TBL vs. Lecture: An Exploration of Student Test Data

In Fall 2012 the first author converted a large (100-150 student) general education course from a lecture based class to a TBL formatted course. This shift entailed restructuring how class time was used and content was delivered, but the content remained the same. Students in both the lecture and TBL classes were assigned the same readings. Additionally, material that had been covered during class in the lecture format (lectures and videos) was delivered outside of class in the TBL format using podcasts, lessons in Moodle, and online access to videos. Students in both formats took the same exams with the same objective (multiple choice, true/false) questions. The TBL sections followed best practices: students were placed on permanent teams; the readiness assessment process was completed for each unit; this was followed by application exercises; and peer feedback was given throughout the semester.

A comparison of test question data after 3 TBL semesters revealed that students in lecture sections got 78% of questions correct while students in TBL sections got 81% of questions correct (this difference was significant). The first author was interested in exploring where this improvement came from and wondered what methods of content delivery in lecture based classes were more or less effective than in a TBL class? For example, did students in TBL sections have a better grasp of material that had previously only been assigned outside of class (e.g. a reading not covered in lecture)? Did they fare worse on material that used to be delivered via lecture because they now reviewed it on their own?

Working with two colleagues, Lisa Walker and Angela Ferrara, we explored this issue by comparing the percentage of students who correctly answered objective (multiple choice/TF) exam questions in these two learning environments (lecture vs. TBL). We analyzed student exam data collected over five semesters. Two of these semesters were lecture based (340 students total). Three semesters were taught using TBL (356 students total). One hundred and one test questions remained constant over these 5 semesters and these were included in our analysis. The first author determined the correct answers and developed a key that was used to machine score the questions.

Questions were categorized based on how students in the lecture based course received the information contained in the question. A test question was categorized as:

"In-class content delivery" if students were only exposed to the information during face-to-face class time.

"Self-directed/outside of class" if students were only exposed to the information outside of class.

"Both" if the information in the test question was covered during face-to-face class time and in a resource students were expected to review outside of class.

Note, questions were not classified based on how content was delivered in TBL sections because all content was initially delivered outside of class. Class time was then spent engaging in the Readiness Assessment Process and application exercises.

The analysis revealed that regardless of how content was delivered in the lecture section, TBL students had higher scores. All differences were statistically significant. Students exposed to TBL scored higher (81%) on the in-class content delivery questions than students in lecture (79%) (t = 1.819, p = 0.035). Students exposed to TBL got 78% of the self-directed/outside of class questions correct, while students in the lecture sections got 72% correct (t = 5.749, p<0.001). Finally, TBL students scored higher (86%) than lecture students (83%) on questions categorized as both (t = 2.261, p = 0.012). These results suggest TBL can be as effective, if not more effective, than lecture based pedagogy.

Coral Wayland

Angie Ferrara

Lisa Walker 

University of North Carolina at Charlotte