limeri et al. (2020) identified several contextual elements in educational settings where students describe differing effects on mindset: experiences with success or failure, observation of peers, and the structure of the school system (e.g., placement in classes based on intelligence). Narrowing the focus to instructors’ teaching behaviors and messages, Kroeper et al. (2022) found that explicit messages about progress and success, opportunities for practice and feedback, instructors’ responses to poor performance, and the value they place on student learning and development play crucial roles in shaping students’ mindsets within the learning environment. Complementing this evidence, longitudinal work on university students demonstrates that mindset beliefs evolve systematically during university education, depending on the subject students are studying, with strong subject-specific differences (Asbury et al., 2025). In addition to the instructional context, the cultural context—particularly cultural beliefs about success and stereotypes about intelligence—has been shown to play a crucial role in shaping mindset beliefs (Muradoglu et al., 2023). Drawing on the mindset literature, de Ruiter and Thomaes (2023) proposed a comprehensive theoretical model that explains the emergence and development of mindsets through socially situated processes. The model (i.e., the process model of mindsets) posits that a mindset comprises interacting facets—beliefs, action tendencies, and situational responses—which are actively constituted through bidirectional sensemaking that occurs during social interaction. Conceptual principles like emergence and constraint suggest that mindsets are not fixed, but rather probabilistic and self-sustaining systems that develop iteratively over time, thereby accounting for the development of multiple, contextually enacted mindsets within an individual (de Ruiter and Thomaes, 2023). Consequently, mindset beliefs are context-dependent, arising through interactions with specific social and instructional contexts.

The role of mindset in students’ learning process in chemistry

In recent years, chemistry education researchers have increasingly focused on mindset beliefs, recognizing their context-dependent nature (Little et al., 2016; Limeri et al., 2020; Kroeper et al., 2022; de Ruiter and Thomaes, 2023; Muradoglu et al., 2023; Asbury et al., 2025) and their role in supporting persistence within science majors (Molden and Dweck, 2006). While researchers’ aims have varied slightly, their collective efforts have significantly contributed to clarifying and measuring the construct, as well as identifying how chemistry-specific mindset beliefs influence a range of variables shaping the learning process.

As a first attempt to validate a chemistry specific mindset construct, Santos et al. (2021) investigated how undergraduate chemistry students understand the terminology (e.g., intelligence, chemistry intelligence, and chemistry ability) and the degree to which the mindset construct is valid within chemistry. Students referred to academic and lab contexts where chemistry intelligence is applicable, which provided evidence for domain specificity of chemistry intelligence. Participants categorized as growth mindset described intelligence as malleable, which requires effort, while this was not the case for fixed mindset students. The authors underscore the necessity for further research involving non-US student populations, which describes the participants of this study, to ascertain how cultural differences may influence the applicability and generalizability of the chemistry mindset as a construct. Validation of the chemistry-specific mindset construct (Santos et al., 2021) informed the development of the Chemistry Mindset Instrument (CheMI), a unidimensional 7-item instrument scored on an 11-point semantic differential scale for use with introductory undergraduate chemistry students (Santos et al., 2022). Santos et al. (2022) emphasized that CheMI is ideal for studies on the pathways involved in student success in introductory college courses considering its length and simplicity, which aligns with the research question of this study. Building on those works, Santos and Mooring (2024) delved into the complexities of chemistry mindset belief through a multiple case study, which revealed the continual and multidimensional nature of the construct. Instead of categorizing students as fixed and growth, qualitative data provided evidence for a continuum of student perspectives reflecting varying degrees of mindset beliefs regarding chemistry learning, ranging from fixed to growth-oriented views. At one end of the continuum, there are students representing fixed or talent-dependent mindsets, perceiving chemistry ability as largely innate and improvement as limited or conditional on natural aptitude. Following this, there are students in the continuum towards the growth end holding more nuanced views that recognize the potential for development while still acknowledging the role of natural ability in facilitating learning. Students closer to the growth end illustrate growth-oriented beliefs moderated by personal confidence, with their effort and mindset shaped by past academic experiences. At the growth-dominant end of the continuum, students exemplify a belief in universal potential for development, emphasizing the importance of personal interest, sustained effort, and the motivational influence of educators. Regarding the multidimensional nature, a useful distinction in students’ mindset beliefs involves separating views about their own chemistry intelligence from beliefs about others. While these perspectives may align, they can also diverge significantly. A two-dimensional framework illustrates this: students may hold a growth mindset for both self and others, doubt their own potential while believing in others' growth, believe in their own potential more than others’, or endorse a fixed mindset for both. This distinction highlights the nuanced and multidimensional nature of mindset beliefs in the context of chemistry learning. However, CheMI's length, simplicity, and established findings continue to make it a viable choice for research.

Growth mindset interventions share several commonalities in their overarching approach; however, they vary in terms of the specific activities employed and the frequency of the activities during implementation. A consistent feature across these interventions is the incorporation of reflective assignments (Fink et al., 2018; Wang et al., 2021; Khandelwal et al., 2024; Nguyen et al., 2024), although the number and structure of these assignments differ among studies. The earliest implementation of a growth mindset intervention in chemistry education was integrated into three homework assignments in a study by Fink et al. (2018). The first assignment involved reading an article that described the brain as malleable and emphasized the role of effort in the development of ability, particularly in response to academic challenges. In the second and third assignments, students reflected on how the concepts presented in the article influenced their study strategies for General Chemistry I examinations. Expanding on this approach, Wang et al. (2021) revised the third assignment, asking students to write advice for future General Chemistry II students, encouraging them to adopt a growth mindset perspective. Findings from these studies indicated that students who received the growth mindset intervention outperformed the students in the control group in both General Chemistry I (Fink et al., 2018) and General Chemistry II (Wang et al., 2021). Furthermore, Fink et al. (2018) reported that one subgroup of students in the intervention group achieved significantly higher final exam scores than those in the control group, while no statistically significant effects were observed among another subgroup of students. In addition, growth mindset intervention was effective in increasing students’ utility value dimension of their chemistry attitude (Wang et al., 2021). These early efforts emphasized the importance of evaluating the local educational context prior to implementing a mindset intervention (Fink et al., 2018), cautioning against treating it as a universal ‘magic bullet’ (Yeager and Walton, 2011). This underscores the significance of the present study, as it represents the first attempt to examine university students’ mindsets in relation to their academic performance and to inform the design of potential interventions in the Turkish context. More importantly, Fink et al. (2018) highlighted the need for context-specific interventions. In response, subsequent studies incorporated more diverse activities and adopted a more explicit approach to embedding growth mindset principles within the specific domains under investigation (Khandelwal et al., 2024; Nguyen et al., 2024). One intervention included two modules promoting growth mindset development early in the semester before students faced difficulties in General Chemistry I and II (Nguyen, et al., 2024). The first module included topics such as neuroplasticity, factors influencing neuroplasticity, and distinctions between growth and fixed mindsets delivered via videos. The second focused on strategies for fostering a growth mindset in the context of general chemistry where students watched videos of interviews featuring a diverse group from the Department of Chemistry and Biochemistry, who shared personal experiences of overcoming academic challenges. Participation in the modules was associated with a shift toward growth mindset beliefs among General Chemistry II students, though no significant mindset change was observed in General Chemistry I students. In addition, General Chemistry I students who engaged in the modules scored higher on the ACS exam compared to non-participants. Both groups demonstrated increased chemistry self-efficacy by the end of the semester. Qualitative responses revealed improved attitudes toward failure and academic challenges. The vast majority of the students found the growth mindset interventions valuable, reported more positive perspectives on challenges, and believed they could master difficult content through effort and persistence (Nguyen, et al., 2024). The other intervention included a series of activities that were implemented to foster growth mindset development (Khandelwal et al., 2024). In a learning reflection homework, students reviewed their current grade and created an improvement plan with strategies for support from peers or the instructor. A brief mid-semester check-in encouraged self-compassion and mindset awareness, particularly in the context of pandemic-related stress. On Exam 1, a short reflective question prompted students to describe an academic challenge and their mindset in addressing it. During the First Day Fears activity, students discussed their anxieties in breakout groups and learned how to reframe them using a growth mindset. A final reflection on the last day of class invited students to articulate insights about themselves as learners, collaborators, and individuals developing a growth mindset (Khandelwal et al., 2024). Those various activities were found to be helpful in developing a stronger growth mindset by the end. Among all the activities, the learning reflection homework, mid-semester check-in, Exam 1 question on academic challenge, and First Day Fears were rated the most helpful by students, which was interpreted as an indication of a need for early interventions on growth mindset.

Considering the association between mindset, student interpretations of challenges, and their behavioral response to challenges, chemistry educators have directed their attention to the role of these relationships in student academic performance. In their study, Limeri et al. (2020) categorized students into three groups based on their experience with academic challenges: no challenge, overcame struggle, and continued struggle. Regarding a change in their mindset beliefs, there was an overall shift from growth to fixed mindset over the semester, with the continued struggle group starting with more fixed beliefs and exhibiting the largest decline in growth mindset. Students primarily attributed changes in mindset to academic experiences. The authors proposed a reciprocal model in which academic experiences influence mindset beliefs, which in turn affect academic outcomes—suggesting this model warrants further investigation despite acknowledged study limitations. Taking a closer look at the type of challenges, Santos and Mooring (2022) identified general and organic chemistry students’ major challenge types, which are time management, motivation, course content, lack of understanding, lack of resources, outside circumstances, teaching strategies, weak foundation, and chemistry ability. Course content, time management, motivation, and chemistry ability were the most frequent challenges. Additionally, the no challenge group not only had more growth mindset beliefs than other groups but also their formative and summative performances were higher. Regarding the challenge type, it was found that students who identified chemistry ability as their primary challenge not only had lower mindset scores indicating a more fixed mindset, but also underperformed on summative and formative measures of performance. Seeing potential in investigating whether students perceive distinct challenges in concert with others, Demirdöğen and Lewis (2023) provided evidence for the existence of three fundamental challenges: individual challenges, teaching-related challenges, and chemistry-related challenges. More importantly, students reporting a fixed mindset rated chemistry-related challenges higher, and students rating chemistry-related challenges higher scored lower on summative performance. Further, students reporting a growth mindset and earning a grade of C+ or lower rated all three fundamental challenges relatively high (Demirdöğen and Lewis, 2023).

The validation of mindset as a construct and its critical role in both academic performance and responses to challenge have prompted chemistry education researchers to examine its simple and complex relationship with other affective variables (Santos et al., 2022; Kattoum et al., 2024; Naibert et al., 2024; Pulukuri et al., 2025; Sun et al., 2025). Individual measures indicated that students reporting higher self-efficacy in their chemistry courses were more likely to believe that they can improve aspects of their chemistry intelligence (Santos et al., 2022; Naibert et al., 2024; Pulukuri et al., 2025). Moreover, growth mindset exerted a positive indirect effect on exam scores through self-efficacy as a mediator (Pulukuri et al., 2025). Similarly, students endorsing a growth mindset were more likely to adopt mastery-approach goals (Santos et al., 2022; Naibert et al., 2024) and less likely to endorse mastery-avoidance (Santos et al., 2022) or performance-avoidance goals (Naibert et al., 2024). Moreover, the influence of mindset on performance-avoidance orientation was moderated by self-efficacy, such that mindset exerted a stronger effect among students with higher self-efficacy and a weaker effect among those with lower self-efficacy. For instance, students with a growth mindset but low self-efficacy were less negatively aligned toward performance-avoidance goals compared to their high self-efficacy counterparts (Naibert et al., 2024). An indirect effect mediated through performance-avoidance was also observed. Specifically, performance goal orientation and mastery-avoidance orientation were significant partial mediators of the relationship between mindset, self-efficacy, and summative scores, whereas both mastery-approach and mastery-avoidance orientations fully mediated the relationship between self-efficacy and formative scores (Naibert et al., 2024). A recent study at the high school level revealed that students who perceived chemistry ability as malleable were more actively engaged in chemistry-related tasks. This relationship was statistically mediated by both academic adaptability and academic buoyancy in chemistry. Additionally, chemistry enjoyment and anxiety served as emotional mediators: a growth mindset was linked to increased enjoyment and reduced anxiety, which were, in turn, associated with higher and lower levels of academic engagement, respectively. Taking a different perspective, Kattoum et al. (2024) examined students’ perceptions of their instructor's mindset and found only indirect effects on course grades when controlling for students’ own mindset and prior preparation. Perceiving an instructor as having a fixed mindset was associated with increased feelings of academic misfit, which in turn predicted lower course grades. Conversely, perceiving an instructor as having a growth mindset was linked to greater utility value of the course, although this did not translate into higher grades. Additionally, students’ perceptions of the instructor's mindset did not predict self-efficacy, nor did self-efficacy predict course grades.

Research goals

Recent research suggests that growth mindset beliefs are often domain-specific, meaning that individuals’ mindset beliefs related to particular academic disciplines (e.g., biology and chemistry) may differ from general growth mindset beliefs. Chemistry specific mindsets have been found to predict student engagement, self-efficacy, goal orientation, and achievement within chemistry (Santos et al., 2022; Kattoum et al., 2024; Naibert et al., 2024; Pulukuri et al., 2025; Sun et al., 2025). Mindset beliefs are increasingly recognized as context-dependent, emerging through interactions within specific social and educational environments (Little et al., 2016; Limeri et al., 2020; Kroeper et al., 2022; de Ruiter and Thomaes, 2023; Muradoglu et al., 2023; Asbury et al., 2025). Despite growing interest in domain-specific mindsets, research on chemistry-related mindset beliefs at the undergraduate level has largely been confined to U.S. settings. To date, to the best of our knowledge, no studies have examined the extent to which Turkish undergraduates view intelligence as malleable or how such beliefs relate to their academic performance in chemistry. This study addresses this gap by exploring the relationship between students’ chemistry mindset and academic performance in a culturally distinct context, contributing to the broader understanding of implicit theories of intelligence (Santos and Mooring, 2022). Additionally, we provide validity and reliability evidence associated with the use of the Chemistry Mindset Instrument (CheMI; Santos et al., 2022) in a new cultural setting, offering insights into cross-cultural differences and supporting the generalizability of personality constructs (Heine and Buchtel, 2009).

These research goals have led to the following research questions that will guide this study:

1. To what extent does the intended factor structure of the Chemistry Mindset Instrument (CheMI; Santos et al., 2022) replicate in Turkey?

2. To what extent does chemistry mindset mediate the relationship between students’ midterm exam performance and their final academic achievement in chemistry across engineering and natural sciences majors in Turkey?

3. To what extent does chemistry mindset mediate the relationship between midterm and final exam performance for subgroups of students with low, medium, and high levels of midterm achievement, respectively?

Methodology

Research setting

The research setting was a large research-intensive university in Turkey. This university was selected for two reasons. First, the administered survey was in English; therefore, we considered universities in which the language of instruction was English. Second, three general chemistry courses (CHEM A, CHEM B, and CHEM C) were offered to students from different majors. This approach permitted the examination of a larger cohort of students and facilitated the analysis of any differences between the courses. Specifically, CHEM B and CHEM C were designed for engineering students, whereas CHEM A was offered to students from the College of Arts and Sciences and College of Education. While CHEM A and CHEM C are the General Chemistry II courses that follow the completion of General Chemistry I, CHEM B is a one-semester course that encompasses many topics typically covered in General Chemistry II. All three courses include a laboratory component. The common topics covered in all three courses are solutions, chemical kinetics, chemical equilibrium, chemical thermodynamics and electrochemistry. In addition to these, additional topics (i.e., atomic and electronic structure, chemical bonding, molecular structure and bonding theories, properties of liquids, solids) are covered in CHEM B. While each course is delivered in individual classes of 45–80 students by different instructors, they are coordinated through standardized syllabi, assignments, and examinations, which are administered simultaneously across all classes. In Turkey, high school seniors participate in a nation-wide university entrance exam that includes questions from subjects like mathematics, chemistry, physics, the Turkish language, and history, among others. Their placement at a particular university and a specific degree program within that university are determined by their performance on this exam combined with their high school cumulative grade point average (CGPA). The university attended by the participants is ranked among the top institutions in the country.

Participants

A total of 871 (see Appendix for the number of students across different majors) university students taking the general chemistry course participated in the study. Enrollment numbers indicated that CHEM A (N = 238) was taken primarily by Physics (117 students) and Biology (77 students) majors. CHEM B (N = 326) was dominated by engineering students, with Mechanical Engineering (N = 162) and Civil Engineering (N = 99) comprising the largest groups. CHEM C (N = 307) included substantial representation from Metallurgical and Materials Engineering (N = 81), Petroleum and Natural Gas Engineering (N = 67), Molecular Biology and Genetics (N = 58), and Geological Engineering (N = 48). These results demonstrate that each chemistry course predominantly serves distinct disciplinary cohorts. Ethical approval for this study was obtained from the Institutional Human Research Ethics Board (Protocol # 568 in 2024). Prior to participation, students were informed about the purpose of the study and the voluntary nature of their involvement, and written informed consent was obtained.

Data collection procedure

The survey was administered on paper to 17 classes (N = 871) after the second midterm and before the final (Fig. 1) during the spring 2024 semester so that students would have sufficient experience to report on their mindset beliefs in this context. The survey was administered in the first 15–20 minutes of a laboratory period in each course. Through this data collection timeline, we were able to build and test reciprocal relationships between academic performance and mindset to explain students’ exam performance hypothesized in the literature (Limeri et al., 2020). A survey including two parts was used to collect data. The first part of the survey was on student mindset while the second part included questions on demographic information.

Fig. 1 Timeline for data collection and grading for measures of summative performance.

The first author conducted cognitive interviews with three students (2 female – Sarah and Emma, 1 male – Michael; all names are pseudonyms) from the research setting prior to finalizing the instrument since items in the instrument were English. Our aim was to determine whether the respondents understood the items in the way intended by the instrument developers. In this way, we attempted to establish response process validation which refers to evaluating how well the construct and the nature of responses fit each other (AERA, APA, and NCME, 2014). The convenience sample of students was in their final year of either the Chemistry or Chemistry Education program. At the time of the interview, they had taken similar chemistry courses such as General Chemistry, Analytical Chemistry, and Organic Chemistry, but they did not participate in the study later. Cognitive interviews were conducted at the end of February 2024. During the interviews, students were asked to read the instruction in the survey at first to check the degree of clarity for guidance while responding. Following the review of instructions, students were prompted to read each item and to provide a response (i.e., a rating). The researcher then requested the interviewees to think-aloud on their response process thereby illuminating whether the item was interpreted as designed for validity evidence. The interviews were audio-recorded and transcribed verbatim.

To assess response-processes in this new context, two of the researchers analyzed transcripts of the three cognitive interviews individually. They particularly examined how the students generated their responses, whether their explanation of each item matched with their rating, what difficulties they experienced while responding, and whether their interpretations were similar to those of the developers. Then, they came together, compared their findings, and reached a consensus. As a result of the analysis, one item emerged as ambiguous (item 4: my ability to master chemistry content is something …). All three students expressed similar difficulties with the item. They could not fully grasp the use of the word “master,” as they usually use this word to refer to further studies after graduation, such as a master's degree, as indicated in the following excerpt from Sarah's interview: “I couldn’t understand it…To me, it sounds like advanced chemistry… It is like doing a master's degree, that's the only meaning I know” (translated from Turkish). Accordingly, to clarify the item for administration in this study, an explanation in parenthesis was added. The final item wording is as follows: My ability to master (fully comprehend) chemistry content is something …

The students appeared to comprehend the remaining items as designed. For instance, Emma explained item 5 (My ability to visualize chemical structures and processes is something…) as

“The shapes formed by compounds—that is, molecules and how they are processed. For example, it could be about reactions, honestly…. It's about seeing the three-dimensional structures they form and how they behave in reactions. In general, the presence of a process, it can be either physical or chemical. For example, the process of establishing equilibrium vapor pressure in a closed container where water vapor pressure forms is also a process. I think it's a question related to imagination, for instance, my ability to visualize a molecule in three dimensions….” (translated from Turkish)

Finally, the analyses of transcripts indicated that the students’ ratings were consistent with their explanations. For instance, Michael explained his rating to the item 1 (My problem-solving ability in chemistry is something…) as

Researcher (R): So, when you look at yourself, do you think your problem-solving ability in chemistry is something that can be improved?

Michael (M): I think it can be improved. I rate my potential to improve around 7–8 out of 10. Problem solving is primarily related to solving numerical and formula-based questions (for example, those involving gas laws, pressure, or volume)…The way chemistry problems are solved in high school and university doesn’t really match. For example, in chemistry now I use a calculator and get results like 13.28 × 10⁹ or something like that—there's no way I’d get such numbers back in high school. Even getting such numbers makes you wonder, like, “Is this too complicated? Am I doing something wrong?” But as you keep studying chemistry and exposing yourself to it, you naturally become more familiar with it—and, yes, your problem-solving techniques improve as well… (translated from Turkish).

To Michael, problem solving in chemistry means using formulas and mathematical operations to find solutions, which is very similar to the response process thinking reported in Santos et al. (2022). He understood the concept clearly and believed his problem-solving ability can develop. Consequently, as a result of the evidence obtained from analyses of cognitive interviews, we finalized the CheMI as indicated in Table 1. The mindset part of the survey was the CheMI (Part I). The CheMI has a unidimensional structure and is a 7-item, semantic differential survey (Santos et al., 2022). Higher CheMI scores indicate a growth mindset, whereas lower scores indicate a fixed mindset. The second part of the survey was on demographic information (i.e., student ID, department, sex and nationality), which helped to match students’ exam scores and survey responses and to conduct course level analysis.

Table 1 The revised Chemistry Mindset Instrument (CheMI) used in this study

	I can’t change at all						I can change a lot
1. My problem-solving ability in chemistry is something…	0	1	2	3	4	5	6	7	8	9	10
2. My ability to understand concepts in chemistry is something…	0	1	2	3	4	5	6	7	8	9	10
3. My ability to apply chemistry knowledge is something…	0	1	2	3	4	5	6	7	8	9	10
4. My ability to master (fully comprehend) chemistry content is something…	0	1	2	3	4	5	6	7	8	9	10
5. My ability to visualize chemical structures and processes is something…	0	1	2	3	4	5	6	7	8	9	10
6. My ability to use mathematical and logical reasoning in chemistry is something…	0	1	2	3	4	5	6	7	8	9	10
7. My overall chemistry intelligence is something…	0	1	2	3	4	5	6	7	8	9	10

---

Performance on course assignments was derived from the instructors who agreed to share the student gradebooks, including scores on midterm exams, the final exam, and the laboratory portion of the course. Two midterm examinations, each contributing 25% to the overall grade, and a comprehensive final examination, accounting for 40% of the overall grade, were conducted in person. The remaining 10% of the grade was allocated to general chemistry laboratory performance. Although the instrument was administered to students enrolled to three courses (CHEM A, CHEM B, and CHEM C), the instructors of CHEM C declined to share the student gradebooks.

Data analysis

Structural validity evidence is crucial both during the development of a measure and when it is adapted or modified for a new context. Internal structure validity examines not only the relationships between items and latent constructs but also how these relationships align with the hypothetical structure of the construct (Worthington and Whittaker, 2006; Arjoon et al., 2013). Alongside evidence from response-process interviews indicating that students comprehended and evaluated the adapted items in a manner consistent with the Turkish context, a confirmatory factor analysis (CFA) using the lavaan package in R (Rosseel, 2012; R Core Team, 2024) was conducted on the collected data from the students enrolled to three courses (CHEM A, CHEM B, and CHEM C) to evaluate the internal structure. The CheMI instrument reports an average score across the seven items for each respondent on a scale from 0 to 10, where 10 is the strongest indication of a growth mindset. CFA of students’ ratings on each CheMI item was used to determine the internal structure of CheMI as perceived by university students in Turkey. Goodness of fit statistics were used to evaluate whether the model fits the data. McDonald's omega (ω) value as a measure of internal consistency was computed for the CheMI because it accounts for unequal factor loadings across items (McDonald, 1999; Komperda et al., 2018).

Structural equation models (SEM) were employed using the lavaan package in R to examine the extent to which mindset mediated the relationship between students' performance on Midterm 2 and the Final Exam in CHEM A and CHEM B. In the models, chemistry mindset was modeled as a latent variable. The outcome variables were performances on Midterm 2 and Final exam, as observed variables. To further understand the extent to which a chemistry mindset mediates the relationship between performance on Midterm 2 and the Final exam among student subgroups with varying achievement levels, regression analyses were conducted using R. In these models, Final Exam performance was treated as the dependent variable, with Midterm 2 performance and chemistry mindset entered as predictors. While CFA included data from students enrolled in CHEM A, CHEM B, and CHEM C, subsequent SEM and regression analyses were limited to CHEM A and CHEM B, as access to CHEM C gradebooks was not permitted by the course instructors.

Results

To what extent does the intended factor structure of the Chemistry Mindset Instrument (CheMI; Santos et al., 2022) replicate in Turkey?

The behavior of 7 individual items on the 11-point Semantic Differential CheMI Likert scale was examined first to determine the appropriateness of data to run CFA. The mean of the average CheMI score is 6.58 with a standard deviation of 1.64 while skewness is −0.882 and kurtosis is 2.458 (Fig. 2). Table 2 summarizes descriptive statistics for each item in CheMI. The average response on items ranged from 5.85 to 6.89 on the 11-point scale with a standard deviation ranging from 2.05 to 2.29. For individual items, the highest skewness is −0.76 and the highest kurtosis is 0.52. Considering skewness and kurtosis values on both individual items and mean of average CheMI, we don’t have substantial evidence of non-normality.

Fig. 2 Response distribution of average CheMI (N = 871).

Table 2 Descriptive statistics for CheMI (N = 871)

Item	Mean	Std. Deviation	Skewness	Kurtosis
1. My problem-solving ability in chemistry is something…	6.76	2.05	−0.64	0.52
2. My ability to understand concepts in chemistry is something…	6.89	2.09	−0.76	0.34
3. My ability to apply chemistry knowledge is something…	6.68	2.08	−0.61	0.29
4. My ability to master (fully comprehend) chemistry content is something…	5.85	2.20	−0.35	−0.03
5. My ability to visualize chemical structures and processes is something…	6.47	2.17	−0.47	0.01
6. My ability to use mathematical and logical reasoning in chemistry is something…	6.84	2.29	−0.69	0.07
7. My overall chemistry intelligence is something…	6.58	2.06	−0.63	0.47

---

The CheMI instrument reports an average score across the seven items for each respondent on a scale from 0 to 10, where 10 is the strongest indication of a growth mindset. The mean of the CheMI score for the courses ranged from 6.54 to 6.63, with a standard deviation ranging from 1.59 to 1.70 (Table 3), which is similar to that reported for the students in US institutions (Santos et al., 2022; Demirdöğen and Lewis, 2023).

Table 3 Descriptive statistics on CheMI data across the courses

Course and N	Mean	Median	S.D.	Min	Max	Skewness	Kurtosis
CHEM A (N = 238)	6.57	6.71	1.70	0.00	10.00	−0.45	3.67
CHEM B (N = 326)	6.54	6.57	1.59	1.14	10.00	−0.39	3.32
CHEM C (N = 307)	6.63	6.86	1.64	0.00	10.00	−0.82	4.48

---

Prior to conducting CFA on CheMI, we estimated the differences in the mean of the average CheMI score among the students enrolled in various courses, if any. During the estimation, we calculated effect sizes as Cohen's d (Table 4), where values of 0.2 represent a small relationship, 0.5 medium, and 0.8 large. All three effect values are lower than small, indicating that the differences between each course pair are not substantial. Furthermore, equivalence testing (Lewis and Lewis, 2005a) provided statistical evidence that the mean CheMI scores were equivalent across the courses. Given these results—the effect sizes lower than small and established equivalence—we were justified in conducting a single CFA on the combined dataset (N = 871).

Table 4 Cohen's d between each course pair

Compared courses	Cohen's d
CHEM A and CHEM B	0.02
CHEM A and CHEM C	−0.04
CHEM B and CHEM C	−0.06

---

Finally, CFA was performed to identify the internal structure of the CheMI (i.e., unidimensional) using data from this research institution, which was also suggested by researchers (Santos et al., 2022). CFA was conducted with one factor structure to identify whether the model fits the data utilizing the lavaan package in the R program (Rosseel, 2012; R Core Team, 2024). Maximum likelihood was chosen as the method of estimation to obtain parameters. Items are specified to load only on one factor. All item measurement errors were estimated, and the variance of a latent factor was fixed to 1 to determine the factor model of CheMI. Full Information Maximum Likelihood Estimation was the default method of handling missing data. Evaluation of the model using four fit indices considerations (X² values closer to 0, CFI values ≥0.95, RMSEA values <0.08, and SRMR values <0.06; Hu and Bentler, 1999) revealed that data are consistent with the one factor model, namely unidimensional (Table 5). The standardized model produced and all paths showed standardized loading coefficients ranging from 0.612 to 0.796, which also supports the unidimensional structure of CheMI (Fig. 3). Moreover, McDonald's omega (ω = 0.89) value indicated that reliability of responses obtained from the single administration of the CheMI is excellent.

Table 5 Goodness-of-fit indicators for the unidimensional model conducted through CFA

Model	X ^2	df	CFI	RMSEA	SRMR
*** p < 0.001. Note. X^2 = chi-square, df = degrees of freedom, CFI = comparative fit index, RMSEA = root mean square error of approximation, SRMR = standardized root mean residual.
Unidimensional	76.794***	14	0.977	0.072	0.024

---

Fig. 3 Standardized CFA for CheMI.

To what extent does chemistry mindset mediate the relationship between students’ midterm exam performance and their final academic achievement in chemistry across engineering and natural sciences majors in Turkey?

Analysis of the descriptive statistics pertaining to exam performance data across the courses (Table 6) prompted us to employ structural equation modeling (SEM) to gain insight into the potential factors influencing the variation in student performance during General Chemistry 2. In CHEM A, student performance was consistent across exams, with means between 56.3 and 64.3 and distributions largely symmetric, although the final exam showed the greatest variability (SD = 25.8). In CHEM B, performance was highest in the laboratory (M = 75.7, SD = 10.3) and lowest in the final exam (M = 42.6, SD = 21.6), with distributions indicating strong performance in labs and midterms but considerable difficulty on the cumulative final. Overall, CHEM A students demonstrated steadier outcomes across assessments, while CHEM B students excelled in laboratory work yet struggled with the final exam. Due to the distinct exam performance patterns across courses, we conducted separate analyses for each course to examine the relationship between student mindset and performance using SEM.

Table 6 Descriptive statistics on exam performance data across the courses

Exam performances (out of)	Mean	Median	S.D.	Min	Max	Skewness	Kurtosis
CHEM A
Midterm 1 (100)	56.3	55	16.9	15	95	−0.060	2.55
Midterm 2 (100)	62.3	65	19.3	0	100	−0.0471	3.05
Final (150)	59.3	60	25.8	0	140	0.128	3.22
Laboratory (100)	64.3	59.4	19.7	9.5	97.2	−0.079	1.70
CHEM B
Midterm 1 (100)	66.9	68	17.1	0	100	−0.748	4.18
Midterm 2 (100)	59.5	62	21.4	0	100	−0.386	2.79
Final (100)	42.6	39.2	21.6	0	97.5	0.336	2.71
Laboratory (100)	75.7	76.3	10.3	33.6	95.1	−0.801	3.90

---

The relationships among performance on Midterm 2, chemistry mindset, and performance on the final exam were examined using structural equation modeling (SEM), reflecting the chronological sequence of data collection (Fig. 1). The results are presented separately for CHEM A and CHEM B in Fig. 4 and 5, respectively. In the model, mindset was specified as a latent construct (represented by circles), while Midterm 2 and Final exam performances were treated as observed variables (represented by squares). To facilitate direct comparison between the two courses, z-scores for Midterm 2 and final exam performance were used, thereby controlling for differences in instructor's grading practices and exam scales. This standardization enhances the validity of cross-course comparisons. Fig. 4 and 5 display the correlations among all constructs, with statistically significant values highlighted in bold red. Effect sizes for significant correlations were estimated using Cohen's f², with values of 0.02, 0.15, and 0.35 interpreted as small, medium, and large effects, respectively.

Fig. 4 SEM model describing the mediation of the relationship between performances on Final and Midterm 2 by chemistry mindset for CHEM A. Bold lines and values in red fonts indicate significant relationships.

The SEM model for CHEM A (natural sciences) (Fig. 4) demonstrated acceptable fit according to Hu and Bentler's (1999) criteria: χ² (26) = 63.810, p < 0.001, CFI = 0.954, RMSEA = 0.081, SRMR = 0.037. As illustrated in Fig. 4, the direct path from Midterm 2 performance to mindset (β = 0.010, p > 0.05) and the direct path from mindset to Final exam performance (β = 0.008, p > 0.05) was non-significant, indicating no evidence of a meaningful predictive relationship between these variables. In contrast, Midterm 2 performance significantly predicted Final exam performance (β = 0.616, p < 0.001), with a large effect size (Cohen's f² = 0.61). This suggests that prior academic performance was the dominant predictor of final exam outcomes for students in natural sciences.

The SEM model for CHEM B (engineers) (Fig. 5) demonstrated acceptable fit according to Hu and Bentler's (1999) criteria: χ² (26) = 60.034, p < 0.001, CFI = 0.962, RMSEA = 0.072, SRMR = 0.033. As shown in Fig. 5, the direct path from Midterm 2 to mindset was not statistically significant (β = 0.161), indicating that Midterm 2 performance did not meaningfully predict engineering students’ mindset. In contrast, mindset significantly predicted Final exam performance (β = 0.066, Cohen's f² = 0.0044), representing an effect size smaller than the conventional threshold for a small effect. This suggests a statistically reliable, albeit modest, contribution of mindset to final achievement for engineering students. The strongest predictor in the model was Midterm 2 performance, which had a significant direct effect on Final exam performance (β = 0.73, Cohen's f² = 1.14), indicating an effect size well above the threshold for a large effect.

Fig. 5 SEM model describing the mediation of the relationship between performances on Final and Midterm 2 by chemistry mindset for CHEM B. Bold lines and values in red fonts indicate significant relationships.

To what extent does chemistry mindset mediate the relationship between midterm and final exam performance for subgroups of students with low, medium, and high levels of midterm achievement, respectively?

Although the SEM model for CHEM B demonstrated overall good fit and highlighted the central role of Midterm 2 performance in predicting Final exam outcomes, the model estimated only average effects across the entire sample. Such an approach may mask important differences in predictive relationships between lower- and higher-performing students. Given that Midterm 2 was a very strong predictor of Final exam performance (β = 0.73, Cohen's f² = 1.14), we sought to determine whether this relationship—and the contribution of mindset—operated similarly across different performance levels. To address this, we moved to regression analyses stratified by low, medium, and high tertiles of performance in Midterm 2 (Table 7) to conserve statistical power. This approach allowed us to test whether mindset plays a more meaningful role for certain subgroups (e.g., high-achieving students) and to evaluate whether the predictive capacity of Midterm 2 varies depending on students’ standing within the course distribution.

Table 7 Hierarchical regression analyses predicting final exam performance in Course B by performance tertiles

Predictor	Low tertile β	Medium tertile β	High tertile β
Note: Standardized regression coefficients (β) are reported. ***p < 0.001.
Midterm 2 (z)	0.473***	1.07***	1.15***
Mindset	0.11	0.06	0.230***
R^2	0.28	0.14	0.49

---

In Course B, regression results revealed distinct patterns across performance tertiles. Among students in the lowest and middle tertile, Midterm 2 scores significantly predicted Final exam performance, whereas mindset was not a significant predictor. However, among the highest-performing students, both Midterm 2 (β = 1.15, p < 0.001) and mindset (β = 0.230, p < 0.001) significantly predicted Final exam outcomes. Notably, mindset demonstrated a unique, positive effect, such that students with higher mindset scores performed better on the Final even after controlling for Midterm 2. The model explained 49% of the variance, indicating that a substantial proportion of Final exam performance was accounted for by these predictors. In summary, Midterm 2 was a robust predictor of Final exam performance across all three performance groups in Course B. However, mindset emerged as an additional predictor only among the highest-performing students, suggesting that mindset plays a differentiating role at the upper end of achievement.

Discussion

This study represents the initial investigation into the relationship between students' chemistry mindset through the use of the CheMI and their academic performance in general chemistry courses at universities in Turkey, with a particular focus on how this relationship may vary across different majors, specifically engineering and natural sciences. The utilization of the CheMI in this study provided a dual benefit: it served as the primary data collection instrument and offered a means to evaluate the construct validity of the chemistry mindset through the interpretation of the resulting data. Consequently, the Discussion section will be structured around two key areas: an examination of the construct validity of the chemistry mindset and an analysis of its correlation with academic performance in general chemistry courses across different majors.

Chemistry education researchers advocated the use of discipline specific measures for mindset to increase the predictive power of the construct for achievement (Santos et al., 2022). Responding to this call, the present study employed the Chemistry Mindset Instrument (CheMI) to examine the relationship between chemistry mindset and academic performance in general chemistry. In addition, this investigation provided an opportunity to test the reciprocal interactions between academic experiences and mindset beliefs, as proposed by Limeri et al. (2020). To do this, we need validity and reliability evidence regarding the CheMI data in this new context first (Komperda et al., 2018; Lewis, 2022). While extensive evidence for response process validity is not within the scope of this study, we took an important step to ensure the instrument's integrity. We conducted cognitive interviews with three students in February 2024 to allow time for potential modification before survey deployment in mid-May. During these interviews, the students explained their thought processes as they responded to each CheMI item, confirming that their interpretations aligned with the intended design. The unidimensional structure of the CheMI, as originally documented by Santos et al. (2022), was supported by a confirmatory factor analysis (CFA) of the data collected in Turkey. This finding provides evidence for the instrument's internal structure validity within this new cultural and educational context (Worthington and Whittaker, 2006; Arjoon et al., 2013). This not only supports the broader generalizability of chemistry mindset as a psychological construct but also allows us to assess potential cross-cultural differences (Heine and Buchtel, 2009). Although emerging research highlights the context-dependent nature of mindset beliefs, which are shaped by specific social and educational environments (Little et al., 2016; Limeri et al., 2020; Kroeper et al., 2022; de Ruiter and Thomaes, 2023; Muradoglu et al., 2023; Asbury et al., 2025), the Turkish students’ experiences when learning chemistry and their beliefs about their ability when learning chemistry do not appear to be different from those in other countries (Santos et al., 2022; Demirdöğen and Lewis, 2023; Kattoum et al., 2024; Naibert et al., 2024; Pulukuri et al., 2025). This finding is further supported by the mean CheMI scores for the courses, which ranged from 6.54 to 6.63, indicating that students have growth mindset (with a standard deviation between 1.59 and 1.70; see Table 3). These values are consistent with scores reported for students in US institutions (Santos et al., 2022; Demirdöğen and Lewis, 2023).

The sequencing of data collection in this study provided an opportunity to investigate the reciprocal association between mindset and academic performance (Limeri et al., 2020) since the CheMI was administered after Midterm 2 and before Final exam. According to this perspective, attaining academic results that exceed one's expectations may foster the development of a growth mindset, while the endorsement of a growth mindset may, in turn, promote persistence in studying and subsequently lead to improved academic performance. Structural equation modeling (SEM) indicated that Midterm 2 performance did not significantly predict students’ mindset for either natural sciences or engineering majors. However, the relationship between mindset and Final exam performance differed across disciplines. Specifically, mindset emerged as a significant predictor of Final exam performance for engineering majors, whereas this association was nonsignificant for natural sciences majors. These findings failed to provide empirical support for the hypothesized reciprocal link between mindset and academic performance (Limeri et al., 2020). The relationship between mindset and academic performance remains inconclusive, with findings varying by educational level, the type of achievement measure employed, and cultural context. For instance, studies have reported null associations between mindset and achievement among Greek elementary and secondary school students (Leondari and Gialamas, 2002), high-achieving U.S. high school students (Wichaidit, 2025), French adults (Dupeyrat and Mariné, 2005), university applicants in the Czech Republic (Bahník and Vranka, 2017), Canadian pre-university students (Bazelais et al., 2018), and U.K. undergraduates (Li and Bates, 2020). Furthermore, irrespective of the educational level, cultural setting, or operationalization of achievement, meta-analytic evidence suggests that the association between students’ mindsets and academic performance is generally weak (Sisk et al., 2018), accounting for 1% of the variance in academic achievement. These findings are consistent with some of the results of our study, which revealed nonsignificant associations not only between Midterm 2 performance and mindset for both majors, but also between mindset and Final exam performance for natural sciences majors. Researchers have suggested several reasons why mindset and performance may not be significantly related. First, chemistry mindset has been characterized as multidimensional in nature (Santos and Mooring, 2024), whereas the unidimensional CheMI was used to measure students’ chemistry mindset in this study (Santos et al., 2022). A more comprehensive representation of the multiple facets that constitute students’ chemistry mindset is likely to yield more accurate predictions of their success in chemistry courses (Santos and Mooring, 2024). One way to address this is through the development and use of a multidimensional chemistry-specific mindset instrument including subscales for the facets: (1) chemistry mindset about self, (2) chemistry mindset about others, (3) interpretation of challenge in chemistry, and (4) behavioral response to challenge in chemistry. Emphasizing the role of challenge and consistent with this multidimensional nature, challenge has been documented as a mediating factor between mindset and performance (Santos and Mooring, 2022; Demirdöğen and Lewis, 2023). Students with a fixed mindset were less able to overcome the challenges they encountered when learning chemistry (Limeri et al., 2020) and demonstrated the lowest performance compared to those who either did not experience challenges or successfully overcame them (Santos and Mooring, 2022). More specifically, students endorsing a fixed mindset reported greater difficulty with chemistry-related challenges, and higher ratings of such challenges were associated with lower scores on summative assessments (Demirdöğen and Lewis, 2023). The study's scope was limited to a direct analysis of the mindset–performance relationship and therefore did not empirically investigate the role of challenge or other factors as potential mediators such as self-efficacy (Pulukuri et al., 2025), goal orientation (Naibert et al., 2024), and students’ perception of their instructor's mindset (Kattoum et al., 2024), which have been revealed as correlated with mindset (Santos et al., 2022). Self-efficacy has been identified as a mediator in both the direct relationship between mindset and exam performance (Pulukuri et al., 2025) and in more complex pathways involving mindset, goal orientations, and summative performance outcomes (Naibert et al., 2024). A growth mindset was found to exert a positive indirect effect on exam performance through self-efficacy as a mediator (Pulukuri et al., 2025). Performance goal orientation and mastery-avoidance orientation served as significant partial mediators in the relationship among mindset, self-efficacy, and summative scores (Naibert et al., 2024). Instead of focusing on student mindset, taking a different perspective, Kattoum et al. (2024) found that students who perceived their instructors as having a fixed mindset experienced greater academic misfit, which in turn predicted lower course grades. Incorporating the aforementioned variables to capture the complex nature of mindset may reveal a different pattern than that suggested by the nonsignificant relationship between mindset and performance.

In contrast to the current research and abovementioned past literature reporting nonsignificant relationships, studies with U.S. undergraduates have shown that chemistry-specific mindset is associated with both summative assessments (Santos et al., 2022; Demirdöğen and Lewis, 2023) and formative assessments (Santos et al., 2022), a pattern that is consistent with our finding of a significant relationship between mindset and Final exam performance among engineering majors. The difference in the nature of relationship between mindset and Final exam performance for natural sciences majors (nonsignificant) and engineering majors (significant) might be related to the pre-selection of students through a high-stakes national exam and the differing pedagogical and assessment methods. The high-stakes YKS examination acts as a selective filter. Admission to prestigious engineering programs at public universities in Turkey requires exceptionally high scores, indicating that these students have already demonstrated strong academic preparation. By contrast, natural science programs admit students with a broader range of scores, generally lower than those of engineering students. This pre-existing difference in academic preparation helps explain subsequent disparities between the cohorts, aligning with research showing that academic preparation is closely linked to performance in general chemistry (Tai et al., 2005; Xu et al., 2013; Frey et al., 2017). Course structure and assessment style may play a role in shaping how students approach chemistry learning (Lewis and Lewis, 2005b; Freeman et al., 2014). Engineering chemistry courses often emphasize cumulative, quantitative problem solving, requiring students to persist through complex, multi-step tasks (Barbera, 2013; Craig, 2013). Such environments may amplify the relevance of students’ beliefs about the malleability of ability, consistent with evidence that mindset effects are stronger under challenging learning conditions (Yeager and Dweck, 2012; Burnette et al., 2013; Paunesku et al., 2015). This aligns with the process model of mindsets, which advocates that the demanding engineering environment may serve as a contextual affordance that supports the development of a deeply entrenched growth mindset attractor for the high-performing students (de Ruiter and Thomaes, 2023). In contrast, natural science chemistry courses frequently stress factual recall and algorithmic procedures, where performance may rely less on perseverance and adaptive strategies, thereby attenuating the influence of mindset (Nakhleh, 1993; Cracolice and Deming, 2001; Smith et al., 2010). Further analyses of Final exam performance for engineering majors indicate that mindset exerts its strongest influence among higher-achieving students, consistent with the evidence that beliefs about the malleability of ability are particularly consequential under challenging learning conditions (Burnette et al., 2013), thereby highlighting its differentiating role at the upper end of the achievement distribution (Table 7). This is further supported by the fact that engineering majors struggled with the Final exam (Table 6).

Limitations

Several limitations of this study should be acknowledged. The study's primary limitation is its correlational nature. While SEM analyses suggested discipline-specific effects, unmeasured confounding factors—such as prior knowledge, motivation, and instructional quality—may have influenced the observed associations. Although the study considered course structure and assessment style, these factors were not experimentally manipulated, and their influence on the mindset–performance relationship was inferred rather than directly tested.

Another significant limitation is the scope of the study. The research was confined to a direct analysis of the mindset–performance relationship. It did not empirically investigate the role of potential mediating factors, such as students' self-efficacy, goal orientation, or their perception of the instructor's mindset. Existing literature suggests that these variables play a crucial role in the complex interplay between mindset and academic success. Without accounting for these factors, the study may be missing a more complete picture of the underlying mechanisms. The study's reliance on a unidimensional instrument (CheMI) to measure chemistry mindset is also a limitation. As noted in the discussion, a multidimensional approach that captures different facets of mindset (Santos and Mooring, 2024)—such as beliefs about self, beliefs about others, interpretation of challenge, and behavioral response to challenge—might yield more nuanced and predictive results. The unidimensional measure may have oversimplified a complex construct, potentially contributing to the nonsignificant findings observed in some groups. Additionally, in this study, we had access to course level data only, which limits our ability to consider the potential influence of pedagogical choices made by individual instructors.

Finally, while the study provides an important context within Turkey, the generalizability of the findings to other cultural or educational contexts may be limited. Although the study found consistency with some U.S. studies, the unique nature of the Turkish educational system, particularly the high-stakes national university entrance exam, could influence the mindset and academic preparedness of students in ways that may not be present in other countries.

Implications

The study highlights the differing roles of mindset across student majors and educational contexts. This implies that a one-size-fits-all approach to mindset interventions may be ineffective since mindset may be considered as a dynamic, context-dependent belief system that evolves through social interactions rather than as a fixed trait residing solely within the individual (the process model of mindsets [PMM]; de Ruiter and Thomaes, 2023). Chemistry educators should consider the academic background and specific challenges faced by their students (e.g., pre-selection through a high-stakes exam) when designing curriculum and implementing support strategies. The PMM suggests that effective mindset interventions require a two-step process: (a) perturbation of the entrenched mindset and (b) guiding the resulting increased variability toward a new, desired developmental trajectory (de Ruiter and Thomaes, 2023). This framework provides a theoretical basis for optimizing the existing intervention ideas in the chemistry education literature. Chemistry education literature provides intervention ideas tailored to specific challenges encountered during learning (Demirdöğen and Lewis, 2023) and specific domains under investigation (Khandelwal et al., 2024; Nguyen et al., 2024). Administering a challenge survey early in the semester could function as a perturbation by encouraging students to consider their current interpretation of the relationship between ability and effort. Once perturbed, the system has the potential for increased variability, which must be guided toward a growth mindset (de Ruiter and Thomaes, 2023). This requires context-level interventions that provide affordances to encourage new, learning-oriented behaviors. Drawing from self-regulation literature, students can set goals, assess their strengths and weaknesses, select and apply strategies, monitor progress, and adjust approaches as needed (Zimmerman, 2000; Schraw et al., 2006). Recognizing that challenges may differ for different majors suggests opportunities to integrate self-reported challenges into instruction (Demirdöğen and Lewis, 2023). Administering a challenge survey early in the semester could help students identify potential obstacles, followed by assignments where they articulate course goals, the impact of each challenge, and strategies to address them. Ongoing reflection papers throughout the semester could further support students in monitoring and adjusting their approaches, fostering active engagement with challenges and promoting goal attainment through an integrated model (Demirdöğen and Lewis, 2023). With a focus on challenge interpretation during their interventions, a more explicit and discipline specific design has the potential to improve the efficacy of mindset intervention (Khandelwal et al., 2024; Nguyen et al., 2024). Video interviews and/or in-person conversations with various department members about their own experiences of overcoming academic challenges might also be helpful (Nguyen et al., 2024). These ongoing discipline-specific, context-level interventions are aligned with the process model of mindsets, since they function as the necessary sustained push to overcome a fixed mindset attractor and/or (re-)entrench a growth mindset attractor (de Ruiter and Thomaes, 2023).

This study highlights the importance of context-specific, discipline-focused investigations into mindset, since mindset relates to final exam performance for engineering majors while there is no evidence of such a relationship for natural science majors. Considering the quantitative nature of this study, a qualitative approach (e.g., interviews, observations, and responses to open ended questions) focusing on student mindset beliefs and factors influencing those beliefs, especially with a focus on academic preparation and course structure, is needed. This qualitative investigation may also enable chemistry education researchers to explore a multidimensional measure of chemistry mindset, capturing facets such as mindset about self, mindset about others, interpretation of challenge, and behavioral responses to challenge, to better predict performance outcomes (Santos and Mooring, 2024).

Although this study focused on the simple relationships between mindset and performance since validation of the CheMI was also a concern, recent research on chemistry mindset has shed light on the relationship among self-efficacy, goal orientation, perception of the instructor's mindset, and academic performance (Santos et al., 2022; Kattoum et al., 2024; Naibert et al., 2024; Pulukuri et al., 2025; Sun et al., 2025). However, more comprehensive research is necessary to move beyond empirical relationships in the literature since each research study focuses on some variables not the whole. Future studies should empirically investigate the proposed mediating roles of challenge, self-efficacy, goal orientation, and students’ perceptions of the instructor's mindset. A more comprehensive model that includes these variables is necessary to fully understand the complex pathway between mindset and academic performance in chemistry.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

The data supporting this study—including survey responses, grade books, and demographic information—are not publicly available because participants did not consent to data sharing, and ethics approval for this study did not permit making the data publicly available. Sharing the data could compromise participant privacy.

Appendix

Sample demographics

Number of students across majors in each course.

Major	Course
	CHEM A	CHEM B	CHEM C	Total
Biology	77	0	0	77
Chemistry	0	0	1	1
Physics	117	0	0	117
Molecular biology and genetics	1	0	58	59
Physics education	21	0	0	21
Elementary science education	22	0	0	22
Civil engineering	0	99	0	99
Electrical and electronics engineering	0	1	0	1
Food engineering	0	59	0	59
Geological engineering	0	0	48	48
Mechanical engineering	0	162	0	162
Metallurgical and materials engineering	0	0	81	81
Mining engineering	0	0	44	44
Petroleum and natural gas engineering	0	0	67	67
Missing	0	5	8	13
Total	238	326	307	871

Acknowledgements

This material is based in part upon work supported by (while serving at) the U.S. National Science Foundation (author JEL). Any opinion, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation or of the Federal government.

References

1. American Educational Research Association, American Psychological Association and National Council on Measurement in Education [AERA, APA, & NCME], (2014), Standards for educational and psychological testing, 5th edn, Washington, DC: American Educational Research Association.
2. Arjoon J. A., Xu X. and Lewis J. E., (2013), Understanding the state of the art for measurement in chemistry education research: examining the psychometric evidence, J. Chem. Educ., 90(5), 536–545.
3. Asbury K., Carstensen B. and Klusmann U., (2025), Changing perspectives: the development of preservice teachers' field-specific ability beliefs across academic disciplines, Br. J. Educ. Psychol., 1–19 DOI:10.1111/bjep.70019.
4. Bahník Š. and Vranka M. A., (2017), Growth mindset is not associated with scholastic aptitude in a large sample of university applicants, Person. Individual Differ., 117, 139–143.
5. Barbera J., (2013), A psychometric analysis of the Chemical Concepts Inventory, J. Chem. Educ., 90(5), 546–553.
6. Bazelais P., Lemay D. J., Doleck T., Hu X. S., Vu A. and Yao J., (2018), Grit, mindset, and academic performance: a study of pre-university science students. EURASIA J. Math., Sci. Technol. Educ., 14(12), em1615.
7. Burnette J. L., O’Boyle E. H., VanEpps E. M., Pollack J. M. and Finkel E. J., (2013), Mind-sets of intelligence and self-regulation: toward a process model of goal setting and persistence, Soc. Person. Psychol. Compass, 7(11), 765–777.
8. Cracolice M. S. and Deming J. C., (2001), Diagnostic teaching in the college chemistry classroom: the role of conceptual understanding and problem-solving skills, J. Chem. Educ., 78(4), 553–556.
9. Craig P. A., (2013), A problem-solving approach to teaching general chemistry, J. Chem. Educ., 90(8), 1032–1038.
10. Demirdöğen B. and Lewis S. E., (2023), Investigating How Chemistry Students’ Reported Challenges Inform the Relationship between Mindset and Academic Performance, J. Chem. Educ., 100(9), 3252–3260.
11. de Ruiter N. M. and Thomaes S., (2023), A process model of mindsets: conceptualizing mindsets of ability as dynamic and socially situated, Psychol. Rev., 130(5), 1326.
12. Diener C. I. and Dweck C. S., (1978), An analysis of learned helplessness: continuous changes in performance, strategy and achievement cognitions following failure, J. Person. Soc. Psychol., 36, 451–462.
13. Diener C. I. and Dweck C. S., (1980), An analysis of learned helplessness: II. The processing of success, J. Person. Soc. Psychol., 39, 940–952.
14. Dinger F. C. and Dickhäuser O., (2013), Does implicit theory of intelligence cause achievement goals? Evidence from an experimental study, Int. J. Educ. Res., 61, 38–47.
15. Dupeyrat C. and Mariné C., (2005), Implicit theories of intelligence, goal orientation, cognitive engagement, and achievement: a test of Dweck's model with returning to school adults, Contemp. Educ. Psychol., 30(1), 43–59.
16. Dweck C. S., (2017), The journey to children's mindsets—and beyond, Child Dev. Perspect., 11(2), 139–144.
17. Dweck C. S., Chiu C. Y. and Hong Y. Y., (1995), Implicit theories: elaboration and extension of the model, Psychol. Inquiry, 6(4), 322–333.
18. Dweck C. S. and Leggett E. L., (1988), A social-cognitive approach to motivation and personality, Psychol. Rev., 95(2), 256–273.
19. Elliott E. S. and Dweck C. S., (1988), Goals: an approach to motivation and achievement, J. Person. Soc. Psychol., 54, 5–12.
20. Fink A., Cahill M. J., McDaniel M. A., Hoffman A. and Frey R. F., (2018), Improving general chemistry performance through a growth mindset intervention: selective effects on underrepresented minorities, Chem. Educ. Res. Pract., 19(3), 783–806.
21. Freeman S., Eddy S. L., McDonough, M., Smith, M. K., Okoroafor, N., Jordt, H. and Wenderoth, M. P., (2014), Active learning increases student performance in science, engineering, and mathematics, Proc. Natl. Acad. Sci. U. S. A., 111(23), 8410–8415.
22. Frey R. F., Cahill M. J. and McDaniel M. A., (2017), Students’ concept-building approaches: a novel predictor of success in chemistry courses, J. Chem. Educ., 94(9), 1185–1194.
23. Heine S. J. and Buchtel E. E., (2009), Personality: the universal and the culturally specific, Annu. Rev. Psychol., 60(1), 369–394.
24. Hu L. T. and Bentler P. M., (1999), Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives, Struct. Eq. Model.: Multidiscip. J., 6(1), 1–55.
25. Kattoum R. N., Montenegro-Montenegro E. and Baillie M. T., (2024), Students who perceive instructors to have a fixed mindset report a greater sense of academic misfit that leads to lower chemistry grades, J. Chem. Educ., 101(9), 3613–3623.
26. Khandelwal N. S., Mulualem M. F. and Theisen K. E., (2024), Growth Mindset as a Student Support Tool during a Pandemic, J. Chem. Educ., 101(4), 1499–1506.
27. Komperda R., Pentecost T. C. and Barbera J., (2018), Moving beyond alpha: a primer on alternative sources of single-administration reliability evidence for quantitative chemistry education research, J. Chem. Educ., 95(9), 1477–1491.
28. Kroeper K., Muenks K., Canning E. A. and Murphy M. C., (2022), An exploratory study of the behaviors that communicate perceived instructor mindset beliefs in college STEM classrooms, Teach. Teacher Educ., 114, 103717.
29. Leondari A. and Gialamas V., (2002), Implicit Theories, Goal Orientations, and Perceived Competence: Impact on Students’ Achievement Behavior, Psychol. Schools, 39, 279–291.
30. Lewis S. E., (2022), Considerations on validity for studies using quantitative data in chemistry education research and practice, Chem. Educ. Res. Pract., 23(4), 764–767.
31. Lewis S. E. and Lewis J. E., (2005a), The same or not the same: equivalence as an issue in educational research, J. Chem. Educ., 82(9), 1408.
32. Lewis S. E. and Lewis J. E., (2005b), Departing from lectures: an evaluation of a peer-led guided inquiry alternative, J. Chem. Educ., 82(1), 135–139.
33. Li Y. and Bates T. C., (2020), Testing the association of growth mindset and grades across a challenging transition: is growth mindset associated with grades? Intelligence, 81, 101471.
34. Limeri L. B., Carter N. T., Choe J., Harper H. G., Martin H. R., Benton A. and Dolan E. L., (2020), Growing a growth mindset: characterizing how and why undergraduate students’ mindsets change, Int. J. STEM Educ., 7(1), 1–19.
35. Little A., Sawtelle V. and Humphrey B., (2016), Mindset in context: Developing new methodologies to study mindset in interview data, Physics Education Research Conference Proceedings, American Association of Physics Teachers, pp. 204–207.
36. Lüftenegger M. and Chen J. A., (2017), Implicit theories: international perspectives from multiple domains, Z. Psychol., 225(2), 97.
37. McDonald R. P., (1999), Test Theory: A Unified Treatment, Lawrence Erlbaum Associates: Mahwah, NJ.
38. McKinney L., Novak H., Hagedorn L. S. and Luna-Torres M., (2019), Giving up on a course: an analysis of course dropping behaviors among community college students, Res. High. Educ., 60(2), 184–202.
39. Molden D. C. and Dweck C. S., (2006), Finding “meaning” in psychology: a lay theories approach to self-regulation, social perception, and social development, Am. Psychol., 61(3), 192–203.
40. Mueller C. M. and Dweck C. S., (1998), Praise for intelligence can undermine children's motivation and performance, J. Person. Soc. Psychol., 75(1), 33–52.
41. Muradoglu M., Arnold S. H., Leslie S.-J. and Cimpian A., (2023), “What does it take to succeed here?”: the belief that success requires brilliance is an obstacle to diversity, Curr. Direct. Psychol. Sci., 32(5), 379–386.
42. Naibert N., Mooring S. R. and Barbera J., (2024), Investigating the relations between students’ chemistry mindset, self-efficacy, and goal orientation in general and organic chemistry lecture courses, J. Chem. Educ., 101(2), 270–282.
43. Nakhleh M. B., (1993), Are our students conceptual thinkers or algorithmic problem solvers? J. Chem. Educ., 70(1), 52–55.
44. Nguyen T. S., Chan J. Y., Ha J. T., Umekwe-Odudu U. and Villafañe S. M., (2024), Development and implementation of chemistry mindset modules in two general chemistry courses at a Hispanic-serving institution: an exploratory study, Chem. Educ. Res. Pract., 25(2), 544–559.
45. Nicholls J. G., (1984), Achievement motivation: conceptions of ability, subjective experience, task choice, and performance, Psychol. Rev., 91, 328–346.
46. Paunesku D., Walton G. M., Romero C., Smith E. N., Yeager D. S. and Dweck C. S., (2015), Mind-set interventions are a scalable treatment for academic underachievement, Psychol. Sci., 26(6), 784–793.
47. Pulukuri S. V., Torres D. and Abrams B., (2025), Disentangling the Impacts of Self-Efficacy, Mindset Beliefs, and Surface Learning in Historically Marginalized Chemistry Students, J. Chem. Educ., 102(8), 3121–3130.
48. R Core Team, (2024), _R: A Language and Environment for Statistical Computing_. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/.
49. Rosseel Y., (2012), Lavaan: an R package for structural equation modeling, J. Stat. Softw., 48(1), 1–36.
50. Santos D. L., Barbera J. and Mooring S. R., (2022), Development of the Chemistry Mindset Instrument (CheMI) for use with introductory undergraduate chemistry students, Chem. Educ. Res. Pract., 23(3), 742–757.
51. Santos D. L., Gallo H., Barbera J. and Mooring S. R., (2021), Student perspectives on chemistry intelligence and their implications for measuring chemistry-specific mindset, Chem. Educ. Res. Pract., 22(4), 905–922.
52. Santos D. L. and Mooring S. R., (2022), Characterizing mindset-related challenges in undergraduate chemistry courses, J. Chem. Educ., 99(8), 2853–2863.
53. Santos D. L. and Mooring S. R., (2024), The complexity of chemistry mindset beliefs: a multiple case study approach, Chem. Educ. Res. Pract., 25(4), 1211–1228.
54. Schraw G., Crippen K. J. and Hartley K., (2006), Promoting self-regulation in science education: metacognition as part of a broader perspective on learning, Res. Sci. Educ., 36(1), 111–139.
55. Sisk V. F., Burgoyne A. P., Sun J., Butler J. L. and Macnamara B. N., (2018), To what extent and under which circumstances are growth mind-sets important to academic achievement? Two meta-analyses, Psychol. Sci., 29(4), 549–571.
56. Smith K. C., Nakhleh M. B., Bretz S. L. and Bretz S. L., (2010), Why are there differences in performance on algorithmic versus conceptual chemistry questions? J. Chem. Educ., 87(7), 753–759.
57. Sun H., Sun W., Liu X. and Liu Y., (2025), The relationship between chemistry growth mindset and chemistry academic engagement: a multiple mediation model, Chem. Educ. Res. Pract., 26(4), 996–1011.
58. Tai R. H., Sadler P. M. and Loehr J. F., (2005), Factors influencing success in introductory college chemistry, J. Res. Sci. Teach., 42(9), 987–1012.
59. Wang Y., Rocabado G. A., Lewis J. E. and Lewis S. E., (2021), Prompts to promote success: evaluating utility value and growth mindset interventions on general chemistry students’ attitude and academic performance, J. Chem. Educ., 98(5), 1476–1488.
60. Wichaidit P. R., (2025), Understanding growth mindset and chemistry mindsets of high-achieving students and the impact of influential language on learning motivation, Chem. Educ. Res. Pract., 26(2), 420–444.
61. Worthington R. L. and Whittaker T. A., (2006), Scale development research: a content analysis and recommendations for best practices. The Counseling Psychol., 34(6), 806–838.
62. Xu X., Villafane S. M. and Lewis J. E., (2013), College students’ attitudes toward chemistry, conceptual knowledge and achievement: structural equation model analysis, Chem. Educ. Res. Pract., 14(2), 188–200.
63. Yeager D. S. and Dweck C. S., (2012), Mindsets that promote resilience: When students believe that personal characteristics can be developed, Educ. Psychol., 47(4), 302–314.
64. Yeager D. S. and Walton G. M., (2011), Social-psychological interventions in education: They’re not magic, Rev. Educ. Res., 81(2), 267–301.
65. Zimmerman B. J., (2000), Attaining self-regulation: A social cognitive perspective, in Boekaerts M., Pintrich P. R. and Zeidner M. (ed.), Handbook of self-regulation, Academic Press, pp. 13–39.

---