Integration of Thematic Coding, Bayesian Methodology, and the General Linear Model

• In Event: 63.072 - Conceptualizing and Utilizing Mixed-Methods Research

Abstract

Bayesian approaches to prediction entail the use of context, which requires the application of prior probabilities when estimating the chance some event will occur. It is often the case that researchers might not be able to work with empirically validated estimates and instead rely on contextual information. Attending to context and subjectivity is well covered ground for qualitative researchers; therefore, we contend that adopting a mixed methods approach to Bayesian analysis can have a role in improving prediction. Therefore, this article focuses on how to mix qualitative methods with Bayesian approaches to prediction. At the analysis stage, once thematic coding is completed, it is possible to quantitize such information and then analyze it using the general linear model.

Authors

• Isadore Newman, Florida International University

• Anthony J. Onwuegbuzie, Sam Houston State University

• John H. Hitchcock, Indiana University - Bloomington