Effect of Seating Arrangement on Class Engagement in Team-based Learning: A Quasi-Experimental Study

Seating Arrangement and Engagement

There is a growing literature on the importance of learning spaces and how it affects teaching and learning[1]. Especially, learning spaces affecting the delivery of higher education[2] and the medical curriculum[3]. The effects of physical space on human activity has been studied where the area of environmental psychology examines subjects such as psychological comfort with space and its incentivising effects[4]. Classroom designs can vary in many forms to strengthen classroom discussions[5]. In traditional lecture hall settings, engagement is worst towards the back of the class[6], grades decreased as distance from the instructor increased, and a smaller proportion of students participated in larger class sizes[7]. On the other hand, semicircular arrangements allow for more comfortable discussion between tutor and students and within students themselves[8]. The students also asked more questions and learnt more compared to traditional row and column organization[9]. Such decentralised layouts encourage participation and show that the classroom’s physical attributes clearly affect the engagement of students.

Cognitive engagement is the psychological construct characterized by active involvement in learning, where students have autonomy[10] and put effort to invest in studying[11-13]. It varies depending on the activity[14], increasing as knowledge is gained, and with more independence in the task involved[15]. It is important in the learning process as cognitive engagement is related to positive outcomes in students, such as improved grades[16], motivation[17] and favorable attitudes towards learning[18].

Team-Based Learning

Team-based learning (TBL) is a structured, learner-centered and instructor-led method of active learning relying on small group interaction and class-wide discussion[19, 20]. TBL use is increasingly popular in medical education[21] with positive outcomes in many areas from attitudes and graded assignments[22], to critical thinking, teamwork and communication[23, 24].

A TBL session comprises of three phases: preparation, readiness assurance and application exercise (AE) phase[25] (Figure 1). The preparation phase involves individual learning outside the classroom using resources provided by the school. The subsequent phases occur in the classroom. Firstly, the students start with an Individual Readiness Assurance Test, a closed book multiple-choice question test, which is then repeated in teams, the Team Readiness Assurance Test, and a “burning questions” (BQ) discussion. Lastly, the AE phase challenges students with practical scenarios, which provides them with an opportunity to apply their knowledge gained. This study was conducted at the Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore. TBL is our main pedagogical method for the first two years of medical undergraduate learning. (Refer to Rajalingam et al[26], large-scale TBL paper for more information. TBL procedures described here may vary between institutions.)

• Download figure
• Open in new tab

Figure 1:

Schematic of a typical Team-Based Learning (TBL) process from the Preparation to Readiness Assurance, Burning questions and Application Exercise phase at the Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore. Descriptions about what each phase entails and their average duration are included. TBL procedures described here may vary between institutions.

A theory about the underlying psychological process[27, 28] through which TBL works is that it increases students’ cognitive engagement and subsequently affects achievement[14].

Previous TBL research found a significant increase in students’ engagement[29-32], where the nature of the tasks promotes participation and discussion while making students process the information[25]. Thus, learning what affects engagement may improve the effective implementation of TBL.

Many medical schools that have implemented TBL have also made physical changes to the classroom to accommodate this instructional approach. In this study, our main interest is the effect of seating arrangement on TBL engagement. However, limited research exists in this field. One study by Espey M[33] examined classroom layouts and their impact on TBL. Although grades were not significantly affected, students felt they worked more effectively in classrooms that easily accommodate group work and attitudes improved. With varying cohort sizes among institutions[34], there may be concerns of large-scale TBL implementation as class size increases, students sitting further away maybe disadvantaged. Furthermore, in a previous local study, cognitive engagement fell significantly during BQ and AE tutor explanation phases compared to the previous phases[14]. Exploring the effects of seating arrangement may promote engagement during these tutor-centred phases. While other factors like gender have been examined in the context of seating position[35], the effect of exposure to TBL has not been explored, such as the differences between novice and experienced students.

In this study, we aimed to investigate two factors associated with seating arrangement. Firstly, the effects of seating distance from the tutor on class engagement in TBL. Secondly, if seating orientation within the team, which is the direction students are facing, affects class engagement in TBL. The hypotheses are: (1) sitting nearer to the tutor results in higher engagement and (2) sitting with their front facing the tutor class results in higher engagement.

Study Design

In this quasi-experimental study, students were assigned to teams of 5-7 based on TBL principles[36]. Our classroom design includes a dual-tiered seating circular layout, six big screens all around the periphery of the room, chairs with wheels around fixed tables and microphones for every team. Usually, the teams’ seating positions are permanent for the entire academic year. Tutors were situated in front and remained static for the purposes of this study, reflecting the default situation during TBL sessions. As teams were seated in a circular arrangement, seating orientation was reported as either sitting with their front or back to the tutor. Seating distance from the teams’ table to the tutor’s table ranged from 2.5m to 11m.

Engagement was assessed at two points: BQ and AE, both after tutor explanation. These were chosen as engagement fell significantly compared to the previous phases[14]. Also, seating distance likely affects engagement when the tutor speaks in front of the class, rather than during intra-team discussion at other phases where students discuss within their teams.

The intervention involved rearranging the students’ seating layout and collecting data before and after this swap (see Supplementary Digital Appendix 1 for lesson topics). After obtaining baseline response, teams were swapped front-to-back, with an average change of 4.5m. Students could choose their own seats at the new table. 12-13 teams moved nearer to the tutor, labelled as Group A while 11-12 teams moved further, labelled as Group B (Figure 2). One TBL session was given for students to adjust to their new positions before collecting responses at the subsequent lesson a week later. Researchers were present to conduct the surveys at appropriate intervals. We used SurveyMonkey platform to collect responses electronically.

• Download figure
• Open in new tab

Figure 2:

Classroom seating arrangement at the Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore. The team numbers are represented by the circles. Left: Original seating positions before the swap. Right: New seating positions after swap. Group A: 13 teams (in orange) moved nearer. Group B: 12 teams (in blue) moved further.

There is a dual-tiered circular seating arrangement with 6 large screens all around the periphery, chairs with wheels around fixed tables and microphones for every team. Tutors are in front of the class and remained static.

Participants

150 first-year and 138 second-year undergraduate medical students were recruited with 85 and 75 responses collected, respectively. With one month of TBL experience prior to the study, the first-year students were now referred as the “novice” group. Second-year students had one year of TBL experience and were referred as the “experienced” group. Their age ranged from 18-23 (M=19.5, SD=1.06), with 103 (64%) males and 57 females.

Assessing cognitive engagement

We used two validated self-reporting instruments to assess TBL engagement.

Firstly, the Situational Cognitive Engagement Measure (SCEM) was used to assess cognitive engagement at various TBL phases[15] (see Supplementary Digital Appendix 2 for complete survey). This 5-point Likert scale 4-question survey captures cognitive engagement of the ongoing activity at that instance. This tool developed items to assess the dynamic aspect of engagement, student’s will to persevere on the activity and how immersed they are which links back to the concepts of cognitive engagement. Previous studies[15] attained adequate reliability, including one in a similar local context[14].

Secondly, the Classroom Engagement Survey (CES) was used when each TBL session ended. This 5-point Likert scale 8-question survey retrospectively assesses overall engagement of the class (see Supplementary Digital Appendix 3 for complete survey). Its construct validity was established from previous TBL studies involving undergraduate nursing and medical students[30-32], obtaining high reliability with Cronbach Alpha of 0.80-0.89.

We decided to use both the SCEM and CES as measures of engagement as both instruments measure slightly different aspects of engagement. The SCEM measures how students perceive their immediate engagement while engaged with each TBL activity, while the CES asks students to retrospectively report how engaged they were throughout the TBL session. Cronbach’s alpha was calculated for internal reliability with values greater than 0.80 being desirable[37]. For the SCEM and CES instruments, our mean values were 0.72 and 0.83 respectively, indicating an acceptable to desirable degree of internal reliability.

Lastly, open-ended response questions were provided to further qualitatively explore students’ preferences with seating arrangement and engagement. Students were asked if they

(1) preferred sitting nearer to the tutor, (2) preferred sitting with their front-facing the tutor, (3) reasons for their seating preferences, and (4) if they preferred their original or new seating position.

To compare pre- and post-swap groups, independent-samples T-test was conducted to assess both the effects of seating distance and seating orientation on engagement. Significance levels were set at p<0.05. We used SPSS version 24 (IBM, Armonk, New York).

Seating distance and engagement

Seating orientation and engagement

In both novice and experienced students, SCEM scores for those with their fronts facing the tutor were lower than those with their backs facing the tutor but were not statistically significant.

In novice students, CES scores for those with their fronts facing the tutor was higher than those with their backs facing the tutor but was not statistically significant. In experienced students, CES scores for those with their fronts facing the tutor was lower than those with their backs facing the tutor but was not statistically significant. (Table 3)

Interaction between seating distance and orientation

A two-way analysis of variance (ANOVA) was performed to assess the interaction between seating distance and orientation which was not statistically significant. Novice group: F(1,81)=0.228, p=0.63. Experienced group: F(1,71)=0.937, p=0.34.

Student preferences

Participants were asked to rate, on a scale of 1 (strongly disagree) to 5 (strongly agree), how much they preferred sitting nearer to the tutor and sitting with their front-facing the tutor.

• Novice students preferred sitting nearer to the tutor compared to sitting further away. (M=3.45 vs M=2.53) (p<0.001) while experienced students had no statistically significant difference in such preferences (M=2.62 vs M=2.61) (p=0.95).

• Both novice and experienced students had a statistically significant preference to sit with their front-facing the tutor compared to facing the back. Novice: M=4.20 vs M=3.56 (p=0.001). Experienced: M=4.21 vs M=3.62 (p=0.003)

Further analysis revealed that 86 (51%) students felt that seating distance did not affect their learning, with common reasons such as screens being all around the room and audio from the speaker being well heard. 104 (61%) preferred sitting with their front-facing the tutor, with common reasons being a preference to look at the tutor while they are speaking and feeling more focused. Some quotes from students include: “I don’t think it (seating position) affects my learning as there are screens around the studio”, “The classroom is 360° so it doesn’t matter”, “audio from the front of the class can be easily heard everywhere” and “acoustics are good enough such that the lesson can be heard clearly”.

Practical implications

The findings of this paper provide important and practical implications on seating arrangement in the TBL classroom. Juxtaposing the strong preference to sit nearer or face the tutor and the insignificant effect of seating arrangement on engagement, there appear to be deeper factors at play. While students displayed a preference to sit nearer or face the tutor, they can overcome such physical inconveniences and feel similarly engaged at their less ideal spot. We postulate that medical students generally have higher motivation levels given the work needed to get into medical school and the commitment to a specific career[39]. With this motivation to learn, they adapt to wherever they seat and pay attention regardless. Also, tutors are relatively less involved compared to team interactions for the overall TBL experience (assessed with CES tool), possibly explaining why students were not affected by seating arrangement and reinforce the rationale of TBL as a learning modality.

Quotes from students showed that classroom design was influential for their learning. The implication is that learning space design does have an impact on learning and engagement. While other institutions are compelled to use traditional learning spaces for TBL, our institution specially designed the classroom, according to well-documented architectural and human factors, to facilitate TBL[26]. The circular layout with dual tiered seating, six screens and microphones at every table allows students to follow and contribute to discussions wherever they are. Learning spaces are only as successful as the students’ motivation and willingness to best make use of it[40]. This motivation allows students to fully utilise the technological space and adapt to less preferred seating positions.

Other studies comparing traditional and active learning classrooms substantiates this, showing improved ease of working in teams[33], higher student engagement[41, 42] and grades[43, 44] in specialised classrooms for interactive pedagogies. Previous reports highlighted that aligning physical learning spaces with the curriculum is valuable to the holistic health professions education[3]. Classroom layouts that allowed students to see everyone contributed to the perception of a safe learning space[45]. Therefore, such team- centric learning spaces provide an engaging setting, favourable for active, collaborative learning[26].