Development and implementation of chemistry mindset modules in two general chemistry courses at a hispanic-serving institution: an exploratory study

Abstract

Retention and underrepresentation of diverse ethnic groups have been and continue to be problematic in the science, technology, engineering, and mathematics (STEM) disciplines in the United States. One foundational course that is required for all STEM majors is general chemistry. One way to increase retention and diversity in STEM majors is by targeting students’ social-psychological beliefs about their academic success through the implementation of social-psychological interventions. These short impactful exercises aim to change students’ thoughts, feelings, and beliefs about their academic success and affective characteristics. In this exploratory study, we designed and implemented two chemistry specific growth-mindset modules (GMMs) in two first-year chemistry courses (general chemistry 1 (GC1) and general chemistry 2 (GC2)) at a Hispanic-Serving Institution (HSI). Students worked on the GMMs asynchronously at two specific time points throughout the semester. Using a mixed-methods approach, we assessed students’: (i) changes in mindset beliefs, chemistry self-efficacy (CSE), and chemistry performance, (ii) perceptions towards failures and challenges, and (iii) perceptions on growth-mindset modules (GMM) after participation in GMMs. Overall, GC2 students shifted towards a growth mindset and away from a fixed mindset, with small to medium effect sizes detected. No statistically significant changes in GC1 students’ mindsets were detected throughout the study period. For both courses, students increased in CSE by the end of semester. Furthermore, GC1 students who participated in any portion of the GMM intervention achieved higher scores on the ACS exam compared to those who didn’t participate. Additionally, students’ written responses highlighted an improved attitudinal change towards failures and challenges after participating in GMMs. For both courses, over 95% of the students agreed that the GMMs were valuable, over 95% students indicated they developed more positive attitudes and perspectives towards challenges, and over 96% students believed they could learn challenging topics with effort, determination, and persistence. While these results show differences in performance, CSE, mindset scores, and attitudinal change after participation in GMMs, it is also important to acknowledge that self-selection into the study may be one of the factors for explaining such differences. Results and implications for practice are discussed.

---

Introduction

The growing issue of attrition in science, technology, engineering, and mathematics (STEM) fields poses threats to the workforce and its diversity. Even though there are more college students entering STEM fields, over 60% of them graduated with a non-STEM major degree in 2012 (Malcom and Feder, 2016). This trend can be attributed to poor performance in first year STEM gateway courses such as general chemistry (GC) and organic chemistry (OC) (Stone et al., 2018; Harris et al., 2020). Poor performance in GC has been shown to negatively impact graduation rates and STEM persistence. Research has shown that marginalized groups (i.e., Hispanic, Pacific Islander, American Indian/Alaska Native and African American) and females have higher attrition rates compared to their counterparts (Malcom and Feder, 2016; National Center for Education Statistics, 2019). When looking at final grades in GC and OC of 25 768 students, Harris et al. (2020) found grade gaps that existed for marginalized groups and sex, with marginalized groups and female students earning an average of 0.28 points lower than their counterparts. Because of these concerning trends in STEM attrition and its potential impacts on the STEM workforce, it is important to focus on what can be done in gateway classes such as GC to support students early on in their STEM careers.

One tool that has been gaining momentum over the years is implementation of social-psychological interventions in the classroom, particularly growth mindset interventions. These interventions target students’ subjective experiences and have the potential to change their thoughts, feelings, and beliefs in their academic performance (Yeager and Walton, 2011). Some studies have shown these interventions to be more impactful for marginalized groups (Fink et al., 2018).

Relationship between mindset beliefs, self-efficacy, and learning in chemistry

Mindset beliefs originated from Dweck's early work on implicit theories of intelligence and how one's beliefs about their ability influences their goal orientations, behavioral patterns, and achievement (Dweck and Elliott, 1983; Dweck and Leggett, 1988). Mindset refers to students’ beliefs about the degree to which ability is a stable trait (fixed mindset) or malleable trait (growth mindset). Students with growth mindsets are more intrinsically motivated and view their ability can be developed through persistence and practicing effective strategies, while those with fixed mindsets view their ability as fixed and are more likely to give up when faced with challenges. Research has shown that having a growth-oriented mindset positively influences many aspects of the learning experience, such as development of higher competence beliefs, task values, enjoyment, (Gunderson et al., 2017), learning goals, effort-based strategies (Blackwell et al., 2007); and engagement in college students (Zhao et al., 2021). Throughout various courses, student settings, and age groups, research has shown that participation in growth mindset workshops improved students’ motivation and achievement (Chiu et al., 1997; Aronson et al., 2002; Blackwell et al., 2007; Yeager and Dweck, 2012; Costa and Faria, 2018; Yeager et al., 2019). Furthermore, the correlation between self-efficacy and mindset has been well documented (Komarraju and Nadler, 2013; Santos et al., 2022). Students with low self-efficacy tend to believe that intelligence is likely innate and relatively unchangeable (i.e., fixed mindset), while students with high self-efficacy tend to believe that intelligence is likely malleable (i.e., growth mindset) (Komarraju and Nadler, 2013). Santos et al. (2022) also observed positive significant correlations between chemistry mindset, students’ self-efficacy, mastery goals, and course performance in first-year undergraduate classes. Since self-efficacy is associated with how one relates to and copes with challenges, those with higher self-efficacy have a higher likelihood of persisting in a difficult task (Bandura, 1977), are better able at recognizing that intelligence is a malleable trait (i.e. growth mindset), which in turn can lead to higher performance (Zusho et al., 2003; Kan and Akbas, 2006; Lalich et al., 2006; Uzuntiryaki-Kondakci and Senay, 2015; Boz et al., 2016). In organic chemistry classes, chemistry self-efficacy (CSE) has been shown to have a direct significant positive effect on exam performance and vice versa throughout the entire semester (Villafañe et al., 2016). Since CSE and mindset play a significant role in chemistry learning and performance, it is important to take a further look at the relationship among these variables.

Design of growth mindset interventions

During the growth mindset interventions, students learn about neuroplasticity (i.e. the idea that our brain is malleable) and its relation to having a growth mindset, watch videos about their peers share stories of how they overcame failures and challenges, work on online tutorials, and engage in discussion or reflective writing assignments (Campbell et al., 2021) The purpose of these interventions is to impart the belief to students that one's skills can be cultivated and improved through persistence and strategic effort and to emphasize the significance of making mistakes and challenges in the development of their abilities. While some interventions claim to improve academic performance, increase persistence in STEM, and eliminate gender and racial achievement gaps (Good et al., 2003; Fink et al., 2018; Yeager et al., 2019), other intervention studies fail to report these students’ benefits (Sriram, 2014; Burnette et al., 2020). Recently, three meta-analyses found a wide range of effect sizes among mindset intervention studies, which suggest students’ mindset is not always predictive of academic achievement (Costa and Faria, 2018; Sisk et al., 2018; Bui et al., 2023). Bui et al. (2023) found that discipline-specific mindset interventions had more positive results compared to general mindset interventions regardless of intervention target, content, or delivery mode, specifically in mathematics education. This suggests the importance of designing growth mindset interventions in the context of the discipline students are learning in.

Very few mindset intervention studies have been conducted in STEM courses (Canning et al., 2019; Campbell et al., 2021), particularly involving chemistry students at the university level (Fink et al., 2018; Wang et al., 2021). In a first-year general chemistry course, Fink et al. (2018) implemented a chemistry specific growth-mindset intervention and showed that the racial achievement gap (marginalized vs. non-marginalized groups) was eliminated. Likewise, Wang et al. (2021) showed that students’ academic performance improved after participating in a growth-mindset intervention in first-year general chemistry course. To date, three other mindset studies in chemistry have been conducted, but none of them included an intervention (Limeri et al., 2020a; 2020b; Santos et al., 2022). In second-semester OC class, Limeri et al. (2020) studied students’ mindsets and found that they became less growth and more fixed over the semester. In particular, those who struggled in the class experienced the most drastic shift towards a fixed mindset and attributed the negative learning experiences as a factor of this change in mindset. In contrast, those who passed the class and overcame challenges attributed their abilities as changeable. Students also associated their prior learning experiences to their current mindset beliefs. Together, these studies indicate the changeability of mindset beliefs and the association between academic performance and mindset beliefs. This emphasizes the importance of enhancing students’ mindsets through timely interventions.

However, in the aforementioned mindset studies, the student demographic makeup is very different from the current study's institution. Given the scarcity of mindset intervention studies in chemistry, particularly in Hispanic-Serving Institutions (HSI), we sought to design and implement chemistry-specific growth-mindset modules (GMMs) in two first-year chemistry courses (first-semester general chemistry (GC1) and second-semester general chemistry (GC2)) and explore students’ mindsets and perceptions towards challenges and failures, and overall perceptions on GMMs. Using a mixed-methods approach, the following research questions were pursued:

1. How does participation in chemistry GMMs influence students’ mindsets, chemistry self-efficacy (CSE), and course performance in chemistry?

2. How does participation in chemistry GMMs affect students’ perceptions towards challenges and failures?

a. To what extent are there differences in students’ perceptions towards challenges and failures between students from marginalized groups compared to students from non-marginalized groups?

3. What are students’ overall perceptions of chemistry GMMs?

Methods

(I) Mindset scale

In the past decades, the Dweck Mindset Instrument (DMI) has been used to measure one's mindsets in general domains of knowledge (Dweck et al., 1995; Dweck, 2006). The original DMI instrument consisted of three items that measured a single construct and has been assessed for reliability and validity across six studies (Dweck et al., 1995). The DMI was later modified and expanded to include a total of eight statements, which measured two separate constructs: fixed and growth (Dweck, 2006). The DMI consists of a six-point Likert scale, ranging from 1 (strongly agree), 2 (agree), 3 (somewhat agree), 4 (somewhat disagree), 5 (disagree), and 6 (strongly disagree). In this later version of the DMI (Dweck, 2006), each student had two mindset scores (fixed, growth). The lower the score on the six-point Likert scale, the more fixed or growth a student's mindset is. For example, a student with a fixed mindset score of 1 indicates a very strong fixed mindset whereas a student with a growth mindset score of 6 indicates a very weak growth mindset.

In the current study, the later version of DMI was used (Dweck, 2006). Two modifications were made to the original DMI. Firstly, each statement on the DMI was modified by adding the word “chemistry” to specify that chemistry intelligence was specifically what was being measured and to avoid response process validity concerns (Appendix 1). Dweck (2000) found that students’ mindsets can vary across academic disciplines and compared to general beliefs about intelligence, a mindset scale consisting of items with discipline-specific beliefs is more accurate and stronger predictor of students’ learning behaviors and achievement (Shively and Ryan, 2013; Gunderson et al., 2017; Costa and Faria, 2018). Another modification that was made to the DMI in this study was the addition of a direct-response item, which read, “This is a control question. Please select “Disagree 5”, as your response to this question.” This question served to screen out participants who responded to survey questions with insufficient effort (DeSimone et al., 2015). Therefore, the modified DMI implemented in our study consists of nine statements (Appendix 1). Students completed the modified DMI at three time points in the study: week 1 (pre-GMMs), week 3 (post-GMMs) and week 13 (delayed post-GMMs) (Fig. 1).

Fig. 1 Timeline of surveys given during study period.

In addition to participating in two GMMs, all students also participated in a learning strategies (LS) workshop designed by the corresponding author (Fig. 1). The design and implementation of the LS workshop is described elsewhere (Nguyen, 2022). This paper only focuses on the design and initial pilot study results from the implementation of GMMs.

(II) Chemistry self-efficacy (CSE)

To measure students’ perception of how well they perform on chemistry specific tasks, the Chemistry Self-Efficacy (CSE) was used (Ferrell and Barbera, 2015) (Appendix 2). The instrument consists of six items asking students to rate how well they understand chemistry concepts (e.g., describe properties of elements, interpret chemical equations, interpret graphs/charts) on a five-point Likert scale ranging from 1 (very poorly) to 5 (very well). Averaging the six items produces a single CSE score, with a higher CSE indicating that students have a higher chemistry self-efficacy. Students completed the CSE at two time points in the study: week 1 and week 13 (Fig. 1).

(III) Setting and participants

The study population involved students in GC1 and GC2 in three semesters (fall 2021, fall 2022, spring 2022) at a Hispanic-Serving primarily undergraduate institution in the southwestern region of the United States. About half of students (52.7%) are from marginalized group backgrounds, slightly more than half (58.4%) are female, and approximately 10% are first-year. Additionally, more than half of the students are from families in which their parents did not earn a four-year college degree, and about half of the students received the federal Pell Grant.

Students in GC1 met either twice a week for 165-minutes each or three times a week for 110-minutes. GC2 students met twice a week for 75-minutes each. All instructors are full-time and tenured or tenure-track, three of which conduct research in chemistry education. Across the two different courses and sections, the instructors met regularly to ensure that the curriculum, class content, and all examinations covered similar material. For GC2, it was taught by one instructor in fall 2021 (N = 83 students) and another instructor in spring 2022 (N = 87 students). For GC1, it was taught by three different instructors in spring 2022 (N = 172) and four different instructors in fall 2022, (N = 151 students). There were two of the same GC1 instructors who taught in both spring 2022 and fall 2022.

For all surveys in this study, student responses were collected online using Qualtrics survey software. Cases with missing data on any of the survey questions were excluded from the analyses. Students’ responses covered the entire range of scales on the surveys. No ceiling or floor effects were observed.

The final number of student participants in the study was calculated by filtering out those who: (i) failed to consent to participate, (ii) did not meet surveys’ deadlines, (iii) participated in a previous implementation of the exact same GMMs in the past (i.e., double exposure to the treatment), and (iv) incorrectly answered the control question on the DMI. Therefore, in GC2, this number was 29 (35% of enrolled) in fall 2021, and 32 (37% enrolled) in spring 2022. In GC1, the final number of students was 34 (20% enrolled) in spring 2022 and 50 (33% enrolled) in fall 2022. To increase statistical power, we combined data across semesters for each course. The implication of these numbers being a fraction of the total population is addressed in Limitations section of this paper.

Students were reminded through e-mail, Canvas, and in lectures of the survey opportunities throughout the semester. As an incentive for completing the surveys, students earned participation points (ranging from 3–5%) toward their course grade. Students who consented to participate and those who did not both earned extra credit by completing the survey. This study was approved by the Institutional Review Board. Human subject consent was obtained from all participants before the start of the study.

(IV) Chemistry achievement (ACS exam scores)

Chemistry exam achievement was measured using the 2020 version of the American Chemical Society (ACS) first-term exam for students taking GC1. The first-term general chemistry ACS exam contains 70 multiple choice questions and students were given a time limit of 110 minutes to complete the exams. All GC1 instructors applied the same method of scaling of the ACS exam for consistency. After scaling was applied, the ACS exam scores were standardized (z-scores) because ACS exam scores were gathered from different semesters and instructors. This was done to account for any differences in means and standard deviation and to ensure a uniform comparison across the different sections. Therefore, the reported ACS exam scores are the standardized ACS exam scores (z-ACS exam scores). For GC2, students did not take the ACS exam and hence no ACS exams were reported for GC2 students.

(V) Design of chemistry specific growth mindset modules (GMMs)

Since context plays a significant role to the success of a growth-mindset interventions, it is strongly recommended to embed principles of growth mindset specifically within the context of domain being studied (Yeager and Dweck, 2020; Bui et al., 2023). By embedding growth mindset principles specifically within the context of chemistry and presenting students with opportunities where they can engage with chemistry content through practice, these interventions can potentially lead to transformative changes in beliefs and academic achievement. Following these recommendations, two novel chemistry-specific GMMs were designed by the corresponding author of this paper (see Appendix 3 for more details). Students participated in two 30-minute GMMs delivered in an asynchronous online instruction mode via Canvas, Learning Management System (LMS) during weeks 1 and 3 of the semester (Fig. 1). Topics covered in these two modules include neuroplasticity, the factors that affect neuroplasticity, characteristics that differentiate between growth and fixed mindset, and strategies for cultivating and embracing a growth mindset during challenges and failures in the context of general chemistry course content. These were conducted through YouTube videos, a TED talk, and pre-recorded semi-structured interviews with one faculty member and students (more detailed description below). To ensure that students watched the videos and engaged with the content posted, students must answer some follow-up questions before they can proceed to the next activity on the GMMs (Appendix 3).

In the second GMM, students viewed short 10–15 minutes pre-recorded semi-structured interviews featuring a small, yet diverse group of students and professor representing various age groups, sex, ethnicity, and cultural backgrounds from the Department of Chemistry and Biochemistry of the corresponding author's institution. This group included a second-year graduate chemistry education student (male, non-URM), third-year undergraduate chemistry major student (female, non-URM), chemistry alumni (male, URM), and an assistant professor in chemistry (male, URM). During these one-on-one interviews, we asked each of them to share stories of how they overcame failures and challenges in the context of teaching and learning chemistry. The purpose was to intervene early and have GC1 and GC2 students learn the importance of cultivating a growth mindset in the face of challenges encountered when learning chemistry through watching these pre-recorded interviews early in the semester before they encounter challenges.

To assess the effectiveness of the modules, students completed surveys immediately before, immediately after, and ten weeks after participating in GMMs to evaluate how their beliefs about intelligence towards chemistry change throughout the semester. Additionally, students were asked to provide feedback for future modules and reflect on how participation in GMMs have influenced their attitude, perception, and motivation toward learning challenges (Appendix 3). Engaging in deliberate reflections where students are asked to connect key concepts learned from the intervention through writing exercises serves to strengthen one's belief in it. This practice, also known as “saying is believing” effect, is common in social-psychological interventions (Higgins and Rholes, 1978; Van den Hurk et al., 2019).

Study design

This study used a mixed-methods design that involved the triangulation of quantitative and qualitative data. All statistical analyses were conducted using SPSS version 28. Confirmatory factor analysis was conducted using MPlus software.

For the written reflection questions, all students’ responses were coded, and a summary of the themes and codes were identified using the constant comparison method (Glaser and Strauss, 1999). Responses were coded by two undergraduate students, two graduate chemistry education students, and the corresponding author. The research group is made up of a diverse team of members representing different ethnic backgrounds, sex, academic background, and age groups. After training and several discussions, an interrater agreement of 88% (for the second research question) and 96% (for the third research question) was reached among all raters. More details will be discussed in the results section under each of these research questions.

Results

(I) Confirmatory factor analysis (for DMI and CSE)

To validate the DMI instrument, Confirmatory Factor Analysis (CFA) using MPlus software was conducted on mindset data collected from implementing GMMs in fall 2021 in CNSM (Think Like Einstein) 101 – a general education course required for all incoming first-year STEM students. This course teaches students how to think like a scientist by practicing critical thinking, argumentation, and logical reasoning applied in chemistry, physics, biology, geology, and math. The student demographics in CNSM 101 is very similar to that of GC1 and GC2. Because of the small sample size of participants in the GC population, it was not feasible to conduct a CFA. The CFA on CNSM 101 data is generalizable to the GC population because all CNSM students will take or are currently taking GC1/GC2 and therefore exposed to the exact same version of GMMs as the GC students.

Data-model fit was assessed using indices such as Comparative Fit Index (CFI) (Bentler, 1990), Root Mean Square Error of Approximation (RMSEA) (Steiger, 2016), and Standardized Root Mean Square Residual (SRMR) (Chen, 2007). CFA conducted showed CFI ranging from 0.888 to 0.915, RMSEA 0.090 to 0.108 and SRMR 0.047 to 0.053 (Appendix 4). According to Hu and Bentler (1999), the acceptable ranges were of CFI values greater than 0.95, RMSEA values less than 0.06, and SRMR values less than 0.08. However, according to McNeish et al. (2018), acceptable CFI values could be as low as 0.775 and RMSEA values as high as 0.20. Naibert et al. (2021) corroborated these values in their CFA. Based on these findings, acceptable data-model fits were assessed using the following cutoff values: CFI (greater than 0.775), RMSEA (less than 0.20), and SRMR (less than 0.08). Because the produced CFA indices fell within the acceptable reported ranges, they confirmed that the DMI measures two constructs (fixed and growth mindsets) (Limeri et al., 2020). Appendices 5, 6, and 7 show the CFA models for three time points, weeks 1, 3, and 13, respectively. The sample sizes used to conduct the CFAs in weeks 1, 3, and 13 were N = 377, N = 339, and N = 346, respectively.

To validate the CSE instrument, the same procedure and range of fit indices (CFI, RMSEA, and SRMR) were used as described above. The analyses found CFI to range from 0.953 to 0.996, RMSEA 0.033 to 0.109 and SRMR 0.020 to 0.032 (Appendix 4). CFA results confirmed a single construct was being measured (CSE) using the instrument (Ferrell and Barbera, 2015). Appendices 8 and 9 show the CFA models for two time points, weeks 1 and 13, respectively. The sample sizes used to conduct the CFAs in weeks 1 and 13 were N = 231 and N = 275, respectively.

(II) Research Question #1: How does participation in GMMs influence students’ mindsets, chemistry self-efficacy, and course performance?

a. Students’ course performance (ACS exam score) for GC1. To determine if there was a difference in course performance between study participants compared to non-study participants, two independent samples t-tests were conducted (Tables 1, 2). Table 1 shows the mean ACS exam scores reported in z-scores between students who participated in at least one GMM during weeks 1 or 3 compared to those who did not participate in any GMMs for GC1. Students who participated in at least one GMM scored 0.15 standard deviations above the mean, while those who did not participate in any GMM scored 0.28 standard deviations below the mean for the ACS exam (t(326) = 3.76, p < 0.001, d = 0.44) (Table 1). Next, we looked to see if there were differences in ACS exam scores for students who participated in both GMMs compared to those who did not participate in both GMMs. Table 2 shows students who participated in both GMMs scored 0.23 standard deviations above the mean, while those who did not participate in both GMMs scored 0.07 standard deviations below the mean for the ACS exam (t(326) = 2.39, p < 0.05, d = 0.32) in GC1.

Table 1 Differences in mean ACS exam scores for students who participated in at least one GMM (weeks 1 or 3) vs. those who did not participate in any GMM for GC1

Participate in at least one GMM	N	Z-ACS exam score	SD	t	df	Effect size (d)
Note: ACS exam scores are z-scored. *** p < 0.001.
Yes	216	0.15	0.98	3.76***	326	0.44
No	112	−0.28	0.99

---

Table 2 Differences in mean ACS exam scores for students who participated in both GMM (weeks 1 and 3) vs. those who did not participate in both GMMs for GC1

Participate in both GMMs	N	Z-ACS exam score	SD	t	df	Effect size (d)
Note: ACS exam scores are z-scored. *p < 0.05.
Yes	79	0.23	0.86	2.39*	326	0.32
No	249	−0.07	1.03

---

Although this study did not test for the causal effect of growth mindset interventions on academic achievement, our findings suggest that the performance improvement may be associated with participation in GMMs. Specifically, students who participated in any portion of the GMM intervention achieved higher on the ACS exam compared to those who didn’t participate. Previous research studies have shown that growth mindset interventions have varying levels of predictive power on achievement measures such as improved grades (Blackwell et al., 2007; Broda et al., 2018, Fink et al., 2018) through adaptive behaviors learned from the intervention (Hong et al., 1999). However, it is important to acknowledge that self-selection into the study may be one of the factors explaining the differences in ACS exam scores. It is unknown whether students who self-selected to participate in the study were inherently more motivated to succeed in the course compared to non-study participants. Thus, it is important to consider the potential impact of self-selection bias as it could influence the interpretation of results.

b. Students’ CSE scores. Building on prior findings by Komarraju and Nadler (2013), which indicated that students with low self-efficacy tend to hold a fixed mindset, perceive intelligence as innate and unchangeable, while students with high self-efficacy tend to hold a growth mindset and perceive intelligence as malleable, we evaluated students’ perceptions about their abilities to perform chemistry-related tasks at two time points: weeks 1 and 13 in the semester for GC1 and GC2 (Appendix 10). Paired samples t-tests were conducted to determine whether students’ chemistry self-efficacy changed after participating in GMM (Appendix 11). Overall, for both GC1 and GC2, there were statistically significant increases in students’ CSE by the end of semester (Fig. 2).

Fig. 2 Chemistry self-efficacy scores across weeks 1 and 13 for GC 1 and GC 2. Error bars represent ± standard deviation. Note: higher scores indicate a stronger chemistry self-efficacy. **p < 0.01, *** p < 0.001.

GC2 students started off with a higher mean CSE score (M = 3.34, SD = 0.57) compared to GC1 students (M = 2.95, SD = 0.59). However, the magnitude of change in CSE throughout the semester was smaller for GC2 compared to GC1. For the GC1 population, a very large effect size was detected (d = 0.97) compared to GC2 population (d = 0.34) (Appendix 11). Similar to our findings, Moreno et al. (2021) found that students enrolled in general chemistry started the semester with a higher chemistry self-efficacy compared to students enrolled in introductory chemistry, but students in introductory chemistry experienced a greater change of CSE after a semester of instruction. However, it is important to acknowledge that self-selection into the study may be one of the factors explaining the differences in CSE scores. This will be further elaborated in the limitations of the manuscript.

c. Students’ overall mindset scores. Students’ mindsets were assessed and quantified into mindset scores (fixed, growth mindset scores) using the DMI at three time points of the semester: week 1 (pre-GMMs), week 3 (post-GMMs) and week 13 (delayed post-GMMs) for GC1 and GC2. Only students who responded to the DMI survey at all three time points were included as part of the data analysis for the first research question. Therefore, after combining data across semesters for each course, the number of students who completed all three DMI surveys was 60 in GC1 and 40 in GC2.

To assess how students’ mindsets changed over the study period, a 2 × 3 repeated measures ANOVA test was conducted with mindset scores (fixed score, growth score) as the dependent variable and time as the within-subjects (independent) variable (weeks, 1, 3, & 13). Overall, there were no statistically significant changes in GC1 students’ fixed and growth mindsets throughout the three time points (Fig. 3 and 4). In contrast, there were statistically significant changes in GC2 students’ fixed and growth mindsets (Fig. 3 and 4) with small to medium effect sizes detected. Specifically, GC2 students’ fixed mindset scores increased (less fixed) from week 1 to 3 (Fig. 3) and GC2 students’ growth mindset scores decreased (more growth) from weeks 1 to 3 and weeks 1 to 13 (Fig. 4). Below is a summary breakdown for each study population.

Fig. 3 Fixed mindset scores across weeks 1, 3, and 13 for GC 1 and GC 2. Error bars represent ± standard deviation. Note: Lower scores indicate a stronger fixed mindset. *p< 0.05, Cohen's d for weeks 1 or 3 for GC2 is 0.29 (small effect size).

Fig. 4 Growth mindset scores across weeks 1, 3, and 13 for GC 1 and GC 2. Error bars represent ± standard deviation. Note: lower scores indicate a stronger growth mindset. *p < 0.05, **p < 0.01, Cohen's d for weeks 1 to 3 for GC2 is 0.35 (small); Cohen's d for weeks 1 to 13 for GC2 is 0.48 (medium).

GC 1. Students’ growth mindset scores fluctuated across the three time points, however this difference was not statistically significant (F(2, 118) = 0.19, p > 0.05). Students’ fixed mindset scores also fluctuated across the three time points, but was also non statistically significant (F(2, 118) = 1.08, p > 0.05).
GC2. Students’ growth mindset scores statistically significantly decreased (more growth) across the three time points (F(2, 78) = 6.38, p < 0.01). Specifically, post hoc pairwise comparison tests showed that between weeks 1 and 3, students adopted a stronger growth mindset (p < 0.05, Cohen's d = 0.35 (small effect size)) and between weeks 1 and 13, students adopted an even stronger growth mindset (p < 0.01, Cohen's d = 0.48 (medium effect size)) (Cohen, 1988). Students’ fixed mindset scores increased (less fixed) across the three time points, but overall, it was non statistically significant (F(1.6, 63.9) = 3.09, p > 0.05). However, post hoc pairwise comparison tests showed between weeks 1 and 3, students adopted a less fixed mindset (p < 0.05, Cohen's d = 0.29 (small effect size)) (Cohen, 1988). No statistically significant differences were found for the other weeks.

The overall trends in mindset scores showed that students shifted towards a stronger growth mindset and a less fixed mindset from weeks 1 to 3 and from weeks 1 to 13 for both GC1 and GC2 students. Interestingly, from weeks 3 to 13, this trend (stronger growth mindset, less fixed mindset) continued for GC2 students. However, for GC1 students during this same time period, the reverse trend was observed (stronger fixed mindset, less growth mindset). During this ten-week period, no additional GMMs were implemented. Students took multiple assessments within the ten weeks and depending on their performance, this may lead them to change their beliefs about their intelligence about chemistry (Limeri et al., 2020b). These shifts in mindset beliefs across the three time points suggest the instability of mindset beliefs and a need to consistently remind students about growth mindset throughout the semester. It is important to note that these are general trends observed, but not all trends were tested to be statistically significant (i.e., GC1) due to low sample sizes and hence, these trends should be interpreted with caution. In contrast to our findings, Limeri et al. (2020b) found that second-semester organic chemistry students became less growth and more fixed by the end of semester and this trajectory of mindset change was related to their experience with academic challenges. This difference in mindset trends may be explained by the very different demographic makeup of the study population in each research study.

It is interesting to note that although the sample size for GC2 was smaller (N = 40) than GC1 (N = 60), statistical significance and small to medium effect sizes were detected (d = 0.29–0.48). One possible reason for why statistical significance was detected in GC2 (and not GC1) may be due to differences in instructor effects. In GC2, there were only two instructors teaching the course; for GC1, there were five different instructors teaching the course at the time of data collection. Since all students participated in GMMs asynchronously, instructor variability across the different sections should not be a concern. However, we did not have control over how much emphasis and reinforcement of growth mindset discussions instructors had in each classroom. Research has shown that instructor's mindsets have a direct influence on students’ mindset and their academic performance (Canning et al., 2019). Whether faculty members adopt a growth or fixed mindset will affect the way they communicate and advise students, the tone and attitude they choose to use, and how they encourage or discourage students to keep trying in the face of challenges. The increased number of different instructors teaching GC1 compared to GC2 may have contributed to an increased variability in students’ responses thereby making it more difficult to detect a statistically significant difference in students’ mindsets (i.e., high noise to low signal ratio).

Additionally, it is also interesting to note that GC2 students started with a more fixed mindset compared to GC1 students. A possible reason for why GC2 students started off with a worse mindset than GC1 students may be influenced by their negative academic experiences encountered from prior chemistry courses. This is important to address because whether students view their abilities as a fixed or growth trait has significant implications for their responses to failure and academic outcomes. Limeri et al. (2020b) found that second-semester organic chemistry students experienced worsening of mindset over the course of a semester (i.e. shift towards a fixed mindset and away from a growth mindset) with many mentioning prior negative academic experiences to be a factor. Together, these findings imply that mindset beliefs are dynamic and can be influenced by prior students’ experiences.

III. Research Question #2: How does participation in GMMs affect students’ perceptions towards challenges and failures? To what extent are there differences in students’ perceptions towards challenges and failures between students from marginalized groups compared to students from non-marginalized groups?

After students participated in both GMMs, they were asked to complete short written reflections regarding how their attitudes and perceptions toward failure and challenges had changed by week 3. For the written reflection questions, all students’ responses were coded, and a summary of the themes were identified using the constant comparison method (Glaser and Strauss, 1999). After training and several discussions, an interrater agreement of 88% was reached among all raters. A total of 103 student responses were analyzed and coded (35 from GC1; 68 from GC2). Analysis of students’ response between both courses (GC1, GC2) across all three semesters showed no major differences, suggesting a consistency in both the implementation of GMMs and students’ interpretation of the GMMs. Table 3 shows a summary of themes gathered in response to students’ attitudes and perceptions about failures and challenges. These themes include acceptance, essential characteristics, behavioral, and neutral. Each student response can be categorized into more than one theme. Appendix 12 provides additional examples of student quotations, extending beyond the ones shown in results section below.

Table 3 Summary of themes from students’ responses to their attitudes and perceptions towards failures and challenges (Week 3) for GC1 and GC2. (Note: each student response can be categorized into more than one theme)

Theme	Description
Acceptance	Indications of development of positive outlook/mindset on failures and challenges (viewing them as opportunity or using them as motivation/encouragement)
	Indications of accepting failures and challenges as part of learning and a willingness to persist and overcome challenges.
Essential characteristics	Indications of students realizing the key characteristics needed to overcome challenges (e.g., having self-belief, determination, self-assurance, hard work, persistence, and strategic practice)
Behavioral	Indications of planning/implementing new approaches/actions when encountering failures and challenges.
Neutral	Indications of unchanged perception/attitude

---

Acceptance. After participating in both GMMs, 71% of GC1 students and 58% of GC2 students indicated they have accepted failure and challenges encountered during the learning experience as opportunities to learn and grow. They also indicated a willingness to persist amidst challenges.

“I have become more accepting towards failure and challenges, and just see them as a means to be optimistic and positive towards myself and the situation”. (non-marginalized group, GC1)

“I have learned that failure is also an important process that is needed in order to learn and better myself. Seeing this I have begun to see failure and challenge as things that are necessary for my growth.” (non-marginalized group, GC2)

Furthermore, for both GC courses, students from the non-marginalized group revealed they were more likely to welcome failures and challenges as part of the learning experience compared to their counterparts (60% non-marginalized group (GC1); 61% non-marginalized group (GC2)).

Essential characteristics. After participating in both GMMs, 6% of GC1 students and 9% of GC2 students recognized the importance of having self-belief, determination, self-assurance, hard-work and strategic practice are essential to overcoming challenges.

“It has made me realize that as long as I have determination and persistence in something that [I’m] trying to learn, I will be able to succeed.” (marginalized group, GC1)

“Always believe in yourself.” (marginalized group, GC2)

When disaggregating the data by minority status, there was no difference in frequency of comments between students from marginalized group vs. non-marginalized group in GC1. However, in GC2, more students from non-marginalized group recognized the necessity of having certain essential characteristics to overcome challenges (62% non-marginalized group, GC2).

Behavioral changes. After participating in GMMs, students gathered new perspectives and alternative approaches to learning. 17% GC1 students and 20% GC2 students indicated the importance of reflecting, adapting, and modifying their study behaviors when they encounter challenges and failures.

“I keep trying and ask for help when needed.” (marginalized group, GC1)

“Definitely change my perspective in how I need to study in order to pass university classes. I need to put in the time in order to successfully learn and pass all my courses.” (non-marginalized group, GC2)

In GC1, there were no differences in frequency of students’ comments between marginalized vs. non-marginalized groups. However, in GC2, more students from the non-marginalized group realized the importance of modifying their study behaviors after participating in GMMs (58% non-marginalized group, GC2).

Neutral. This theme represents students who show a neutral disposition or no change towards attitudes and/or perceptions towards challenges and failures. Only a small portion of students (9% GC1; 19% GC2) indicated they experienced no changes in their attitudes and/or perceptions towards challenges and failures after participating in GMMs. Some students elaborated that they had previously been exposed to growth mindset before and/or they always had a positive attitude towards learning challenges.

“It has not really. I still am pretty focused when I need to be.” (marginalized group, GC1)

“My mindset about failure and challenges have not changed that much because I have always had the mindset that challenges are good learning opportunities and failure is a requirement to succeed.” (non-marginalized group, GC2)

In both courses, GMMs had a stronger impact on transforming the attitudes and perceptions of marginalized students towards challenges and failures when compared to non-marginalized group students (33% marginalized groups, GC1; 19% marginalized group, GC2).

Overall, students from the non-marginalized group were more accepting of failures and challenges and their comments indicated they were more able to recognize the importance of reflecting, adapting, and modifying their study behaviors when they encounter challenges and failures. In addition, students from the non-marginalized group were better at recognizing the importance of having specific traits for overcoming challenges. While students from the non-marginalized group showed higher awareness of these factors, the GMMs had a less pronounced impact on them. Instead, students from the marginalized group were more impacted by the GMMs, particularly in terms of transforming their attitudes and perceptions towards challenges. Research has shown that students with higher self-efficacy are more likely to persist and display positive affect when going through challenging tasks and exercise more self-regulated behaviors compared to students with lower self-efficacy (Bandura, 1977; Zimmerman, 2000; Komarraju and Nadler, 2013). The frequency of students’ comments suggests that non-marginalized students may have higher self-efficacy compared to marginalized students. In a semester of a college-level preparatory chemistry course, Villafañe et al. (2016) showed that Black and Hispanic males started the semester with a higher initial chemistry self-efficacy. However, as the semester progressed, their chemistry self-efficacy decreased compared to White males (Villafañe et al., 2014). This finding suggests that chemistry self-efficacy is influenced by students’ prior experiences and different subgroups of students experience the course in different ways throughout the semester. Consequently, these different experiences impact students’ beliefs about their ability to overcome challenges, the way in which they approach and respond to challenges in their coursework.

The themes that emerged from this study corroborated with mindset-related constructs found in prior research studies on growth mindset (Dai and Cromley, 2014; Smiley et al., 2016; Fink et al., 2018) and highlight different stages of the cognition-affect-behavior chain (Dweck and Leggett, 1988). The cognition-affect-behavior chain reflects the different stages of mindset development and how one's beliefs about intelligence influences their emotions and actions. It is important to note that many students described their ability to transition from thinking negatively about themselves to accepting mistakes and persevering during challenging experiences after attending the modules. This transition led to a positive outlook on failures and the belief that success comes from hard work as explained by a student below:

I always had a fixed mindset. I thought I was a failure and will never succeed in life. Instead of challenging myself, I just give up and start to think negatively about myself. However, after participating in these[modules], I’m beginning to think that maybe I just need to be more patient, and if I work hard enough, I will ultimately reach my full potential. To grow as a person, I need to start challenging myself and realize that it's okay to make mistakes and fail. (GC2, non-marginalized)

Triangulating students’ mindsets scores with their written responses revealed development of mindset-related constructs such as accepting challenges and failures as part of the learning process, gaining new perspectives, adopting new approaches towards studying, and developing an overall positive attitude towards failure and challenges (Table 3).

(IV) Research Question #3: What are students’ overall perceptions of GMMs?

To understand students’ overall perceptions of GMMs, we investigated: (i) the impact of GMMs on students’ learning in chemistry classes and (ii) students’ feedback of GMMs.

(i) Impact of GMMs on students’ learning in chemistry classes. Students were asked to read and rate three statements regarding the impact of GMMs on their learning in chemistry classes on a six-point Likert scale ranging from 1 (strongly agree) to 6 (strongly disagree) (Tables 4 and 5). For both courses and across the three semesters, over 95% of the students agreed (strongly agreed, agree, somewhat agree) that the GMMs were valuable (97.2% GC1; 95.7% GC2). In addition, after participating in GMMs, over 95% students indicated they developed more positive attitudes and perspectives towards challenges (96.3% GC1, 95.8% GC2). Over 96% students believed they could learn challenging topics with effort, determination, and persistence (96.2% GC1, 98.7% GC2). These findings corroborate with themes extracted from students’ responses with regards to students’ perceptions towards challenges and failures (Table 3). The overwhelmingly positive responses highlighted students’ recognition for potential benefits of GMMs on their learning.

Table 4 Students’ responses to statements assessing impacts of GMMs on their learning in GC1

	After the growth mindset sessions, I believe I can learn challenging topics with effort, determination, and persistence	I believe the growth mindset sessions were valuable	After the growth mindset sessions, when I encounter challenging situations, I now approach them with a more positive attitude and perspective
Strongly disagree	0	0	0
Disagree	1 (0.9%)	1 (0.9%)	1 (0.9%)
Somewhat disagree	3 (2.7%)	2 (1.8%)	3 (2.7%)
Somewhat agree	12 (10.9%)	13 (11.8%)	18 (16.3%)
Agree	40 (36.3%)	41 (37.2%)	46 (41.8%)
Strongly agree	54 (49%)	53 (48.2%)	42 (38.2%)
Total responses	110

---

Table 5 Students’ responses to statements assessing impacts of GMMs on their learning in GC2

	After the growth mindset sessions, I believe I can learn challenging topics with effort, determination, and persistence	I believe the growth mindset sessions were valuable	After the growth mindset sessions, when I encounter challenging situations, I now approach them with a more positive attitude and perspective
Strongly disagree	0 (0%)	0 (0%)	0 (0%)
Disagree	0 (0%)	1 (1.4%)	1 (1.4%)
Somewhat disagree	1 (1.4%)	2 (2.8%)	2 (2.8%)
Somewhat agree	6 (8.5%)	13 (18.3%)	15 (21.1%)
Agree	34 (47.9%)	23 (32.4%)	26 (36.7%)
Strongly agree	30 (42.3%)	32 (45%)	27 (38%)
Total responses	71

---

(ii) Students’ feedback of GMMs. To evaluate the implementation of GMMs, we asked for student feedback about the modules and asked them to identify areas of strengths and improvements in week 3 (after they have participated in both GMMs). A total of 88 student responses were analyzed and coded (29 from GC1; 59 from GC2) and an interrater agreement of 96% was reached. Analysis of students’ response between both courses across the semesters showed no major differences, suggesting a consistency in the implementation of GMMs and students’ interpretation of the GMMs across the different populations. Table 6 shows a summary of themes and codes gathered in response to students’ feedback about the GMMs. These themes include areas of strengths, areas of improvement, and neutral. Each student response can be categorized into more than one theme and code. Appendix 13 provides additional examples of student quotations, extending beyond the ones shown in the results section below.

Table 6 Summary of themes and subthemes from students’ responses to feedback of GMMs (Week 3) for GC1 and GC2. (Note: each student responses can be categorized into more than more than one theme and code)

Theme	Sub-theme	Description
Areas of strengths	Resources	Students appreciate articles, videos, and outside resources provided.
	Development of new perspectives	Students gain positive experiences, perspectives, mindsets, and ideas
	Content	Students appreciate the type of information presented and the structure workshop
Areas of improvement	Interaction	Students request for more interactive activities.
	Modification	Students request for modifying content of modules by adding more information about growth-mindset and/or specific assignments that help them practice growth mindset
Neutral		No suggestions or improvements needed

---

Areas of strengths

Many students described the use of different resources (articles, videos, TED talks, YouTube videos), development of new affective perspectives, and modules’ content as key strengths of the GMMs. 51.7% GC1 students and 43.7% GC2 students found that the variety of resources shared in the GMMs to be very informative and helpful.

I think strengths in the [modules] were the videos since they were entertaining and had a lot of information. (GC2)

Some of the strengths were the video and the reflection questions […] (GC1)

Furthermore, students valued the written reflective questions throughout the modules as the questions promoted deep internalization of the concept of a growth mindset and the ability to immediately apply the concept within their individual contexts. Reflective writing, a common and recognized social-psychological intervention, has been well documented to be effective in reinforcing one's beliefs (i.e. “saying is believing” effect) (Higgins and Rholes, 1978). In addition, 34.5% GC1 and 30.7% GC2 students reported having more positive outlooks on failures and challenges as an area of strength garnered from after attending the GMMs. It is worthy to note that students could highlight this new perspective towards challenges and failure without the help of a prompt. In the second research question, we specifically prompted students to describe their perspectives towards challenges and failures after participating in GMMs, and the themes “acceptance” and “essential characteristics” emerged (Table 3). These themes corroborated with “development of new affective perspectives” described within this section (Table 6). For example, some students commented on the following:

I really like how encouraging these [modules] were. Shows me that no matter what, I can accomplish anything if I tried […] (GC2)

Some strengths are helping others see that growth mindsets are essential to improving. (GC1)

24.1% GC1 and 41.1% GC2 students also mentioned the content presented as another strength of the GMMs. For example,

I also thought that the real life videos of people from different stages of life being able to share their experiences of failures, allows me to understand that we are not perfect and that we are bound to make mistakes and that is okay. (GC1)

I think that the open ended questions about what the students have personally struggled with are really good because they leave a lot of room for interpretation of different experiences. (GC1)

This student appreciated the personal anecdotes and stories shared by current students and professor. Learning from real people and hearing about their personal experiences were more impactful and resonated with the students more than being taught theories or concepts about growth mindsets.

Areas of improvement

Students requested having more interactive activities and modification of content as areas of improvements of the GMMs. 26.7% GC1 and 17.9% GC2 students commented about including more interactive activities in the GMMs. 53.3% GC1 and 46.8% GC2 commented on having additional content to supplement the current existing content in GMMs. For example,

An improvement I suggest for future [modules] is to try to have more interactive activities to test our knowledge of growth mindsets. (GC2)

I would suggest more examples on how you could change your mindset, as well as the benefits, other than the anatomical benefits. (GC1)

These students’ comments highlighted the necessity of converting asynchronous modules into interactive in-person workshops and integrating additional intentional interactive activities to help students put what they learn into practice, particularly in the asynchronous environment to build more engagement. For example, instructors can design small reflection assignments that utilizes growth-mindset related strategies before and after each exam to actively remind students about the importance of having growth-mindset during challenging times throughout the semester. Furthermore, some students were interested in learning more about practical ways to shift their mindsets, understanding the benefits of cultivating a growth mindset, and delving into a more in-depth understanding of the connection between growth mindset and neuroplasticity.

Neutral

A smaller proportion of students commented there were no improvements needed and/or they had no further suggestions for the GMMs (20% GC1 and 37% GC2). Many of the students also followed up by mentioning it was a positive educational experience. For example, a student mentioned:

In my opinion, I don’t think there are any improvements that should happen for future [modules] because I believe it was amazing and helped me really think [for] the first time. (GC2)

I did not really see any points of weakness that would need improvements. (GC1)

Overall, using a mixed-methods study design which involved a pre- post-, and delayed-post data collection process, the effect of GMMs on student learning could be more holistically studied. Quantitatively, statistical analysis revealed no statistically significant differences in mindset scores among GC1 students across the study period. However, statistically significant increases in growth mindset and decreases in fixed mindset were detected for GC2 students. While quantitative findings did not distinctly show the effectiveness of the GMMs for the two courses, students’ written responses provided much more in-depth insights into the effectiveness of GMMs.

Conclusions

In the current study, two novel chemistry-specific GMMs were developed and implemented in two first-year chemistry courses across three semesters at a Hispanic Serving Institution. The intentional structural design of the two-part series of GMMs ensures that students first learn about the association between neuroplasticity and the concept of growth mindset in the first GMM. Subsequently, they learn about the importance of cultivating a growth mindset when faced with challenges while learning chemistry in the second GMM by viewing pre-recorded interviews from a diverse group of students and professor. By watching these pre-recorded interviews detailing personal stories of overcoming failures and challenges, students gain valuable vicarious experiences, which has been shown to increase self-efficacy (Bandura, 1977; Usher and Pajares, 2008). Throughout the two GMMs, students also engaged in reflective writing exercises which scaffolded them to actively apply their knowledge of mindsets to learning challenging topics in their chemistry course. The GMMs were designed as asynchronous online modules and delivered via the institution's learning management system platform, Canvas. Implementing the GMMs on Canvas platform makes the intervention material easily accessible, sustainable, and transferable to other chemistry instructor's classes. It is also cost-effective and allows students the flexibility to work on the modules at their own pace.

Using a mixed-methods study design which involved a pre-, post-, and delayed-post (weeks 1, 3, 13) data collection process, we examined: (i) students’ mindset beliefs, chemistry self-efficacy, and chemistry performance, (ii) perspectives towards challenges and failures, and (iii) overall perceptions of the novel GMMs.

Overall, there were no statistically significant changes in GC1 students’ mindsets throughout the study period. In contrast, there were statistically significant changes in GC2 students’ mindsets with small to medium effect sizes detected. GC2 students became more growth in mindset from weeks 1 to 3 and from weeks 1 to 13 and became less fixed in mindset from week 1 to 3. For both GC1 and GC2, there were statistically significant increases in students’ chemistry self-efficacy by the end of semester after participating in GMMs. Furthermore, our analysis showed that students who participated in any portion of the GMM intervention achieved higher on the ACS exam compared to those who didn’t participate. While results show that ACS exam scores, CSE, scores, and mindset scores improved after participation in GMMs, it is also important to acknowledge that self-selection into the study may be one of the factors for explaining the differences in these scores detected as it is unknown whether motivation may be a confounding variable. Further discussion is provided in the limitations section of this manuscript.

Students’ written responses highlighted an improved and new perspective towards viewing failures and challenges. Students from non-marginalized group showed a higher degree of: (i) acceptance of failures and challenges (acceptance), (ii) belief in their abilities to succeed and overcome challenges (essential characteristics), and (iii) active planning or finding strategies to conquer challenges and failures (behavioral). The higher frequency of students’ comments in these categories suggests that non-marginalized students may have higher self-efficacy compared to marginalized students, show a higher tendency to persist through challenges, and demonstrate a greater ability to cope with challenges.

Furthermore, students appreciated the resources shared, content delivered, and new perspectives gained as key strengths of the GMMs. Some students requested inclusion of more interactive activities and additional content to supplement the existing content in GMMs as areas of improvements. Overall, for both courses, students recognized the strength, value, and utility of the GMMs. Over 95% of students agreed that the GMMs were valuable, over 95% students indicated they developed more positive attitudes and perspectives towards challenges, and over 96% students believed they could learn challenging topics with effort, determination, and persistence in chemistry.

Together, the quantitative and qualitative results from this exploratory study highlight the potential benefits of chemistry specific GMMs implemented in two first-year chemistry courses. However, given the low sample size in our current study, it's important to acknowledge the data presented may not entirely represent the characteristics of the target population. Further discussion is provided in the limitations section of this manuscript.

Implications

What are the practical implications of the findings from this research? The primary message is that instructors should teach students early in the semester about the importance of having a growth-mindset and provide them with necessary skills and resources on how to actively practice having a growth-mindset throughout the semester. This is particularly important before and after exam period, when academically at-risk students are most vulnerable of dropping out of the course. Some students’ comments suggest the importance of incorporating additional assignments into the course that provide practical application of growth-mindset concepts, such as including more intentional reflective exercises and opportunities for actively practicing a growth mindset. We recommend instructors provide scaffolded experiences in the classroom where students have multiple opportunities to fail, practice, and demonstrate mastery learning by coupling metacognitive exercises to allow students to reflect and connect their experiences by using a growth-mindset. We recommend instructors can do this by incorporating specifications grading and pre/post exam reflections into their practice. In specifications grading, students are given multiple opportunities to demonstrate successful mastery of each course learning objective that is tied to short low-stakes assessments (Nilson, 2014). Providing students with multiple attempts to demonstrate mastery of each learning objective allows for greater flexibility in time to reach mastery and alleviates students’ test anxiety. Additionally, this alternative form of grading can reinforce growth-mindset skills and self-efficacy by helping students realize that success relies on persistence and strategic effort, rather than on innate abilities and/or intelligence. Furthermore, recent evidence supports the idea that pre-exam reflections can improve students’ study strategies, course performance, and development of growth mindsets in college first-year chemistry course (Fink et al., 2018).

Another way instructors can provide students with opportunities to encounter challenges and practice a growth mindset is by integrating course-based undergraduate research experience (CUREs) into the lab setting. In CUREs, students engage in challenge-based learning by performing authentic research that is intrinsically fraught with setbacks. Gin et al. (2018) showed that a CURE explicitly designed to embed failure and practice into the laboratory resulted in students’ increased ability to go through scientific challenges compared to students who participated in CUREs that did not embed failures. In addition, students who failed developed more resilience and understanding that setbacks during the learning experience were normal and should be viewed as learning opportunities instead of “failures”.

Many student comments highlighted a key strength of the GMMs was learning from real people and hearing about how they overcame personal academic challenges through watching the pre-recorded interview videos. Students’ comments revealed they felt “validated”, “less alone”, and “can easily relate” when they heard about the academic challenges that students and professor went through. Limeri et al. (2020b) also found that students’ beliefs about intelligence were influenced by observing and hearing stories of their peers overcoming academic challenges and recognizing that it was possible for them to overcome them as well. Therefore, future mindset interventions could emphasize on this social element by having peers share personal stories of how they overcame challenges and encourage students to view the ability to overcome challenges in learning chemistry through a growth-oriented approach that encompasses developing the correct type of study strategies, consulting the right resources, and practicing strategic effort.

Additionally, several student comments highlighted they would like to learn more about the integration of neuroplasticity, cognition, and nature of science in addition to what is already covered in the GMMs. Limeri et al. (2020b) found that another factor that influenced students’ beliefs about intelligence was reasoning from scientific principles, such as teaching students about brain plasticity. To effectively engage the college-level student audience who may have a more sophisticated knowledge of brain plasticity, it may be necessary to include more details about the neurobiology of learning and emphasize on the physical changes that occur in the brain for future mindset interventions.

Finally, instructors’ beliefs about their students’ intelligence and ability are equally important as it plays a key role in students’ academic achievement. When students believe that their instructors believe in their ability to learn and overcome challenges (i.e. have a growth mindset), they are more inspired to perform and achieve more (Canning et al., 2019; LaCosse et al., 2020; Muenks et al., 2020). This implies how and what we choose to communicate to students in the classroom is highly important as it shapes their motivation and achievement. If instructors can create more opportunities to cultivate growth mindset cultures in their classes, this could potentially inspire more students to pursue STEM fields.

Limitations

First, it is important to acknowledge the small sample size in our study. One explanation for why the sample size is small is because students who had prior exposure to a very similar version of GMMs in past semesters were filtered out to avoid a ceiling effect in the results. Another explanation for why the sample size is small is due to COVID-19 pandemic which forced classes to be online and this caused communication issues between instructor and students. In spring of 2022, students were forced to attend classes virtually for the first two weeks of the semester due to the spreading of COVID-19 new variant, before switching back to in-person mode. This switch also contributed to low participation rates at the beginning of spring 2022 semester. With only 20–37% of enrolled students participating (i.e., completing all surveys), the results of this study should be interpreted judiciously as it may limit the generalizability of the study's findings.

Second, it is important to note that the results reported herewith in are limited to a single HSI in the United States and cannot be generalizable to other institutions and settings. The data collected in this study were from a diverse sample of students across several sections and semesters of general chemistry lecture courses. This diverse sample, representing a wide spectrum of perspectives and experiences, can provide additional insights into the views of the target population.

Third, it is important to acknowledge the limitation of self-reported data used in surveys, such as social desirability and acquiescence bias. Whether students provided responses in which they truly developed and held throughout the study period is unknown. Future work can include follow-up interviews with students to further probe their views on mindset beliefs which may provide more insight on such limitations.

Fourth, it's important to acknowledge that self-selection bias could affect the interpretation of results. Prior to the intervention, study participants and non-study participants were not compared on incoming metrics, such as self-efficacy, motivation, or prior knowledge. Therefore, it is unknown whether students who self-selected to participate in the study had similar background characteristics compared to non-study participants. To assess if the two groups are equivalent or not, a prior knowledge test and or survey can be distributed prior to the intervention. Additionally, future mindset interventions can consider using a randomized controlled trial (RCT) study design to test the true effect of the intervention by minimizing the impact of confounding variables (e.g., student motivation).

Fifth, students participated in GMMs early in the semester before taking any high-stakes assessments. The purpose of having students participate in GMMs in the first three weeks of the semester was to intervene early and have students learn the importance of cultivating a growth mindset before they encounter challenges, such as taking high-stakes exams. Therefore, most of the student survey data and reflection data collected occurred before exam 1. It is important to note if the GMMs were implemented later in the semester, after exams, the results reported herewith in may differ.

Finally, we acknowledge the limitations of using the term “chemistry intelligence” to replace “intelligence” in the modified DMI. While Costa and Faria (2018), Gunderson et al. (2017), and Shively and Ryan (2013) found that the mindset scale is more accurate and predictive of academic outcomes when items are context-specific to the course being studied, Santos et al. (2021) and Limeri et al. (2020a) found that modifying mindset instrument with the addition of the discipline name, “chemistry intelligence” led to a more growth-mindset centered distribution compared to other STEM-specific studies. These findings imply that students interpret general intelligence and chemistry intelligence in a broad range of ways, which may lead to potential response process concerns. One way to improve measurements on chemistry intelligence is to incorporate more specific terms which align with common perceptions of “intelligence” and “ability” in chemistry. These terms include “knowledge”, “ability to apply”, and “understanding” on certain chemistry-related tasks (Santos et al., 2021). Future work can include collecting mindset data using the CheMI instrument and comparing it to mindset data collected using the modified DMI instrument to identify if there are any differences in the way students are interpreting “chemistry intelligence”.

Author contributions

TSN: data collection, data analysis, writing; JH: data analysis; UUO: data analysis; SVG: data analysis (confirmatory factor analysis); JYKC: conceptualization, methodology, oversight of data collection and analysis, writing.

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

The authors would like to thank all the students for sharing their perspectives and time and course instructors (Drs Joya Cooley, Allyson Fry-Petit, Barbara Gonzalez, Paula Hudson, Sissi Li, Macy Shen, Sachel Villafañe) for facilitating this study. We would also like to thank James Park, Natalie Tran, Geon Hwang, and Joachim Kavalakatt for helping out with the interrater reliability tests.

References

1. Aronson J., Fried C. B. and Good C., (2002), Reducing the effects of stereotype threat on African American college students by shaping theories of intelligence, J. Exp. Soc. Psychol., 38(2), 113–125 DOI:10.1006/jesp.2001.1491.
2. Bandura A., (1977), Self-efficacy: Toward a unifying theory of behavioral change, Psychol. Rev., 84(2), 191–215 DOI:10.1037//0033-295x.84.2.191.
3. Bentler P., (1990), Comparative fit indexes in structural models, Psychol. Bull., 107(2),238–246 DOI:10.1037/0033-2909.107.2.238.
4. Blackwell L. S., Trzesniewski K. H. and Dweck C. S., (2007), Implicit theories of intelligence predict achievement across an adolescent transition: A longitudinal study and an intervention, Child Dev., 78(1), 246–263 DOI:10.1111/j.1467-8624.2007.00995.x.
5. Broda M., Yun J., Schneider B., Yeager D. S., Walton G. M. and Diemer M., (2018), Reducing Inequality in Academic Success for Incoming College Students: A Randomized Trial of Growth Mindset and Belonging Interventions, J. Res. Educ. Eff., 11(3), 317–338 DOI:10.1080/19345747.2018.1429037.
6. Boz Y., Yerdelen-Damar S., Aydemir N. and Aydemir M., (2016), Investigating the relationships among students’ self-efficacy beliefs, their perceptions of classroom learning environment, gender, and chemistry achievement through structural equation modeling, Res. Sci. Technol. Educ., 34(3), 307–324 DOI:10.1080/02635143.2016.1174931.
7. Bui P., Pongsakdi N., McMullen J., Lehtinen E., Hannula-Sormunen M. M., (2023), A systematic review of mindset interventions in mathematics classrooms: What works and what does not? Educ. Res. Rev., 40, 1–31 DOI:10.1016/j.edurev.2023.100554.
8. Burnette J. L., Hoyt C. L., Russell V. M., Lawson B., Dweck C. S. and Finkel E., (2020), A growth mind-set intervention improves interest but not academic performance in the field of computer science, Soc. Psychol. Pers. Sci., 11(1), 107–116 DOI:10.1177/1948550619841631.
9. Campbell A. L., Direito I. & Mokhithi M., (2021), Developing growth mindsets in engineering students: a systematic literature review of interventions, Eur. J. Eng. Educ., 46(4), 503–527 DOI:10.1080/03043797.2021.1903835.
10. Canning E. A., Muenks K., Green D. J. and Murphy M. C., (2019), STEM faculty who believe ability is fixed have larger racial achievement gaps and inspire less student motivation in their classes, Sci. Adv., 5, 1–7 DOI:10.1126/sciadv.aau4734.
11. Chen F. F., (2007), Sensitivity of goodness of fit indexes to lack of measurement invariance, Struct. Equation Modeling, 14(3), 464–504 DOI:10.1080/10705510701301834.
12. Chiu C., Hong Y. and Dweck C. S., (1997), Lay dispositions and implicit theories of personality, J. Pers. Soc. Psychol., 73(1), 19–30 DOI:10.1037/0022-3514.73.1.19.
13. Cohen J., (1988), Statistical power analysis for the behavioral sciences, 2nd edn, L. Erlbaum Associates.
14. Costa A. & Faria L., (2018), Implicit theories of intelligence and academic achievement: A meta-analytic review, Front. Psychol., 9(829), 1–16 DOI:10.3389/fpsyg.2018.00829.
15. Dai T. and Cromley J. G., (2014), Changes in implicit theories of ability in biology and dropout from STEM majors: A latent growth curve approach, Contemp. Educ. Psychol., 39(3), 233–247 DOI:10.1016/j.cedpsych.2014.06.003.
16. DeSimone J. A. Harms P. D. and Desimone A. J., (2015), Best practice recommendations for data screening, J. Organ. Behav., 36(2), 171–181 DOI:10.1002/job.1962.
17. Dweck C. S., (2000), Self-Theories: Their Role in Motivation, Personality, and Development, Philadelphia, USA: Psychology Press.
18. Dweck C. S., (2006), Mindset: The new psychology of success, 1st edn, Random House.
19. Dweck C. S. and Elliott E. S., (1983), Achievement motivation, Handb. Child Psychol., 4, 643–691.
20. Dweck C. S. and Leggett E. L., (1988), A social-cognitive approach to motivation and personality, Psychol. Rev., 95(2), 256–273 DOI:10.1037/0033-295x.95.2.256.
21. Dweck C. S., Chiu C. and Hong Y., (1995), Implicit theories and their role in judgments and reactions: A world from two perspectives, Psychol. Inquiry, 6(4), 267–285 DOI:10.1207/s15327965pli0604_1.
22. Ferrell B. and Barbera J., (2015), Analysis of students’ self-efficacy, interest, and effort beliefs in general chemistry, Chem. Educ. Res. Pract., 16, 318–337 10.1039/C4RP00152D.
23. 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 10.1039/C7RP00244K.
24. Gin L. E.; Rowland A. A.; Steinwand B.; Bruno J.; Corwin L. A., (2018), Students who fail to achieve predefined research goals may still experience many positive outcomes as a result of CURE participation, CBE – Life Sciences Education, 17(4), a57 DOI:10.1187/cbe.18-03-0036.
25. Glaser B. G. and Strauss A. L., (1999), The discovery of grounded theory: Strategies for qualitative research, Aldine Transaction.
26. Good C., Aronson J. and Inzlicht M., (2003), Improving adolescents’ standardized test performance: An intervention to reduce the effects of stereotype threat, J. Appl. Dev. Psychol., 24, 645–662 DOI:10.1016/j.appdev.2003.09.002.
27. Gunderson E. A., Hamdan N., Sorhagen N. S. and D’Esterre A. P., (2017), Who needs innate ability to succeed in math and literacy? Academic-domain-specific theories of intelligence about peers versus adults, Dev. Psychol., 53(6), 1188–1205 DOI:10.1037/dev0000282.
28. Harris R. B., Mack M. R., Bryant J., Theobald E. J. and Freeman S., (2020), Reducing achievement gaps in undergraduate general chemistry could lift underrepresented students into a “hyperpersistent zone”, Sci. Adv., 6(24) DOI:10.1126/sciadv.aaz5687.
29. Higgins E. T. & Rholes W. S. (1978) “Saying is believing”: effects of message modification on memory and liking for the person described, J. Exp. Soc. Psychol., 14(4), 363–378 DOI:10.1016/0022-1031(78)90032-X.
30. Hong Y.-y, Chiu C.-y, Dweck C. S., Lin D. M.-S. and Wan W., (1999), Implicit theories, attributions, and coping: A meaning system approach, J. Pers. Soc. Psychol., 77(3), 588–599 DOI:10.1037/0022-3514.77.3.588.
31. Hu L. T. and Bentler P. M., (1999), Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives, Struct. Equation Modeling, 6(1), 1–55 DOI:10.1080/10705519909540118.
32. Kan A. and Akbas A., (2006), Affective factors that influence chemistry achievement (attitude and self efficacy) and the power of these factors to predict chemistry achievement, J. Turk. Sci. Educ., 3(1), 30.
33. Komarraju M. and Nadler D., (2013), Self-efficacy and academic achievement: Why do implicit beliefs, goals, and effort regulation matter? Learn. Individ. Differ., 25, 67–72 DOI:10.1016/j.lindif.2013.01.005.
34. LaCosse J., Murphy M. C., Garcia J. A. and Zirkel S., (2020), The role of STEM professors’ mindset beliefs on students’ anticipated psychological experiences and course interest, J. Educ. Psychol., 113(5), 949–971 DOI:10.1037/edu0000620.
35. Lalich I. J., Taylor M. J. and Pribyl J. R., (2006), Identification of the correlation between student self-efficacy and final course percentage in a general chemistry course, Paper presented at the conference in Minnesota State University, Mankato (April, 2006).
36. Limeri L. B., Choe J., Harper H. G., Martin H. R., Benton A. and Dolan E. L., (2020a), Knowledge or Abilities? How Undergraduates Define Intelligence, CBE – Life Sci. Educ., 19(5), 1–12 DOI:10.1187/cbe.19-09-0169.
37. Limeri L. B., Carter N. T., Choe J., Harper H. G., Martin H. R., Benton A. and Dolan E. L., (2020b), Growing a growth mindset: characterizing how and why undergraduate students’ mindsets change, Int. J. STEM Educ., 7(35), 1–19 DOI:10.1186/s40594-020-00227-2.
38. Malcom S. M. and Feder M. A., (2016), Barriers and opportunities for 2-year and 4-year stem degrees: Systemic change to support students’ diverse pathways, The National Academies Press.
39. McNeish D., An J. and Hancock G. R., (2018), The thorny relation between measurement quality and fit index cutoffs in latent variable models, J. Pers. Assess., 100(1), 43–52 DOI:10.1080/00223891.2017.1281286.
40. Moreno C., Pham D. and Ye L., (2021), Chemistry self-efficacy in lower-division chemistry courses: Changes after a semester of instruction and gaps still remain between student groups, Chem. Educ. Res. Pract., 22(3), 772–785 10.1039/d0rp00345j.
41. Muenks K., Canning E. A., LaCosse J., Green D. J., Zirkel S., Garcia J. A. and Murphy M. C., (2020), Does my professor think my ability can change? Students’ perceptions of their STEM professors’ mindset beliefs predict their psychological vulnerability, engagement, and performance in class, J. Exp. Psychol.: General, 149(11), 2119–2144 DOI:10.1037/xge0000763.
42. Naibert N., Duck K. D., Phillips M. M. and Barbera J., (2021), Multi-institutional Study of Self Efficacy within Flipped Chemistry Courses, J. Chem. Educ., 98(5), 1489–1502 DOI:10.1021/acs.jchemed.0c01361.
43. National Center for Education Statistics, (2019), Status and trends in the education of racial and ethnic groups: Indicator 26 STEM degrees, Retrieved from https://nces.ed.gov/programs/raceindicators/indicator_reg.asp.
44. Nguyen T., (2022), Development and implementation of mindset and metacognitive learning strategies workshops in two first-year chemistry courses, Master's thesis, Zenodo: California State University Fullerton DOI:10.5281/zenodo.7407093.
45. Nilson L. B., Specifications Grading: Restoring Rigor, Motivating Students, and Saving Faculty Time, 47933 Stylus Publishing, Llc, 2014.
46. 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, 742–757 10.1039/D2RP00102K.
47. 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, 905–922 10.1039/D1RP00092F.
48. Shively R. L. and Ryan C. S., (2013), Longitudinal changes in college math students’ implicit theories of intelligence, Soc. Psychol. Educ., 16(2), 241–256 DOI:10.1007/s11218-012-9208-0.
49. 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 DOI:10.1177/095679761773970.
50. Smiley P. A., Buttitta K. V., Chung S. Y., Dubon V. X. and Chang L. K., (2016), Mediation models of implicit theories and achievement goals predict planning and withdrawal after failure, Motiv. Emot., 40(6), 878–894 DOI:10.1007/s11031-016-9575-5.
51. Sriram R., (2014), Rethinking intelligence: The role of mindset in promoting success for academically high-risk students, J. Coll. Stud. Retention: Res., Theory Pract., 15(4), 515–536 DOI:10.2190/CS.15.4.c.
52. Steiger J. H., (2016), Notes on the Steiger-Lind (1980) handout, Struct. Equation Modeling, 23(6), 777–781 DOI:10.1080/10705511.2016.1217487.
53. Stone K., Shaner S. and Fendrick C., (2018), Improving the success of first term general chemistry students at a liberal arts institution, Educ. Sci., 8(1), 5 DOI:10.3390/educsci8010005.
54. Usher E. L., Pajares F., (2008), Sources of self-efficacy in school: Critical review of the literature and future directions, Rev. Educ. Res., 78(4), 751–796 DOI:10.3102/0034654308321.
55. Uzuntiryaki-Kondakci E. and Senay A., (2015), Predicting chemistry achievement through task value, goal orientations, and self-efficacy: a structural model, Croat. J. Educ., 17(3), 725–753 DOI:10.15516/cje.v17i3.1555.
56. Van den Hurk A., Meelissen M. and van Langen A., (2019), Interventions in education to prevent STEM pipeline leakage, Int. J. Sci. Educ., 41(2), 150–164 DOI:10.1080/09500693.2018.1540897.
57. Villafañe S. M., Garcia C. A. and Lewis J. E., (2014), Exploring diverse students’ trends in chemistry self-efficacy throughout a semester of college-level preparatory chemistry, Chem. Educ. Res. Pract., 15, 114–127 10.1039/C3RP00141E.
58. Villafañe S. M., Xu X. and Raker J. R., (2016), Self-efficacy and academic performance in first-semester organic chemistry: testing a model of reciprocal causation, Chem. Educ. Res. Pract., 17, 973–984 10.1039/C6RP00119J.
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 DOI:10.1021/acs.jchemed.0c01497.
60. 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 DOI:10.1080/00461520.2012.722805.
61. Yeager D. S. and Dweck C. S., (2020), What can be learned from growth mindset controversies? Am. Psychol., 75(9), 1269–1284 DOI:10.1037/amp0000794.
62. Yeager D. S. and Walton G. A., (2011), Social-Psychological Interventions in Education: They’re Not Magic, Rev. Educ. Res., 81(2), 267–301 DOI:10.3102/00346543114059.
63. Yeager D. S., Hanselman P., Walton G. M., Murray J. S., Crosnoe R., Muller C., Tipton E., Schneider B., Hulleman C. S., Hinojosa C. P., Paunesku D., Romero C., Flint K., Roberts A., Trott J., Iachan R., Buontempo J., Yang S. M., Carvalho C. M. and Dweck C. S., (2019), A national experiment reveals where a growth mindset improves achievement, Nature, 573(7774), 364–369 DOI:10.1038/s41586-019-1466-y.
64. Zhao H., Xiong J., Zhang Z. and Qi C., (2021), Growth mindset and college students’ learning engagement during the COVID-19 pandemic: A serial mediation model, Front. Psychol., 12, 621094–621094 DOI:10.3389/fpsyg.2021.621094.
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 DOI:10.1016/B978-012109890-2/50031-7.
66. Zusho A., Pintrich P. R. and Coppola B., (2003), Skill and will: the role of motivation and cognition in the learning of college chemistry, Int. J. Sci. Educ., 25(9), 1081–1094 DOI:10.1080/0950069032000052207.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3rp00352c

---