Categories

Categories are analytic units developed by qualitative researchers to conceptually organize findings related to a phenomenon or human experience that is under investigation. The novice researcher typically makes few distinctions among analyzing for themes, categories, and codes. If one were to think about a micro-, meso-, or macro-level analysis, coding starts at the micro level, the generation of categories moves the investigator to the meso level, and themes that bear out lessons learned or truths that reflect the findings are indicative of a macro-level analysis. Although themes, categories, and codes might build on each other, there are distinct differences.

The qualitative researcher almost always arrives at some point in the research process when developing categories is necessary. Category development can be done either inductively or deductively. To generate categories inductively, the researcher approaches data analysis without a preset list of categories and analyzes the data to identify analytic units that conceptually match the phenomenon portrayed in the data set. When categories are generated deductively, they emerge not from the data but rather from prior studies, relevant literature, research questions, and the researcher's own experience with and knowledge of the phenomenon. Under this approach, there is a chance that the categories generated from other sources will not be relevant or accurately reflect the qualitative data set at hand. Although inductive and deductive approaches work, both novice and senior researchers fluctuate back and forth between inductive and deductive analyses because blending the two methods helps the researcher to fully interrogate the data.

Recalling that both qualitative and quantitative researchers use categories when designing research studies, it is important to remember that there are various data sources that can be used to develop categories for a qualitative study or qualitative component of a mixed-methods study. Sources include formal and informal interview data, focus group interviews, observation notes, emails, journals, newspapers, primary documents (e.g., memos, internal organizational reports, diaries), and open-ended questionnaires. Potentially, a single study may have multiple data sources that must be analyzed before the researcher can reach any plausible conclusions.

Given the potential for diverse data sources that could be critical for a single study, constructing well-defined categories can be overwhelming for the researcher who must wade through what may appear to be mountains of data. The work is tedious and time-consuming but has the potential to yield great insights. Computer software is available to help keep qualitative categorical data analysis manageable. Qualitative software is helpful in managing and organizing categories. Although some programs tend to be more linear and less flexible in their capacity to reflect nonlinear findings, overall they are very instrumental in managing large qualitative data sets.

One benefit of using qualitative software (e.g., ATLAS.ti, NVivo) to conduct categorical analysis is that doing so gives multiple research team members the ability to collaborate during this phase of the research [Page 72]process. As more researchers gain a growing appreciation for qualitative methods, investigators from different institutions, disciplines, countries, and cultures will form more collaborative efforts that require multiple analyses examining concepts at the categorical level.

...