Anxiety is increasingly reported among college students in the United States (Bledsoe & Baskin, 2014; Kitzrow, 2003). An analysis of student anxiety based on data in a mental health database found that distress related to academic concerns was responsible for most of the variation in student anxiety levels (Jones et al., 2018). Several investigators have explored student anxiety in undergraduate classrooms, specifically finding that active learning practices are related to differential levels of anxiety among students in the same class (Broeckelman-Post et al., 2016; Cooper et al., 2018; England et al., 2017). One unexplored question is the extent to which different types of active learning anxiety vary by where students sit in a biology classroom.

Spatial variation in anxiety has been found in both classroom and nonclassroom settings. A study by Rebeta et al. (1993) measured state anxiety (transitory anxiety felt in a particular context), trait anxiety (tendency to experience anxiety based on personality), and test anxiety (anxiety in response to assessment) and found that trait anxiety increased significantly in students who sat in the back versus the first row. Luck (2006) found that patients with anxiety or depression had different left-versus-right seating preferences than patients without these conditions. These studies provide evidence that students with different anxiety levels may sit in different locations, but they do not indicate why this may be.

Losonczy-Marshall and Marshall (2013) identified five factors underlying student seating preference: performance, social, asocial, noticeability, and environment. Of the five, performance (conditions that students believed help them succeed) and noticeability (being seen by others and the professor) were identified most often by students who preferred the first row, asocial (not wanting to stand out or sit near others) was identified most often by students who preferred the back of the room, and environment (proximity to doors, windows, or vents) was identified most often by students who preferred to sit in the center. Although this suggests personality plays a role in seating preference, a nonclassroom study by Karev (2000) found that cinema attendees preferred seats on the right and suggested that handedness and emotional processing plays a role. Harms et al. (2015), however, found that students preferred seats on the left side of the classroom and attributed it to left- versus right-brain processing of information.

Regardless of what drives student preferences about where to sit, seating location has been linked to student performance and engagement in the classroom. Benedict and Hoag (2010) found that students who preferred to sit in the back of the classroom were more likely to receive a D or F in the course, even when those students did not sit in the back. Conversely, students who preferred the front had a higher probability of receiving an A in the course. Marshall and Losonczy-Marshall (2010) studied 1,829 students over 15 years and found higher grades and attendance for students seated in the middle of classrooms. Parker et al. (2011) studied 55 students to determine whether class participation differences were due to student personality (linked to seat choice) or the influence of location on student behavior. One group was assigned seats in the classroom for the entire semester, while the second group was assigned a new seat each day. At the end of the study, students assigned to the front row all semester participated more than those assigned to the back, but there were no differences between the front and the back for those who moved each day. Their results suggested that participation frequency is an interaction between personality preference and influence of seating location.

Given the research on anxiety, seating location preferences, engagement, and performance, we investigated whether students’ preferred seating location in introductory biology related to active learning anxiety levels and why students preferred certain seating locations. The results will add to the literature about how student anxiety may interact with active learning, even at smaller spatial scales within a classroom. The study was guided by two research questions:

1. Are there differences in active learning anxiety levels among students who prefer to sit in one of nine different classroom locations?
2. Does the reason for seating location preference relate to potential anxiety differences?

Methods

Context of the study

This study was conducted in spring 2017 at a large, public research university in the south. In week 4 of the class, students in seven introductory biology lecture courses (three organismal and four cellular) were provided information about the study and an online survey link by their instructor. The survey was the same for each class. Course sizes were: 75 (N = 2), 100, 150, 175, and 225 (N = 2). Students who chose to complete the survey received 1 to 3 bonus points (depending on the instructor) toward their 1,000-point total in the class.

Consistent with the Vision and Change in Undergraduate Biology Education (AAAS, 2011) report, these classes were taught using active learning, typically consisting of clicker questions, instructors’ verbal questions that students answered voluntarily or were called on to answer (cold calling), or in-class written work. Much of this work was done in small student groups. Six classes were taught in traditional stadium-style lecture classrooms (four classes in two of the same rooms). All the stadium-style classrooms had two central aisles. All except one of the rooms had student entrances at both the front and back. These entrances were symmetrical—with either one entrance at each corner of the room or one large central entrance and exit. The seventh classroom was not a stadium-style lecture class, but a large, flat (i.e., non-stadium style or tiered) classroom with long tables and chairs that seated about 75 students. This room had one entrance at the front left and one central aisle. Three of the rooms had centrally located presentation podiums, and three of the rooms had presentation podiums at the front right (relative to the instructor), although all instructors moved around the front of the room while they taught. Classroom layouts are shown in Figure 1.

Data collection

The survey (Figure 2) collected students’ self-reported information about their active learning anxiety levels, their preferred seating location, why they preferred to sit there, and demographic information (gender, year in school, racial and ethnic identity, and instructor). The survey included four closed-response questions, 25 Likert scale questions, and two open-ended questions. There were also other questions not used as part of this study. The total time required to complete the survey was 10 to 15 minutes.

Students’ perceptions of their general anxiety toward the course were assessed using a survey modified from Papanastasiou and Zembylas (2008) by England et al. (2019). We also measured test, communication, and social anxiety, as these are common in large introductory courses (Culler & Holahan, 1980; Gerwing et al., 2015; Zeidner & Matthews, 2005). For test anxiety, the items from the Motivated Strategies for Learning Questionnaire (MSLQ; Pintrich et al., 1991) were used. We also used the communication anxiety (talking in front of the class) and social anxiety (working with peers) scales from the Personal Report of Communication Apprehension-24 (PRCA-24; McCroskey, 1982) instrument. To measure student anxiety about specific active learning practices, single-item measures asked students to rank their anxiety about four active learning practices (England et al., 2017).

In addition to measures of anxiety, students were asked, “Where do you prefer to sit in your biology lecture class?” The preferred seating location was based on nine potential locations: front-right, front-middle, front-left, middle-right, middle-middle, middle-left, back-right, back-middle, back-left. Students were provided a reference to the teacher (right and left were in reference to how the teacher saw the class) and asked to select one preferred location from the nine options. We explicitly indicated that their preferred location may differ from where they actually sat for various reasons. Students then answered an open-ended question explaining why they preferred that location (“Why do you prefer to sit in the location you indicated?”).

All procedures for this study were approved by the institution’s human subjects review board prior to the start of the research (IRB-16-03181-XP).

Data analysis

Anxiety levels and seating location

The survey data were downloaded to an external program for statistical analysis (SPSS, 2017). Any responses from students who did not consent to participate in the study or who were younger than 18 were removed from the data set. There were 665 students who completed the survey; however, not every student answered every question. The respondents of the survey were mostly freshmen (62%), female (67%), and White (76%) (Table 1). Response rates for each class ranged from 28% to almost 100% of students in the class (N = 25 to N = 227 per class).

Students’ preferred seating locations were coded as a number from 1 to 9, each aligned with one of the nine classroom locations. Students who selected more than one location were removed from the data set. We pooled the data from the individual lecture classes because the sample sizes for each of the nine locations in each class were too small for robust statistical analysis.

Individual student responses to the four general anxiety items were averaged to arrive at a mean general anxiety score for each student (with 7 being the highest level of anxiety). These student anxiety scores were then sorted by preferred seating location (1 to 9) for each student, and the scores of the students in each of the nine locations were averaged. Student responses to the test anxiety scale were also averaged, then sorted and averaged by room location. Student scores for communication and social anxiety were calculated based on the protocol provided by McCroskey (1982). Communication and social anxiety scores for individual students ranged from 6 to 30, with higher scores indicating higher levels of anxiety. Once again, students were sorted into their preferred room locations, and scores for communication and social anxiety were averaged for each location. For the single-item measures of anxiety toward particular active learning practices, averages for students who preferred each of the nine room locations were calculated (once again, a score of 7 was the highest level of anxiety for each practice). The dependent variable for all analyses was the average anxiety level for each of the nine room locations. The constructs of general, test, communication, and social anxiety were validated with this population in fall 2016 (England et al., 2019).

We conducted a multivariate analysis of covariance (MANCOVA) with the dependent anxiety variables, room location as a fixed effect, and year in school as a covariate to test whether there was a difference in the dependent variables by location and whether the student’s year impacted the model. Based on the MANCOVA results, we then conducted separate analyses of variance (ANOVAs) with post-hoc tests (Tukey’s honestly significant difference [HSD]) for the anxiety variables with room location as a fixed effect.

Student explanations for preferred seating location

A total of 590 students indicated where they preferred to sit in the room and why they chose to sit there, with preferences as follows: 23 in the back-left, 27 in the back-middle, 29 in the back-right, 52 in the middle-left, 107 in the middle-middle, 59 in the middle-right, 69 in the front-left, 151 in the front-middle, and 73 in the front-right.

Responses indicating why students preferred to sit in a particular location underwent inductive thematic analysis (Braun & Clarke, 2006). Through this process, we were blind as to the seating preference of the students. The first and second authors each read the explanations independently and made lists of potential codes that would capture most of the student responses. We then met to generate a consensus list of codes into which we would sort the student responses.

We then independently assigned codes to the student responses. Students sometimes gave multiple reasons for their seating preference, in which case students received multiple codes for their response (e.g., to see, to hear, and to be close to the professor were separate codes). As we coded, we had regular discussions about any coding difficulties and revised the coding sheet and recoded responses to reconcile any confusion.

After independently coding all student responses, one researcher compared all codes and identified those with and without agreement between coders. We had an initial agreement of 86.7%. All codes that were different between the coders were discussed until agreement was reached. The agreed upon and reconciled codes were used as the final codes for each student response.

After codes were assigned, the data were sorted by preferred seating location, and a frequency for each code in each of the nine room locations was generated. As we organized these data, we compiled the codes into overarching themes that described the codes. For example, there were many codes about wanting to be near the instructor to hear, near the screen to see the slides, and near the front to stay focused. These were grouped into an “engagement” theme. Using the same technique, we grouped students who said they did not want to be noticed, or wanted to have more space, into an “Avoidance” theme.

Results

Anxiety levels and students’ preferred seating locations

The number of student responses for each classroom location varied from a low of 14 in the back-left to a high of 146 in the front-middle; however, responses varied by location and question. Response rate ranges are indicated in Figure 3.

For the overall sample, students’ general anxiety averaged slightly less than the midpoint on the 7-point Likert scale (3.07). Of the four active learning practices, student anxiety was highest for volunteering to answer a question (3.96 out of 7) and cold calling (4.73 out of 7) and lowest for clicker questions (2.35 out of 7) and group work (2.70 out of 7). Average test anxiety was almost at the middle of the 7-point Likert scale range (3.95). Both social anxiety (20.05) and communication anxiety (19.97) were mid-range on the scale that ranged from 6 to 30.

The MANCOVA revealed differences among the anxiety variables by location (Wilks’ lambda 0.784; F(64, 2095) = 1.408, p = .019) and indicated that year was not significant. For the ANOVA results, there were no differences in general anxiety among the students who preferred to sit in different room locations. There was a difference in test anxiety by location (p = .022), with higher levels of test anxiety in the back-left and front-middle (throughout this article, a statement of room location is inferred to mean “students who preferred to sit in X room location”), and lower levels of test anxiety in the back-right (see Figure 3 for anxiety values and significant differences by location; Table 2 shows the significant post-hoc room location comparisons with p-values). The overall communication anxiety model was significant (p = .045), but the post-hoc analyses did not identify significant differences among specific locations. Social anxiety was also significant overall (p = .004), with the back-left having higher levels of social anxiety, and all front-row locations and the middle-middle having lower levels of social anxiety.

Among the four active learning practices, each had significant differences among particular room locations. Anxiety toward the use of clickers was significantly different among the nine locations (p = .002), with the back-left, middle-right, and front-middle having higher levels of clicker anxiety than the back-right. There was an overall significant difference among locations for anxiety about volunteering to answer a question (p = .012), with the middle-right having higher levels of anxiety and the front-right having lower levels of anxiety. There was also a significant difference among locations for anxiety about cold calling (p = .024), with higher levels of cold call anxiety in the middle-right and lower levels of cold call anxiety in the back-right. Anxiety toward group work was significantly different among the nine locations (p = .012), with the back-left having higher levels of group work anxiety and the back-middle, back-right, middle-middle, and front-right having lower levels of group work anxiety.

The only location in the room that did not have higher or lower anxiety than other locations was the middle-left (Figure 3).

Explanations for seating location

The codes were organized into five overarching themes: engagement, proximity, ease, avoidance, and view (Table 3). Students choosing a location for engagement wanted to see the screen, hear the instructor, and pay attention to the content and thought that location would be less distracting than others. The proximity theme reflected a code of students wanting to be close, but not too close, to the instructor and content. An ease theme reflected students’ desire to be relaxed and comfortable and able to find a seat easily and leave the classroom when needed. The avoidance theme reflected codes of wanting to sit apart from others, get away from the crowd, spread out, and not be noticed. Finally, the view theme encompassed codes of wanting to see the landscape of the classroom and view all students at once.

The assignment of themes to each room location was determined by identifying the top three highest frequency codes for each of the nine areas and matching the theme to those codes (e.g., if “need to see” was in the top three in the middle-middle, this indicated “engagement” as a theme in that location). In one case, ties in the top three codes resulted in four themes for one location, or in other cases all three codes were in the same theme, resulting in only one theme for a room location. The themes for each location and summary of quantitative results are shown in Figure 4.

For the engagement theme, as student seating preference moved from the front to the back of the classroom, there was an overall decrease in the prevalence of engagement codes; however, engagement codes were among the top three most frequent codes in each location except the back-right. The codes related to the proximity theme were most frequent in the middle-middle and middle-right portions of the classroom. Avoidance codes were rare in the front (never rising into the top three in frequency) and increased toward the back of the room, but they only broke into the top three codes in the back-left of the classroom. Ease codes were also infrequent among students who preferred sitting in the front of the classroom; none of these codes were in the top three for any of the front room locations. These codes increased in proportion as seating preference moved toward the back of the classroom, with this theme identified for all back-row locations and the middle-left. View codes were more likely to be mentioned by students who preferred the back-left and back-middle.

Elisabeth E. Schussler (eschussl@utk.edu) is a professor in the Department of Ecology and Evolutionary Biology and associate head for undergraduate programs at the University of Tennessee, Knoxville (UTK); Brianna Reynolds earned a bachelor’s of science degree from the Department of Ecology and Evolutionary Biology at UTK in 2019; Benjamin J. England is an assistant professor in the Biology Department at Saint Louis University in St. Louis, Missouri; and Jennifer R. Brigati is an associate professor in the Department of Biology and chair of the Division of Natural Sciences at Maryville College in Maryville, Tennessee.