A Unique Mixed-Methods Design for the Analysis of Large Data Sets

• In Event: AERA Roundtable Session Saturday 11:40 am Four Seasons Ballroom 2-3
  In Roundtable Session: From Quantitizing to AI: Exploring Current Issues in Mixed-Methods Research Design (Table 4)

Abstract

Debate surrounds the “quantitizing” of qualitative data. Some scholars reject data conversion while others suggest that it facilitates pattern recognition. Often researchers resort to quasi-statistical words such as “more,” “most,” and “all” when presenting qualitative findings. This paper explains a unique research design using a QUAN→qual pragmatic lens that can identify significant differences while maintaining contextual integrity. We first used validated word count software to compare text to user-defined software dictionaries. We then used principal components factor analysis and multivariate modeling to find statistical differences. Finally, we cross-referenced the most frequent words to the dictionary to identify actual text. This mixing of linguistic and qualitative software programs is a tool to identify significant findings in non-numerical text.

Authors

• Barbara Lee, Keiser University

• Carol A. Isaac, Mercer University