Zickmund SZ (VA Pittsburgh Healthcare System, Center for Health Equity Research and Promotion), Hanusa BH
(VA Pittsburgh Healthcare System, Center for Health Equity Research and Promotion)
Health services researchers increasingly employ qualitative methods. Yet these qualitative findings can be difficult to integrate into the other quantitative data that a study can produce. Researchers are now exploring mixed method approaches to convert qualitative information into quantitative data. Yet rarely are such data incorporated into the final statistical model. Barriers include the difficulty of transforming themes into quantifiable variables and the confusion that can exist over how to treat qualitative themes statistically. To reduce these barriers, this workshop will focus on: (1) how to tranform qualitative themes into variables; and (2) how to integrate qualitative findings into statistical models.
In the first half of the workshop, we provide examples and guidelines for translating qualitative findings into empirical data. Specific techniques include: (1) the development of codebooks that facilitate the transformation of qualitative themes into numeric datasets; and (2) the use of computer programs for the recording, storing, and transferring of qualitative codes into a spreadsheet format. Attention will also be given to methods for retaining and organizing textual quotations that are critical to the development of a thematic analysis. In the second half of the workshop, we provide the statistical techniques needed to describe, compare, and model qualitative data. Specific techniques include: (1) learning when qualitative results function as outcome or predictor variables; (2) determining whether the qualitative variables are predictors with equal value to the quantitative variables; and (3) demonstrating specifically how to integrate qualitative variables into statistical models.
Investigators and methodologists who are interested in integrating qualitative and quantitative datasets.
Assumed Audience Familiarity with Topic: