Analyzing Constructed Responses in Educational Survey With Latent Dirichlet Allocation: A Demonstration on Students’ Career Aspiration

• In Event: Thursday Roundtable Session 9:00 am
  In Roundtable Session: AI, Psychometrics, and Data Analysis (Table 24)

Abstract

This study investigates the effectiveness of Latent Dirichlet Allocation (LDA; Blei et al., 2003) as a tool for analyzing constructed response data in educational surveys. The traditional manual analysis of such data can be labor-intensive and time-consuming, especially when dealing with large sample sizes and long textual responses. LDA, a statistical algorithm from Natural Language Processing (NLP), provides an efficient and effective solution for discovering major topics within a large collection of documents. In this study, using students’ constructed responses regarding their career aspirations, we demonstrated the utility of LDA in obtaining insights from these responses and transforming textual data into numerical variables which can be used in subsequent statistical analyses.

Authors

• Yuxiao Zhang, Purdue University

• Nielsen Pereira, Purdue University/Gifted Education Research & Resource Institute

• David Arthur, Purdue University

• Hua-Hua Chang, Purdue University

• Zafer Ozen, Purdue University

• Hernan Castillo-Hermosilla, Purdue University

• Brenda Cavalcante Matos, Purdue University

• Tugce Karatas, Purdue University

• Shahnaz Safitri, Purdue University