Constant Comparison

The Process

Constant comparison can be thought of as a qualitative approach that resembles the quantitative methods of factor analysis or multiple regression in that every data bit is compared with every other data bit; however, two major differences exist. First, rather than a computer, the analyzing instrument is the researcher's brain. Second, as the theory begins to take shape, the researcher is free to alter her or his study question. For example, a researcher may enter the field looking at the experience of loss associated with home fire victims and find that those emotions are exacerbated by the rituals followed by their social nexus—rituals that can be connected or unconnected to what the victims need.

If the researcher is using interview and observational data, each episode is coded and compared with every other episode for similarities and differences to answer the question, “What is going on here?” In this method, data gathering and analysis go on simultaneously rather then at the end of data collection. The interview schedule and observation site will evolve as the developing theory begins to take shape. When the researcher is able to group analyzed data into categories, those categories are examined for how they are related to one another and then collapsed under a higher level category until the central category that explains most of variation in the data is revealed or “discovered,” as Glaser and Strauss put it. The process holds for other kinds of data, documents, and records as well as the work of other authors.

Memorandums. The mental calisthenics required of the researcher using grounded theory necessitates careful record keeping in the form of memos that represent the researcher's thoughts about the data, how the theory is coming together, and possible next steps. Later in the process these memos, through constant comparison, are sorted into categories and subcategories until it becomes clear that one central category holds the developing theory together and subcategories become properties of the central category. Sorted memos become the basis for the research report.

Theoretical Codes. Theoretical codes are regulators that transform descriptive data into an abstract or [Page 115]theoretical structure. There are families of codes; for example, Glaser wrote of the “six Cs”: context, condition, cause, consequence, covariance, and contingency.

Comparison With Existing Literature. A grounded theory is verified by its acceptance by the targeted population, but by comparing the new knowledge with existing theory, the researcher can place her or his theory within the context of knowledge development. For example, in a study of home loss by fire, Phyllis Stern and June Kerry looked for other works on loss rituals.

...