Graduate Student Seminar: Analyzing Free Response Data Using TextRank and GPT

In studies that involve human feedback, accurately analyzing textual data is crucial for drawing meaningful insights. In this presentation, I will explore two powerful methods for analyzing free response data: TextRank and GPT. I will begin by explaining the motivation for analysis and the collection method of this free response data. I will then explain the underlying processes of TextRank and GPT, detailing the algorithmic structure of TextRank and the general architecture of GPT. I will then present the application where I have utilized these methods, comparing their effectiveness in extracting key themes and insights from free response data. Finally, I will share the results of this analysis, highlighting the continued relevance of TextRank in light of the advanced capabilities of GPT, and discuss when and why one method may be preferred over the other.