Is Support in the Anxiety of the Beholder? How Anxiety Interacts with Perceptions of Instructor Support in Introductory Biology Classes

INTRODUCTION

Undergraduate students experience a variety of emotions as part of their collegiate experience, including enjoyment, pride, frustration, and boredom, to name a few (Pekrun et al., 2023). Of these, anxiety is one of the most common (Pekrun et al., 2002), and its prevalence among undergraduates has increased in recent years. In national surveys of over 35,000 students at more than 50 institutions, 29% of undergraduate students in 2019 and 34% in 2022 reported being treated or diagnosed with anxiety in the previous year (American College Health Association [ACHA], 2019; 2022). While moderate anxiety levels can be a positive motivator (Yerkes and Dodson, 1908), high levels of anxiety tend to have negative impacts on cognition (Grossberg, 2009; Gonzalez et al., 2019), course engagement (Pekrun and Linnenbrink-Garcia, 2012), and performance (Kim and Pekrun, 2014; Mainhard et al., 2018; Pekrun, 1992, 2006). Thus, the relationship between student anxiety levels and student success is an important area of investigation, particularly in undergraduate science, technology, engineering, and mathematics (STEM) courses with traditionally high rates of attrition (PCAST, 2012).

Because anxiety levels differ among students, anxiety has been investigated as one factor contributing to differential student outcomes in undergraduate science classrooms. Students who have higher anxiety are less likely to persist in their science majors (Respondek et al., 2017; England et al., 2019). Many factors contribute to the variation in levels of student anxiety. At the student level, certain demographic factors, such as gender identity, generation status, and year of study, have been found to predict levels of anxiety (England et al., 2019; Gaudier-Diaz et al., 2019) and these differential anxiety levels interact with class events such as exams or active learning practices (England et al., 2017; Cooper et al., 2018; Harris et al., 2019; Cooper and Brownell, 2020; Mohammed et al., 2021). Within an individual class, student anxiety levels have a bell curve distribution (England et al., 2017), but the mean level of student anxiety also varies among classes, resulting in larger numbers of students with high anxiety in some classes versus others (England et al., 2017; Schussler et al., 2021). Therefore, one potential factor contributing to the variation in student anxiety is instructors and their practices.

Instructors play a vital role in the generation of student emotion (Mainhard et al., 2018), putting them in a position to reduce the effects of anxiety on student success in their courses (Pekrun, 1992; 2006). A 2018 meta-analysis of 65 studies indicated a strong correlation between instructor support and positive academic emotions such as enjoyment, interest, hope, and pride (Lei et al., 2018). However, the opposite can also be true, with a lack of instructor immediacy (i.e., perception of approachability), clarity, and communication competence being associated with negative student emotions such as anger, anxiety, boredom, shame, and hopelessness (Mazer et al., 2014; Goetz et al., 2019). At our institution, instructors who are viewed by students as providing a higher level of support have classes with lower mean levels of anxiety (Schussler et al., 2021). Therefore, understanding the link between specific instructor practices and student anxiety in individual classes may help identify ways to reduce disparities in the student anxiety experience and thus remove one factor contributing to differentials in student success.

Theoretical Framework

Our work on anxiety is theoretically guided by the control-value theory of Achievement Emotion outlined by Pekrun (2006; herein, control-value theory). This theory posits that students’ emotional outcomes to achievement-related tasks (e.g., coursework or exams) are proximally generated by students’ personal appraisals of the value of the tasks that are being asked of them and their perceived control over the achievement of that task. Specific emotions arise from different appraisals. For example, anxiety generally arises when students appraise task value as high but are uncertain about their level of control to achieve the task. These control and value appraisals are distally guided by contextual features such as the classroom environment or prior student experiences (Pekrun et al., 2007). Emotions can be moderated by emotion regulating processes such as coping skills (Gross and Jazaieri, 2014; Gross, 2015). Control-value theory broadly describes how student emotion is generated and processed; however, it is important to further study context-specific applications of the theory, such as students’ anxiety experiences within introductory biology classrooms.

Our previous work studying student anxiety in introductory biology courses at our institution used control-value theory to propose the Instructor Practices-Student Anxiety (IPSA) framework. IPSA postulates that students view instructor practices as cues to inform their perceptions of instructor support, which inform students’ appraisals of control and value that generate anxiety levels (Figure 1; Schussler et al., 2021). We suggest that students are likely to have higher anxiety when they are unsure about how supportive their instructor is, leading to uncertain appraisals of their ability to control their task performance (Pekrun, 2006; Goetz et al., 2019). It also follows from the control-value theory that these relationships between emotion and appraisals are iterative and continue to influence each other over time. For example, a student who feels confused in class may have increased anxiety levels that in turn may increase uncertainty of control.

Schussler et al. (2021) confirmed our hypothesis that class-level anxiety was inversely related to class-level perception of instructor support. To identify instructor practices that might relate to anxiety, we asked students to explain why they rated instructor support the way they did. We found that interactions with students (relational cues), instrumental support they provided to students (instrumental cues), pedagogical practices (pedagogical cues), instructor personality, and uncertainty were mentioned by students as impacting their support ratings. Thus, we conceptualize “instructor practices” broadly in IPSA as anything the instructor does that impacts students’ perceptions of support. Importantly, the framework focuses on student perceptions of practices and not reality; what drives students’ appraisals and emotional outcomes is how supportive students perceive instructor practices to be.

While the IPSA framework is aligned with findings from previous work (e.g., Allen et al., 2006; Reeve, 2009; Seidel et al., 2015; Schussler et al., 2021), there are many aspects that still need testing in introductory biology classes. In this study, we investigated the potential relationship between student perceptions of instructor support and anxiety levels within individual classes and whether those relationships are dynamic across time. Notably, we did not test the potential mediating factors of student appraisals of control and value, nor emotion regulation; these should be explored in future work.

Rationale for Current Study

The IPSA framework was informed by our previous study, which used data collected from six introductory biology classes to identify a significant, negative relationship between levels of student anxiety and perceptions of instructor support (Schussler et al., 2021). However, this analysis combined data across all six classes and did not analyze the relationship between student anxiety levels and perceptions of instructor support within individual classes. Exploring IPSA at the level of an individual class is critical to understand how student emotional experiences differ when controlling for instructor practices; disparities in emotional experiences among students in the same class would not be a result of differential instructor practices, instead, it would be an outcome of how students differentially perceive those practices. Therefore, to better understand the role of differential appraisals of instructor practices in the IPSA framework, this study examined student anxiety levels and perceptions of instructor support within three individual classes and investigated the relationship between these two variables by class. As with our previous study, this study collected qualitative data to understand which aspects of instructor practice students perceived as supportive, allowing us to ascertain whether students with relatively higher versus lower anxiety in a single class may perceive the same instructor practices differently.

Control-value theory posits that emotions and their drivers are dynamic and situational, interacting and varying with each other across time (Pekrun, 2006; Goetz et al., 2013). However, our previous study only sampled student anxiety levels and instructor support perceptions at one timepoint, so we could not test whether these relationships change over time. Extant course experiences (such as instructor practices) and emotions become part of the contextual antecedents that inform future control and value appraisals. Changes in student appraisals could range from no change, to slow shifts over time, to “turning point events” where negative or positive encounters between students and instructors change students’ perceptions of the cognitive, affective, and motivational aspects of the class (Docan-Morgan and Manusov, 2009). Research suggests that changes in perceptions of instructor support over the course of a semester, such as perceptions of instructor immediacy, autonomy support, and instructor-student rapport, are common (Frymier, 1993; Black and Deci, 2000; Jang et al., 2016; Lammers et al, 2017). Thus, we would anticipate that the relationship between perceptions of instructor support and anxiety levels could iteratively change throughout the semester. To investigate this aspect of the framework, the current study collected perceptions of instructor support and student anxiety levels at two timepoints (wk 4 and 14 of the semester).

Objectives of Current Study

Based on previous work at our institution, we know that some instructor practices are related to students’ perception of instructor support, and these perceptions are related to student anxiety at the level of a class (England et al., 2017; Schussler et al., 2021). However, despite instructors using the same practices for all students, we know that student anxiety levels vary within an individual class. We suggest that anxiety may be a factor in the differential perceptions of instructor practices, and that these perceptions vary over time (Hoyt et al., 2021; Weatherton et al., 2024). Thus, we see a need to investigate the relationship between student anxiety and perceptions of instructor support within a class and over time. This study surveyed students in three introductory biology classes to investigate how students with relatively higher and lower anxiety levels in each class perceived different aspects of instructor support at two timepoints during the semester, and then compared these results among classes. Our study sought to refine the IPSA framework and add to our understanding of the types of instructional practices that should be investigated in relation to anxiety outcomes. Thus, this study asked two research questions:

1. What is the relationship between student perceptions of instructor support and student anxiety levels over time in each class?

2. How do students with relatively higher and lower anxiety levels in each class describe the support of their instructor?

In exploring the dynamic between instructor practices and student emotion in individual classes, we are not seeking to test causal relationships, but instead hoping to understand classroom conditions that may decrease disparities in student anxiety levels and, thus, facilitate more positive outcomes for all students.

MATERIALS AND METHODS

Study Context

This study was conducted in the fall of 2021 at a large research university in the southeastern United States. The study was approved by the Institutional Review Board (IRB #16-03181-XP). We surveyed students in three large (∼220-student) introductory biology classes, two of which were 3-credit cellular biology courses (classes 1 and 2) and one of which was a 3-credit organismal biology course (class 3). These represent two of the three introductory biology courses taken by biology and other science majors at this institution. Organismal biology is typically taken first in the series, and cellular biology second. Each class was taught in an in-person classroom setting by a single instructor; all of the instructors had taught the class multiple times before. Instructors 1 and 2 were both White women and had both taught the class for more than 10 years at this institution. Instructor 3 was a White man and had taught the class for more than 15 years at this institution.

Data Collection

Data were collected via an online survey distributed by instructors to their students at wk 4 and 14 of the semester. We chose to survey students at these timepoints because wk 4 of the semester is generally after students have taken their first assessment and have had time to make judgments about the class and their instructors. To explore how these perceptions may have changed over the course of the semester, we surveyed students again at wk 14, which is usually the last full week of instruction before final exams.

The instructors sent IRB-approved email verbiage along with a survey link to students; the link took students to a consent form followed by the survey text. Instructors could offer up to 5 points to their students for completing the surveys, though this was at the discretion of the instructor and made up a very small part of the overall 1,000 points of the courses. Students could take the survey to earn points but opt out of the study. The survey collected data related to students’ perceived anxiety in relation to the lecture class, students’ perceptions of instructor supportiveness, students’ demographic information, and intention to persist in their major.

Students’ overall anxiety level in lecture was assessed via four survey items: “Biology lecture makes me nervous,” “Biology lecture is stressful,” “Biology lecture makes me anxious,” and “Biology lecture is scary.” Students provided their answers on a Likert scale from 1 to 7, where 1 was low anxiety and 7 was high anxiety. These items have been used multiple times on the introductory biology student population at this institution. Validity evidence of these items was published in England et al. (2019); however, given that this study was conducted post-COVID we re-evaluated them using the wk 4 survey data. Confirmatory factor analysis returned comparative fit index (CFI) and Tucker–Lewis index (TLI) values that indicated reasonable fit (CFI: 0.97, TLI: 0.91), where reasonable fit is defined as a CFI and TLI > 0.90 (Hu and Bentler, 1999; Shi et al., 2019). Furthermore, the standardized root mean squared error (SRMR) value indicated reasonable fit (0.03), where reasonable fit is defined as values ≦ 0.05 (Shi et al., 2019). Root mean square error of approximation (RMSEA) scores indicate good fit when lower than 0.08; our RMSEA score was outside of the ideal range (0.27), however, RMSEA is notably sensitive to small sample size (Hu and Bentler, 1999; Savalei, 2012; Shi et al., 2019). Calculating Chronbach's alpha for the data indicated high internal consistency (α = 0.94), well above the accepted cutoff (α = 0.70) (Nunnally, 1978).

To collect data on students’ perception of their instructor's support, students were first asked “How supportive is the instructor of your biology class?” Students responded on a scale from 1 (not supportive) to 10 (very supportive). Following this, students were asked two open-ended questions: “What specific things has your instructor said or done that led to the support rating you gave them?” and “Besides specific things your instructor said or did, what else informed your rating of their supportiveness?” Previous iterations of these questions asked students to explain “why you rated your instructor's support the way you did” (Schussler et al., 2021); these new variations were designed to elicit more specific instances of instructor behavior, actions, or verbal utterances that gave rise to student support ratings. We gathered validity evidence for these new questions through expert review, examining the relationship between the responses and participants’ support ratings (i.e., evaluating concurrent validity), and comparing the support characteristics generated by these questions with the ones identified from the previous study (American Education Research Association, American Psychological Association, and National Council for Measurement in Education, 2014).

Finally, the survey asked students to self-identify the following demographic variables: year in school, gender identity, racial identity and ethnicity, their major, and their instructor's name. We also asked students how many years it had been since they last took a biology course (exclusive of their current course). From the institution, we requested class grade point average (GPA) as a measure of overall class final grades. We collected students’ university emails and names so we could match participants’ responses at wk 4 and 14. Only matched respondents were included in the analyses.

Data Analysis

All quantitative analyses and visualizations were created in RStudio software using the tidyverse and stats packages (Wickham et al., 2019; RStudio Team, 2020).

What is the Relationship Between Student Perceptions of Instructor Support and Student Anxiety Levels over Time within Individual Classes?

To answer this question, we calculated the mean anxiety level and perception of instructor support for each class. To examine the relationship between these two variables, we calculated the Pearson's R correlation value at both wk 4 and 14 for each class (Chao, 2017).

How do Students with Relatively Higher and Lower Anxiety in Each Class Describe the Support of Their Instructor?

To answer this question, we first had to categorize how participants described instructor support by using the student open-ended questions where they explained their support ratings. After this qualitative analysis was complete, we analyzed the differences in instructor support codes between students who had relatively higher versus lower anxiety in each class at wk 4 and 14.

Consensual Qualitative Coding.

To identify categories of student instructor support perceptions, we followed the theoretical and methodological framework of Hill and colleagues (1997) called “consensual qualitative research,” which engages groups of researchers in data analyses to reach a consensus about the codes. We had a team of six coders for this process (MW, ES, JB, HF, IB, BE). Iterative analysis of the data is a hallmark of this method, with multiple rounds being employed by one group of researchers until they feel they have adequately represented the meaning of the data, and then a second team auditing the final codes. This method relies on group consensus versus calculating an interrater agreement to assert the value of free thinking within the team versus a focus on getting a code “right.” Throughout our description, we will use “theme” to refer to a more encompassing idea being expressed in the dataset, and codes as more specific ideas within those broader themes.

We started our analysis using the codebook from our previous work on student perceptions of instructor support (Schussler et al., 2021), but also employed an inductive process to identify any new codes for two reasons: 1) the question prompts were slightly different and 2) the previous codes were generated from classes taught pre-COVID, and we wanted to account for the fact that student perceptions of instructor support may have changed over the intervening period. Thus, code generation for this study was both deductive and inductive.

To create the codebook for this study, we used subsets of the student responses (from both open-ended questions) and a coding team to generate and test codes until consensus was reached. Two coders (JB and BE) who had been a part of the previous support coding independently examined 150 student responses (out of 280 matched wk 4 and 14 student responses) and then met to discuss modification of the former codebook using the new dataset. They used 150 student responses to ensure enough data to reach saturation (Saldaña, 2009). They maintained the former codebook themes of relational, instrumental, pedagogical, but folded the former theme of instructor personality into a code within the relational theme. Surprisingly, there were few instances of the former “uncertain” theme; the themes “neutral” and “other” were the final two themes. The new open-ended questions and a desire for more details about support perceptions led to the creation of new codes within the relational, instrumental, and pedagogical themes. Each code was a distinct variation within the broader theme that gave more insight into the specific student-perceived instructor actions or behaviors that they said were supportive or not supportive within the theme. In particular, students were mentioning more types of instrumental support, and more negative perceptions of instructor support, than seen in the pre-COVID dataset. JB and BE made initial recommendations regarding these codes for the rest of the coding team. As in our previous codebook, codes could be assigned as positive (supportive), negative (not supportive) or neutral (could not tell) based on the student comment.

To test the new codebook, the full coding team met to create a plan. MW, ES, HF, and IB first tested the initial theme and code ideas from JB and BE by reviewing 100 student responses, 50% of which overlapped with another coder's assigned responses. Each researcher independently coded the data using the suggested codebook and made notes about items they were confused about or codes that they thought needed to be added or clarified. Pairs with the same student responses met to discuss their subset, and then the larger group reconvened to discuss and make changes to the codebook. The group agreed they were ready to code when agreement between coders was about 70%. The final codebook can be found in Supplemental Table S1 (Supplemental Materials).

To conduct the coding, MW, ES, HF, and IB each received a new set of 140 student responses. As before, 50% of every coders’ responses overlapped with another coders’, so that each student response was coded by two independent researchers. Researchers assigned as many codes to a student's response as necessary to reflect their perceptions, with each code counting once for that student regardless of how many times they mentioned the same idea. Upon completion, ES received all codes and identified disagreements. JB and BE then reconciled disagreements at the level of theme and moved two codes from the pedagogy theme to the relational theme with the permission of the coding team. ES made final decisions on all codes. The codes from both support perception questions at wk 4 or 14 were merged to provide a single set of codes regarding a student's perceptions of their instructor's support at each time period.

Support Perceptions of Students with Higher and Lower Anxiety.

To distinguish between students with relatively higher versus lower anxiety in each class, we grouped participants by whether they were above or below the median anxiety score at wk 4 and 14 in each class. This allowed us to investigate how perceptions of instructor support might differ across a range of anxieties in one class, versus setting a single value on a scale from 1 to 7 as the cutoff for “high anxiety.” This is consistent with our previous research in these classes indicating the negative impacts of higher anxiety, not necessarily anxiety above a certain threshold value (England et al., 2019; Weatherton et al., 2024). Calculating these higher- and lower-anxiety groups separately for wk 4 and 14 also recognizes that some students might move between the higher- and lower-anxiety groups over time.

Once we created these higher- and lower-anxiety groups for each class at each time period, we tallied the frequency of each support code for each group and calculated the relative prevalence of each code by dividing the frequency by the total number of students in each group. Because student responses could only be coded with each support code once, this relative prevalence calculation is equivalent to the number of students who reported a particular code divided by the number of students in the respective anxiety group. To start to answer our research question about how students with relatively higher and lower anxiety in each class differentially viewed their instructor's support, we identified the three codes with the highest differences in prevalence between these groups for each instructor at each time period.

To identify significant differences in the perceptions of students with higher and lower anxiety in each class, we used two-proportion tests. Two-proportion tests are used to determine whether there is a significant difference in a binary outcome between two independent groups (Chen et al., 2000). Specifically, we used these tests to compare code prevalence between relatively higher- and lower-anxiety groups in each class at each time period. Thus, there was the potential to run up to 240 tests (i.e., 40 codes * 3 classes * 2 timepoints), although in practice this number was lower because some codes were not present in a class or time period. We used Benjamini and Hochberg corrections to correct for multiple comparisons, and solely report corrected p-values in our results (Thissen et al., 2002).

RESULTS

Participants

There were 558 students who completed the wk 4 survey and 489 students who completed the wk 14 survey. After matching presurvey and postsurvey respondents, we had a final dataset of 280 participants. Of those, 204 came from the two cellular biology classes (N = 94; class 1 and N = 110; class 2), and 76 from the organismal class (N = 76; class 3).

Our sample was predominantly White (75.36%), second-year (53.93%), women (75%) who had taken another biology class within the past year (52.5%). We report the demographic composition of our overall sample, as well as the composition within each class (Table 1). The demographic composition of each class was relatively similar, with a notable exception that class 3, the organismal biology class, had many more first-year students who had not taken a biology class in the past year, compared with classes 1 and 2, the cellular biology classes (Table 1). This is because most first-year students at this institution start their biology course series with the organismal class.

We used institutional and survey data to characterize achievement and potential attrition from the major for each class in this study. In fall 2021, classes 1 and 3 had very similar class GPAs (2.50 and 2.52, respectively), while class 2 had a class GPA of 3.38. Classes 1 and 3 also had higher intended attrition from the biology major by wk 14. In class 1, 35.1% of the students indicated they were biology majors, with 10.6% of those indicating they were intending to leave the major (a potential attrition rate of 30%). In class 3, 13% of the 39.5% biology majors in the class said they were intending to leave the major (33% attrition). In comparison, only 4.5% of the 31.8% biology majors in class 2 indicated they were intending to leave the major by wk 14 (14% attrition).

What is the Relationship Between Student Perceptions of Instructor Support and Student Anxiety Levels over Time within Individual Classes?

Across all participants, the mean level of anxiety at wk 4 was 3.54 (where 7 indicates high anxiety). Class 3 had the highest mean level of anxiety, while class 2 had the lowest (Figure 2; Supplemental Table S2, Supplemental Materials). Over the course of the semester, the mean level of anxiety increased in classes 1 and 3, while the mean level of student anxiety decreased in class 2.

For student perceptions of instructor support, the overall average was 7.56 (out of the highest rating of 10) at wk 4 (Figure 3; Supplemental Table S3, Supplemental Materials). Class 3 had the lowest rating of instruction support at wk 4 (6.41) and class 2 had the highest (9.09). All three classes had a decrease in average rating of instructor support over the course of the semester, though this change was much smaller in class 2 than in classes 1 and 3.

To explore the relationship between students’ perceptions of instructor support and their level of anxiety, we examined the correlation between the two variables at wk 4 and 14 for each class (Figure 4). At wk 4, all classes experienced a negative, significant correlation between instructor support rating and student anxiety level, where students with higher anxiety tended to rate their instructors as less supportive. For class 1, the R value was −0.25 (p = 0.014); for class 2, the R value was −0.36 (p = 0.0001); for class 3, the R value was −0.3 (p = 0.0087).

At wk 14, the relationship between instructor support and student anxiety in classes 1 and 3 became more negative and more significant compared with wk 4. In class 1, the relationship had an R value of −0.42 (p < 0.001) and in class 3 there was an R value of −0.56 (p < 0.001). In class 2, however, the correlation between instructor support and student anxiety was not statistically significant at wk 14.

How do Students with Relatively Higher and Lower Anxiety in Each Class Describe the Support of Their Instructor?

Student-perceived Instructor Support Themes and Codes.

We created five overall instructor support themes from student responses: relational, instrumental, pedagogical, neutral, and other. Because of their lack of prevalence or utility for the study, we dropped the “neutral” (not knowing how to describe the support of the instructor because of the size of the class or not interacting with the instructor) and “other” (random, uncategorizable comments) themes from further analysis. Each of the remaining three themes had subsets of instructor support characteristics called codes: the relational theme had four codes, the instrumental theme had 10 codes, and the pedagogical theme had six codes (Figure 5). Codes could be positive, negative, or neutral; due to low incidence, we removed neutral codes from further analysis. Thus, there were 20 positive support codes and 20 negative support codes across three themes that we used for analysis.

The relational theme included four codes (A1, A2, A3, A4) that consolidated student expressions about how the instructor supported or did not support them in terms of how they: 1) interacted with the students personally, 2) expressed concern for their success, 3) contributed to the class climate in terms of how they made everyone feel, and 4) expressed who the instructor was as a person to the students.

The instrumental theme captured student comments about something the instructor did for them or provided to them that conveyed support or a lack of support to them. This included codes (B1 through B10) of: 1) holding office hours, 2) answering questions, 3) responding to emails, 4) offering materials to help them succeed, 5) being flexible with due dates, 6) giving them feedback on assignments, 7) encouraging attendance at a supplemental instruction session, 8) providing extra points, 9) sending them updates on items that are due, and 10) offering generic expressions of help. These were all different types of provisioning that the instructor was doing for the students that they perceived as either helping or, through the denial of these things, hurting their success in the course.

The pedagogical theme was used when students mentioned things that were specifically happening in class sessions, such as codes (C1 through C6) of: 1) how instructors were explaining information, 2) instructor ability to recognize student struggle, 3) the pace of the lecture, 4) the organization of the class, 5) the nature of the exams, or 6) other pedagogical practices such as active learning activities. These were specifically about how the instructor was helping them to learn or not learn in the context of the class.

Differences in Perceptions of Instructor Support by Higher and Lower Anxiety Students in Each Class.

The relative prevalence of each of the 20 positive and 20 negative support codes for students with relatively higher and lower anxiety at wk 4 and 14 in each class are shown in Supplemental Materials (Supplemental Tables S4 and S5). To identify qualitative patterns in how student perceptions of instructor support differed between relatively higher- and lower-anxiety students in each class, we highlighted the three codes in each class that had the largest difference in prevalence between higher- and lower-anxiety students in Table 2 (wk 4) and Table 3 (wk 14). At wk 4, positive features of instructor personality (class 1), answering questions (class 3), offering materials (class 1), clarity of explanations (classes 2 and 3), and other pedagogical practices (classes 2 and 3) were mentioned more frequently by relatively lower-anxiety students. However, higher-anxiety students mentioned more positive codes related to instructor answering questions (class 2) and the nature of the exams (class 1). At wk 14, students with relatively lower anxiety were more likely to mention positive perceptions of class climate (classes 2 and 3), responding to emails (class 3), and offering materials (class 1) compared with higher-anxiety students in these classes. Interestingly, at wk 14 lower-anxiety students in classes 1 and 3 mentioned more positive comments about instructor answering questions, while in class 2 it was higher-anxiety students who mentioned more positive comments about their instructor answering questions (Table 3); higher-anxiety students in class 2 also mentioned positive perceptions of pedagogical practices more frequently than lower-anxiety students. Although these trends are interesting and should be further explored, the small sample sizes mean most differences were not statistically different.

Two-proportion tests found three significant differences in relative prevalence of instructor support codes between higher- and lower-anxiety students in any class at either time period. All of these significant differences occurred at week 14. In class 1, lower-anxiety students were significantly more likely to mention positive perceptions of the instructor answering questions than higher-anxiety students (p = 0.004). Lower-anxiety students in class 3 also mentioned more positive perceptions about the instructor answering questions (p = 0.0368). In class 2, higher-anxiety students were significantly more likely to mention positive perceptions of other types of pedagogical practices (e.g., using clickers or worksheets, or other active learning practices) than lower-anxiety students in the class (p = 0.0368). None of the other code prevalences differed significantly between higher- and lower-anxiety students in these classes.

DISCUSSION

This study explored differences in how students with relatively higher versus lower anxiety levels perceived instructor support in three introductory biology classes at two timepoints during a semester. In two of the three classes, there was a significant, negative relationship between students’ anxiety levels and their ratings of instructor support at week 4 and 14. However, in class 2, the negative relationship between student anxiety and perceived instructor support present at week 4 was not significant at week 14, indicating that higher- and lower-anxiety students in this class became more similar in their perceptions of this instructor's support. In exploring student explanations of instructor support characteristics between students with higher and lower anxiety in each class, we found interclass differences between classes 1 and 3 and class 2 in two instructional aspects: answering questions and other pedagogical practices. Given these findings, we suggest additional research on these instructor practices as a way to identify strategies to equalize the disparate emotional landscape in introductory biology, both between and within classes.

Refining the IPSA Framework

Our IPSA framework (Schussler et al., 2021) posited that student perceptions of supportive instructor practices inform their appraisals of control and value related to class-based achievement tasks, which impacts their anxiety levels (Figure 1). While our previous work highlighted how perceptions of instructor support and class anxiety levels varied among different instructors, this study showed that student perceptions of instructor support vary even for a single class where the instructor practices are the same. Furthermore, our results provide empirical support for the generally negative, but malleable relationship between these instructor support perceptions and student anxiety levels at the class level. Using these findings, we modified our IPSA framework, which we detail below, and pose hypotheses that could be investigated to further advance IPSA.

Our modified IPSA framework now delineates actual instructor practices from student perceptions of supportive instructor practices and adds a feedback loop to show that higher and lower anxiety levels are differentially linked with student perceptions of support (Figure 6). The feedback loop also represents that the relationships of these constructs is iterative, given that students with relatively higher and lower anxiety in the same class can, over time, come to see instructor support more similarly, as in class 2. The inclusion of this iterative loop is supported by Pekrun's control-value theory (2006), research linking instructor support perceptions and student emotions (Lei et al., 2018; Mazer et al., 2014), and how time can impact the relationship between student emotions and perceptions of their instructor (Small et al., 1982). Overall, this framework is now better able to reflect how student perceptions and emotions influence each other across and within classes.

Unexplored in this study, but critical to understanding the findings are how anxiety and perceptions of instructor support are related to student appraisals of control, given the central role of control in the generation of anxiety. Control is conceptualized as a student's feeling of agency over achieving success in a task (Pekrun, 2006; Stolz et al., 2020), with uncertainty in control leading to higher anxiety. We draw from Cavanagh et al. (2018), who showed that instructor trust—feelings of understanding, acceptance, care—were related to student course achievement, and from Canning et al. (2024), who showed that instructor growth mindset messages increased student achievement, to suggest that instructor communication is a key area to investigate. Docan-Morgan and Manusov's work (2009) implies that control could also be impacted by specific moments of instructor interaction for some students. It may be that the instructor of class 2 used words, actions, and policies that unambiguously and explicitly messaged support, and this helped all students over time to feel more in control; by contrast, the instructors of class 1 and 3 may have more ambiguously messaged support resulting in differential perceptions of control in the class. The concept of an instructor using the same practices but resulting in different student perceptions is supported by the work of Alvi and colleagues (2020) who documented that individuals with higher anxiety have lower empathetic accuracy (i.e., the ability to infer the intentions of an individual). Their work suggests that students with higher anxiety may be less able to accurately infer the support of an instructor, which may explain why higher-anxiety students in our study described fewer positive aspects of instructor support. We offer these as ideas worthy of future exploration in seeking to understand differential student anxiety in biology classes. Broadly, we follow the lead of Mainhard and colleagues (2018) in suggesting that instructor practices are critical to the emotional climate of a classroom. Our study hints at some specific practices that could be investigated in future studies to examine how practices related to the generation of student anxiety.

Teaching Practices to Study in Relation to Student Anxiety

We found two instructor support codes that significantly differed between higher- and lower-anxiety students at week 14: how the instructor answered their questions (classes 1 and 3), and how the instructor used pedagogical practices to enhance their learning (class 2).

In classes 1 and 3, lower-anxiety students were significantly more likely to mention positive perceptions of how their instructors answered questions, and higher-anxiety students in classes 1 and 3 mentioned more negative comments about this practice. We note the absence of these differences between students with relatively higher and lower anxiety in class 2. Because we did not observe these classes, we do not know what aspect of answering questions drove these differential perceptions. It may be that higher-anxiety students in these classes were less satisfied with the instructor's openness to questions or the clarity of responses, or maybe they desired answers in a different format (in class vs. via email, for example). Micari and Calkins (2021) found that instructor openness to questions was significantly related to higher grades in several large introductory science and social science courses, with students indicating that providing time for asking questions, encouraging question asking and multiple methods for asking questions were related to perceptions of openness. Therefore, it may be valuable to probe question-asking behaviors through surveys such as the scale developed by Walker and Fraser (2005), or the one used by Micari and Calkins (2021) or through class observations and examine any interactions with student anxiety levels and instructor support perceptions.

In class 2, students with relatively higher anxiety were significantly more likely to express positive perceptions of “other pedagogical practices” (e.g., clicker questions, worksheets or other active engagement techniques), compared with lower-anxiety students in the same class. There were no significant differences in the prevalence of this code across student anxiety levels in classes 1 and 3. Thus, it may be that the way the instructor of class 2 used active learning practices in their class was interpreted as especially supportive by students with higher levels of anxiety. This is consistent with the potential advantages of appropriately-enacted active learning for students (Freeman et al., 2014; Theobald et al., 2020), but should be interpreted with caution given reports of active learning increasing anxiety for some students (Cooper et al., 2018; Cooper and Brownell, 2020; Mohammed et al., 2021). We also know from prior work on active learning and anxiety at our institution, that there is wide variation in the levels of anxiety generated from the same active learning practices among instructors (England et al., 2017) and that some types of anxiety can be beneficial to outcomes (England et al., 2019), suggesting that the manner of implementation is paramount (Stains and Vickrey, 2017). We know from prior observations that each of the instructors in this study utilize active learning in their classrooms, but future observational studies are needed to document potential differences in the types of active learning used and their implementation to identify potential practices seen as supportive by students with higher anxiety.

Reflecting on These Ideas in the Classroom

These findings have potential implications for developing and reflecting on teaching practices, although our relatively small sample size and lack of teaching observations constrain our ability to generalize or make definitive suggestions. However, we offer here what we would personally take from our research and apply in our own programs and courses. Many instructors ask for mid-semester feedback from students, and we propose that they could add prompts related to specific support characteristics, like how they answer questions or the types of pedagogical practices they use. Instructors may also wish to use our anxiety and instructor support survey questions to see for themselves how student answers may differ between those with lower and higher anxiety in their classes. For programs that wish to increase degree-level retention, instructors of introductory courses could work together to collect student data and observe one another's teaching to collectively identify teaching behaviors that seem to support all students more equitably (Wilson, 2022). We also suggest that instructors self-reflect on the types of support they provide to students and the extent to which they are explicit in messaging these things to students. Being clear and consistent in messaging and leaving as little for students to infer as possible may be one way to make sure all students feel as supported as possible in our classes.

Limitations and Opportunities for Future Research

A limitation of this study is the lack of class observational data, which prevents us from making inferences about how practices differed between the instructors. However, IPSA and control-value theory suggest that the reality of instructor practice is not the variable of interest, it is how students perceive and appraise those practices. Thus, we take this opportunity to affirm a reality about students and class experiences: how students feel and their perceptions of their reality drive their emotional reactions, and these can impact outcomes. Lammers et al. (2017), for example, found that students who perceived student-instructor rapport as decreasing over the course of a semester had lower final grades than those who perceived rapport as stable or increasing. Overall, the correlative nature of this study makes it impossible to propose causative links between student emotion, student instructor support perceptions, and class outcomes; however, there is an urgent need to investigate these relationships, given the high attrition from science majors (PCAST, 2012).

The timing of our surveys was another factor that potentially limits our inferences. We chose week 4 to measure initial student emotion, assuming that students would need a few weeks to assess their instructor's practices so that their perception of anxiety would more fully reflect their experiences in their current Biology class, as opposed to being principally informed by previous experiences. However, we do not know the timing of how perceptions of a current instructor are integrated into the contextual background that informs student appraisals of instructor practice. Future research could collect data at several early timepoints and include student interviews to determine how students integrate the perceptions of their current instructor with their past experiences in an iterative manner.

We also note that there were several contextual differences between the classes; class 3 had proportionally more first-year students than classes 2 and 1, and class 2 had a higher class GPA and less attrition than classes 1 and 3. These contextual differences among classes and lack of ability to determine causality lead to many potential interpretations. For example, one could attribute lower performance and retention in classes 1 and 3 to instructor practices that were viewed as less supportive and increased student anxiety, which interferes with cognition and motivation (Grossberg, 2009; Kim and Pekrun, 2014; England et al., 2019; Gonzalez et al., 2019). One could attribute the opposite to class 2; however, it may be that class 2 was less difficult and higher grades by wk 14 led to lower anxiety and a more positive perception of instructor support. Furthermore, this could have led to a “ceiling” effect, making it more difficult to discern meaningful differences among student perceptions. Overall, the complexity of the classroom environment and the myriad of variables that may be acting on outcomes means much more work is needed to disentangle causal relationships among variables of interest.

Our work is bounded by a small subsample of students at one institution and our results may not be broadly generalizable to all students; future work should examine the relationship between anxiety, perceptions of support, and instructor practices across many different contexts and institutions. Finally, as with all qualitative studies, our categorizations of instructor support are bounded by the perceptions of our participants, as well as the individuals conducting the analysis.

CONCLUSION

Our work sought to add more nuance to our understanding of the relationship between student anxiety and student perceptions of instructor support in introductory biology courses. We found that students with relatively higher versus lower anxiety in the same class have different perceptions of their instructor's support, and that there may be instructor practices that help close these gaps, such as answering questions or supportive pedagogical practices. These practices may help students feel more in control of their success and thus may lower student anxiety. We suggest additional research on these practices, but also suggest instructors self-assess their teaching practices in these areas and measure student anxiety and support perceptions if they are concerned about negative impacts of this emotion on student outcomes. Our work adds to the literature suggesting that instructor practices are central to student emotions, and our study shows that instructor practices may not be perceived similarly by all students, but that this can change over time. Given the role of emotion as an often-hidden factor in student success, we believe it is critical for future work to further investigate the links between instructor practices and student emotions.

REFERENCES

• Allen, M., Witt, P. L., & Wheeless, L. R. (2006). The role of teacher immediacy as a motivational factor in student learning: Using meta-analysis to test a causal model. Communication Education, 55(1), 21–31. https://doi.org/10.1080/03634520500343368 Google Scholar
• Alvi, T., Kouros, C. D., Lee, J., Fulford, D., & Tabak, B. A. (2020). Social anxiety is negatively associated with theory of mind and empathic accuracy. Journal of Abnormal Psychology, 129(1), 108–113. https://doi.org/10.1037/abn0000493 Medline, Google Scholar
• American College Health Association. (2019). NCHA-III Fall 2019 Undergraduate Reference Group Executive Summary [PDF file]. Retrieved January 23, 2023 from https://www.acha.org/NCHA/ACHA-NCHA_Data/Publications_and_Reports/NCHA/Data/Reports_ACHA-NCHAIII.aspx Google Scholar
• American College Health Association. (2022). NCHA-III Spring 2022 Undergraduate Reference Group Executive Summary [PDF file]. Retrieved January 23, 2023 from https://www.acha.org/NCHA/ACHA-NCHA_Data/Publications_and_Reports/NCHA/Data/Reports_ACHA-NCHAIII.aspx Google Scholar
• American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for educational and psychological testing. Google Scholar
• Black, A. E., & Deci, E. L. (2000). The effects of instructors' autonomy support and students' autonomous motivation on learning organic chemistry: A self-determination theory perspective. Science Education, 84(6), 740–756. https://doi.org/10.1002/1098-237X(200011)84:6≤740::AID-SCE4≥3.0.CO;2-3 Google Scholar
• Canning, E. A., White, M., & Davis, W. B. (2024). Growth mindset messages from instructors improve academic performance among first-generation college students. CBE—Life Sciences Education, 23(2), ar14. https://doi.org/10.1187/cbe.23-07-0131 Medline, Google Scholar
• Cavanagh, A. J., Chen, X., Bathgate, M., Frederick, J., Hanauer, D. I., & Graham, M. J. (2018). Trust, growth mindset, and student commitment to active learning in a college science course. CBE—Life Sciences Education, 17(1), ar10. https://doi.org/10.1187/cbe.17-06-0107 Link, Google Scholar
• Chao, C. (2017). Correlation, Pearson. In: M. Allen (Ed.), The SAGE Encyclopedia of Communication Research Methods, 4, (pp. 267–270). Thousand Oaks, CA: SAGE Publications, Inc. https://doi.org/10.4135/9781483381411 Google Scholar
• Chen, J. J., Tsong, Y., & Kang, S. H. (2000). Tests for equivalence or noninferiority between two proportions. Drug Information Journal, 34(2), 569–578. https://doi.org/10.1177/009286150003400225 Google Scholar
• Cooper, K. M., & Brownell, S. E. (2020). Student anxiety and fear of negative evaluation in active learning science classrooms. In: E. WalterJ. J. Mintzes (Eds.), Active Learning in College Science, the Case for Evidence-based Practice. Heidelberg, DE: Springer Nature. Google Scholar
• Cooper, K. M., Downing, V. R., & Brownell, S. E. (2018). The influence of active learning practices on student anxiety in large-enrollment college science classrooms. International Journal of STEM Education, 5, 23. https://doi.org/10.1186/s40594-018-0123-6 Medline, Google Scholar
• Docan-Morgan, T., & Manusov, V. (2009). Relational turning point events and their outcomes in college teacher-student relationships from students' perspectives. Communication Education, 58(2), 155–188. https://doi.org/10.1080/03634520802515713 Google Scholar
• England, B. J., Brigati, J. R., & Schussler, E. E. (2017). Student anxiety in introductory biology classrooms: Perceptions about active learning and persistence in the major. PLoS One, 12(8), e0182506. https://doi.org/10.1371/journal.pone.0182506 Medline, Google Scholar
• England, B. J., Brigati, J. R., Schussler, E. E., & Chen, M. M. (2019). Student anxiety and perception of difficulty impact performance and persistence in introductory biology courses. CBE—Life Sciences Education, 18(2), ar21. https://doi.org/10.1187/cbe.17-12-0284 Link, Google Scholar
• Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K., Wenderoth, M. P., Okoroafor, N., & Jordt, H. (2014). Active learning increases student performance in science, engineering, and mathematics. Proceedings of the National Academy of Science USA, 111(23), 8410–8415. https://doi.org/10.1073/pnas.1319030111 Medline, Google Scholar
• Frymier, A. B. (1993). The impact of teacher immediacy on students' motivation: Is it the same for all students? Communication Quarterly, 41(4), 454–464. https://doi.org/10.1080/01463379309369905 Google Scholar
• Gaudier-Diaz, M. M., Sinisterra, M., & Muscatell, K. A. (2019). Motivation, belongingness, and anxiety in neuroscience undergraduates: Emphasizing first-generation college students. Journal of Undergraduate Neuroscience Education: JUNE: A publication of FUN, Faculty for Undergraduate Neuroscience, 17(2), A145–A152. Medline, Google Scholar
• Goetz, T., Keller, M. M., Lüdtke, O., Nett, U. E., & Lipnevich, A. A. (2019). The dynamics of real-time classroom emotions: Appraisals mediate the relation between students’ perceptions of teaching and their emotions. Journal of Educational Psychology, 112(6), 1243–1260. http://dx.doi.org/10.1037/edu0000415 Google Scholar
• Goetz, T., Lüdtke, O., Nett, U. E., Keller, M. M., & Lipnevich, A. A. (2013). Characteristics of teaching and students’ emotions in the classroom: Investigating differences across domains. Contemporary Educational Psychology, 38(4), 383–394. https://doi.org/10.1016/j.cedpsych.2013.08.001 Google Scholar
• Gonzalez, A. A., Bottenhorn, K. L., Bartley, J. E., Hayes, T., Riedel, M. C., Salo, T., … & Laird, A. R. (2019). Sex differences in brain correlates of STEM anxiety. NPJ Science of Learning, 4(1), 18. https://doi.org/10.1038/s41539-019-0058-9 Medline, Google Scholar
• Gross, J. J. (2015). Emotion regulation: Current status and future prospects. Psychological Inquiry, 26(1), 1–26. https://doi.org/10.1080/1047840X.2014.940781 Google Scholar
• Gross, J. J., & Jazaieri, H. (2014). Emotion, emotion regulation, and psychopathology: An affective science perspective. Clinical Psychological Science, 2(4), 387–401. https://doi.org/10.1177/2167702614536164 Google Scholar
• Grossberg, S. (2009). Cortical and subcortical predictive dynamics and learning during perception, cognition, emotion, and action. Philosophical Transactions: Biological Sciences, 364, 1223–1234. https://doi.org/10.1098/rstb.2008.0307 Medline, Google Scholar
• Harris, R. B., Grunspan, D. Z., Pelch, M. A., Fernandes, G., Ramirez, G., & Freeman, S. (2019). Can test anxiety interventions alleviate a gender gap in an undergraduate STEM course? CBE—Life Sciences Education, 18(3), ar35. https://doi.org/10.1187/cbe.18-05-0083 Link, Google Scholar
• Hill, C. E., Thompson, B. J., & Williams, E. N. (1997). A guide to conducting consensual qualitative research. The Counseling Psychologist, 25(4), 517–572. https://doi.org/10.1177/0011000097254001 Google Scholar
• Hoyt, L. T., Cohen, A. K., Dull, B., Castro, E. M., & Yazdani, N. (2021). “Constant stress has become the new normal”: Stress and anxiety inequalities among US college students in the time of COVID-19. Journal of Adolescent Health, 68(2), 270–276. https://doi.org/10.1016/j.jadohealth.2020.10.030 Medline, Google Scholar
• Hu, L.-t., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. https://doi.org/10.1080/10705519909540118 Google Scholar
• Jang, H., Kim, E. J., & Reeve, J. (2016). Why students become more engaged or more disengaged during the semester: A self-determination theory dual-process model. Learning and Instruction, 43, 27–38. https://doi.org/10.1016/j.learninstruc.2016.01.002 Google Scholar
• Kim, C., & Pekrun, R. (2014). Emotions and motivation in learning and performance. In: J. M. Spector (Ed.), Handbook of Research on Educational Communications and Technology (pp., 65–75). New York: Springer Science+Business Media LLC. Google Scholar
• Lammers, W. J., Gillaspy, Jr, J. A., & Hancock, F. (2017). Predicting academic success with early, middle, and late semester assessment of student–instructor rapport. Teaching of Psychology, 44(2), 145–149. https://doi.org/10.1177/0098628317692618 Google Scholar
• Lei, H., Cui, Y., & Chiu, M. M. (2018). The relationship between teacher support and students' academic emotions: A meta-analysis. Frontiers in Psychology, 8, 2288. https://doi.org/10.3389/fpsyg.2017.02288 Medline, Google Scholar
• Mainhard, T., Oudman, S., Hornstra, L., Bosker, R. J., & Goetz, T. (2018). Student emotions in class: The relative importance of teachers and their interpersonal relations with students. Learning and Instruction, 53, 109–119. https://doi.org/10.1016/j.learninstruc.2017.07.011 Google Scholar
• Mainhard, T., Oudman, S., Hornstra, L., Bosker, R. J., & Goetz, T. (2018). Student emotions in class: The relative importance of teachers and their interpersonal relations with students. Learning and Instruction, 53, 109–119. https://doi.org/10.1016/j.learninstruc.2017.07.011 Google Scholar
• Mazer, J. P., McKenna-Buchanan, T. P., Quinlan, M. M., & Titsworth, S. (2014). The dark side of emotion in the classroom: Emotional processes as mediators of teacher communication behaviors and student negative emotions. Communication Education, 63(3), 149–168. https://doi.org/10.1080/03634523.2014.904047 Google Scholar
• Micari, M., & Calkins, S. (2021). Is it OK to ask? The impact of instructor openness to questions on student help-seeking and academic outcomes. Active Learning in Higher Education, 22(2), 143–157. https://doi.org/10.1177/1469787419846620 Google Scholar
• Mohammed, T. F., Nadile, E. M., Busch, C. A., Brister, D., Brownell, S. E., Claiborne, C. T., … & Cooper, K. M. (2021). Aspects of large-enrollment online college science courses that exacerbate and alleviate student anxiety. CBE—Life Sciences Education, 20(4), ar69. https://doi.org/10.1187/cbe.21-05-0132 Medline, Google Scholar
• Nunnally, J. C. (1978). Psychometric Theory, 2nd ed. New York, NY: McGraw-Hill. Google Scholar
• PCAST. (2012). Engage to excel: Producing one million additional college graduates with degrees in Science, Technology, Engineering, and Mathematics. Report to the President. Executive Office of the President. Washington, D.C. Retrieved January 23, 2024 from https://obamawhitehouse.archives.gov/sites/default/files/microsites/ostp/pcast-engage-to-excel-final_2-25-12.pdf Google Scholar
• Pekrun, R. (1992). The impact of emotions on learning and achievement: Towards a theory of cognitive/motivational mediators. Applied Psychology: An International Review, 41(4), 359–376. https://doi.org/10.1111/j.1464-0597.1992.tb00712.x Google Scholar
• Pekrun, R. (2006). The control-value theory of achievement emotions: Assumptions, corollaries, and implications for educational research and practice. Educational Psychology Review, 18(4), 315–341. https://doi.org/10.1007/s10648-006-9029-9 Google Scholar
• Pekrun, R., Frenzel, A. C., Goetz, T., & Perry, R. P. (2007). The control-value theory of achievement emotions: An integrative approach to emotions in education. In: P. A. SchutzR. Pekrun (Eds.), Emotion in education (pp. 13–36). Elsevier Academic Press. https://doi.org/10.1016/B978-012372545-5/50003-4 Google Scholar
• Pekrun, R., Goetz, T., Titz, W., & Perry, R. P. (2002). Academic emotions in students' self-regulated learning and achievement: A program of qualitative and quantitative research. Educational Psychologist, 37(2), 91–105. https://doi.org/10.1207/S15326985EP3702_4 Google Scholar
• Pekrun, R., & Linnenbrink-Garcia, L. (2012). Academic emotions and student engagement. In: S. ChristensonA. ReschlyC. Wylie (Eds.), Handbook of Research on Student Engagement (pp. 259–282). Boston, MA: Springer US. Google Scholar
• Pekrun, R., Marsh, H. W., Elliot, A. J., Stockinger, K., Perry, R. P., Vogl, E., & Vispoel, W. P. (2023). A three-dimensional taxonomy of achievement emotions. Journal of Personality and Social Psychology, 124(1), 145. https://doi.org/10.1037/pspp0000448 Medline, Google Scholar
• Reeve, J. (2009). Why teachers adopt a controlling motivating style toward students and how they can become more autonomy supportive. Educational Psychologist, 44(3), 159–175. 10.1080/00461520903028990 Google Scholar
• Respondek, L., Seufert, T., Stupinsky, R., & Nett, U. E. (2017). Perceived academic control and academic emotions predict undergraduate university student success: Examining effects on dropout intention and achievement. Frontiers in Psychology, 8, 243. https://doi.org/10.3389/fpsyg.2017.00243 Medline, Google Scholar
• RStudio Team. (2020). RStudio: Integrated development for R. RStudio, PBC, Boston, MA. Retrieved from http://www.rstudio.com/ Google Scholar
• Saldaña, J. (2009). The Coding Manual for Qualitative Researchers. Thousand Oaks, CA: Sage Publications Ltd. Google Scholar
• Savalei, V. (2012). The relationship between root mean square error of approximation and model misspecification in confirmatory factor analysis models. Educational and Psychological Measurement, 72, 910–932. https://doi.org/10.1177/0013164412452564 Google Scholar
• Schussler, E. E., Weatherton, M., Chen Musgrove, M. M., Brigati, J. R., & England, B. J. (2021). Student perceptions of instructor supportiveness: What characteristics make a difference? CBE—Life Sciences Education, 20(2), ar29. https://doi.org/10.1187/cbe.20-10-0238 Link, Google Scholar
• Seidel, S. B., Reggi, A. L., Schinske, J. N., Burrus, L. W., & Tanner, K. D. (2015). Beyond the biology: A systematic investigation of noncontent instructor talk in an introductory biology course. CBE—Life Sciences Education,, 14(4), ar43. https://doi.org/10.1187/cbe.15-03-0049 Link, Google Scholar
• Shi, D., Lee, T., & Maydeu-Olivares, A. (2019). Understanding the model size effect on SEM fit indices. Educational and Psychological Measurement, 79(2), 310–334. https://doi.org/10.1177/0013164418783530 Medline, Google Scholar
• Small, A. C., Hollenbeck, A. R., & Haley, R. L. (1982). The effect of emotional state on student ratings of instructors. Teaching of Psychology, 9(4), 205–211. https://doi.org/10.1207/s15328023top0904_3 Google Scholar
• Stains, M., & Vickrey, T. (2017). Fidelity of implementation: An overlooked yet critical construct to establish effectiveness of evidence-based instructional practices. CBE—Life Sciences Education, 16(1), rm1. https://doi.org/10.1187/cbe.16-03-0113 Link, Google Scholar
• Stolz, D. S., Müller-Pinzler, L., Krach, S., & Paulus, F. M. (2020). Internal control beliefs shape positive affect and associated neural dynamics during outcome valuation. Nature Communications, 11(1), 1230. https://doi.org/10.1038/s41467-020-14800-4 Medline, Google Scholar
• Theobald, E. J., Hill, M. J., Tran, E., Agrawal, S., Arroyo, E. N., Behling, S., … & Freeman, S. (2020). Active learning narrows achievement gaps for underrepresented students in undergraduate science, technology, engineering, and math. Proceedings of the National Academy of Science USA, 117(12), 6476–6483. https://doi.org/10.1073/pnas.1916903117 Medline, Google Scholar
• Thissen, D., Steinberg, L., & Kuang, D. (2002). Quick and easy implementation of the Benjamini-Hochberg procedure for controlling the false positive rate in multiple comparisons. Journal of Educational and Behavioral Statistics, 27(1), 77–83. https://doi.org/10.3102/10769986027001077 Google Scholar
• Walker, S. L., & Fraser, B. J. (2005). Development and validation of an instrument for assessing distance education learning environments in higher education: The Distance Education Learning Environments Survey (DELES). Learning Environments Research, 8, 289–308. https://doi.org/10.1007/s10984-005-1568-3 Google Scholar
• Weatherton, M., Rosenberg, J. M., Rocconi, L., & Schussler, E. E. (2024). Beyond pre-post surveys: Exploring validity evidence for the use of experience sampling methods to measure student anxiety in introductory biology. bioRxiv, 2024–2006. https://doi.org/10.1101/2024.06.06.597780 Google Scholar
• Wickham, H., Averick, M., Bryan, J., Chang, W., McGowan, L. D., François, R., … Yutani, H. (2019). “Welcome to the tidyverse.” Journal of Open Source Software, 4(43), 1686. https://doi.org/10.21105/joss.01686 Google Scholar
• Wilson, J. (2022). Initial steps in developing classroom observation rubrics designed around instructional practices that support equity and access in classrooms with potential for “success”. Teachers College Record, 124(11), 179–217. https://doi.org/10.1177/01614681221140963 Google Scholar
• Yerkes, R. M., & Dodson, J. D. (1908). The relationship of strength of stimulus to rapidity of habit formation. Journal of Comparative Neurology and Psychology, 18(5), 459–482. https://doi.org/10.1002/cne.920180503 Google Scholar