A Master of Arts thesis in Teaching English to Speakers of Other Languages (TESOL) by Anuja Mariyam Thomas entitled, “Rethinking the Bottom Bun: Assessing Topic Closers”, submitted in May 2021. Thesis advisor is Dr. Philip McCarthy. Soft copy is available (Thesis, Completion Certificate, Approval Signatures, and AUS Archives Consent Form).
This study focusses on the learnability of a model for topic closers. Topic closers are the various approaches through which a paragraph can be ended. The model comprises a total of 10 topic closers, each of which is termed a Label. The current study explores students’ ability to recognize the various Labels of the model. To assess the learnability of the model, I used a pretest-intervention-posttest experimental design. For the pretest, I presented participants with 50 topic closer sentences and asked them to rate each sentence on a scale of 1-6 for effectiveness. Following the pretest, I presented participants with an explanation document that details all relevant information. The posttest assessment repeated the method used in the pretest with a different set of 50 sentences. The results were analyzed using two sets of criteria: Average scores and variance scores. Average is the mean of the scores for each Label and is used to assess whether overall scores for each Label rise or fall. Variance is the distance from the mean for each Label and is used to calculate how consistently students recognize each label. The results were conducted using R statistical software to assess both between and within ANOVA analyses. Averages of the Labels deemed more effective saw a rise in posttest values, while averages of the Labels deemed less effective saw a fall. The results for average scoring suggest that students can distinguish more effective examples of topic closers from less effective examples of the same. Additionally, there was a significant fall in variance scores across all data. The results for variance scoring suggest that students gained confidence in their evaluations. As such, we can conclude that the intervention resulted in a significant level of learning. Qualitative analysis was also conducted with the analyses supporting the quantitative results. The results of this study have implications for both classroom learning as well as for intelligent tutoring systems.