When teacher-centered instructors are assigned to student-centered classrooms

Technology-rich student-centered classrooms such as SCALE-UP and TEAL are designed to actively engage students. We examine what happens when the design of the classroom (conventional or teacher-centered versus student-centered classroom spaces) is consistent or inconsistent with the teacher’s epistemic beliefs about learning and teaching (traditional or teacher-centered versus student-centered pedagogies). We compare two types of pedagogical approaches and two types of classroom settings through a quasiexperimental $2\times 2$ factorial design. We collected data from 214 students registered in eight sections of an introductory calculus-based mechanics course given at a Canadian publicly funded two-year college. All students were given the Force Concept Inventory at the beginning and at the end of the 15-week-long course. We then focused on six teachers assigned to teach in the student-centered classroom spaces. We used qualitative observations and the Approaches to Teaching Inventory (ATI), a self-reported questionnaire, to determine the teachers’ epistemic beliefs (teacher-centered or student-centered) and how these beliefs affected their use of the space and their students’ conceptual learning. We report four main findings. First, the student-centered classroom spaces are most effective when used with student-centered pedagogies. Second, student-centered classrooms are ineffective when used with teacher-centered pedagogies and may have negative effects for students with low prior knowledge. Third, we find a strong correlation between six instructors’ self-reported epistemic beliefs of student centeredness and their classes’ average normalized gain ($r=0.91$; $p=0.012$). Last, we find that some instructors are more willing to adopt student-centered teaching practices after using student-centered classroom spaces. These data suggest that student-centered classrooms are effective only when instructors’ epistemic framework of teaching and learning is consistent with a student-centered pedagogy. However, the use of the student-centered classrooms may change instructors’ epistemic frameworks over time. Further research should focus on how to better support teachers with shifting epistemic frameworks as well as helping students with lower prior knowledge in student-centered classroom spaces.

Figure 1

The difference pedagogy makes on conceptual normalized gain is significant. Classroom settings alone do not contribute significantly to conceptual gain. Error bars represent the standard error to the mean.

Figure 2

Independent of classroom type, students in active learning pedagogies have significantly greater mean post-FCI score, when adjusted for pre-FCI score. Classroom settings alone do not contribute significantly to end of semester conceptual knowledge. Also, students in traditional pedagogies achieve better results in conventional classrooms than in sociotechnological classrooms. Error bars represent the standard error to the mean.

Figure 3

Students in active learning student-centered pedagogies achieve higher normalized gains than students in traditional teacher-centered instruction regardless of whether they have low, medium, or high prior knowledge. Error bars represent the standard error to the mean.

Figure 4

Students in active learning student-centered pedagogies achieve higher post-FCI mean scores (ANCOVA adjusted for pre FCI) than students in traditional teacher-centered instruction. However, the difference in achievement for high prior knowledge students is not statistically significant. Error bars represent the standard error to the mean.

Figure 5

Differences in normalized gain for students with low, medium, and high prior knowledge in either sociotechnological or conventional classrooms. Error bars represent the standard error to the mean.

Figure 6

Differences in mean post-FCI score for students with low, medium, and high prior knowledge in either sociotechnological or conventional classrooms. Error bars represent the standard error to the mean.