Negative Case Analysis

Negative case analysis is a central data analytic approach in qualitative methods and is essential to the rigor of most data analytic plans. It is seen in what grounded theorists call constant comparison procedure and what Michael Agar refers to as “breakdown” and “resolution.” Negative case analysis is necessitated by purposely sought or spontaneously appearing pieces of data that differ from the researcher's expectations, assumptions, or working theories. Although there is always some dread attached to the appearance of cases that appear to call into question one's carefully constructed analytic framework, negative cases are integral to strengthening findings. As Matthew Miles and Michael Huberman suggest in their seminal text on qualitative data analysis, the outcome of negative case analyses can run the gamut from refuting to refining findings.

Whether actively sought (“occasioned breakdown” in Agar's terms) or spontaneously occurring (what Agar calls “mandated breakdown”), negative cases are not a rare occurrence, but rather are a natural part of any study. It would be highly unlikely in real life for everything to fall exactly in line and act the same, and particularly unlikely in the early stages of analysis and hypothesis creation. Finding and understanding negative cases not only strengthens a good study, but these cases protect against researcher biases in what and how data are seen and reported. These negative cases can be, among other things, people, places, or events that differ in meaningful ways from other data points. Miles and Huberman describe how well-organized data displays can simplify the process of locating outliers by a simple scan of the data array.

The proper response to these negative cases is to seek to understand where and how these new data diverge from the rest and from the standing theory, to make the necessary revisions to the theory to include these unique findings, and then to test these revisions in the iterative manner that is synonymous with qualitative analysis. It is important to note that, unlike in traditional quantitative methods, in qualitative methods, outliers are neither ignored nor is the working theory necessarily rejected when countering evidence is found. In fact, Miles and Huberman caution not to discard the working hypothesis too quickly, pending analysis of the proportion of positive and negative evidence.

Even on the rare chance that no spontaneous negative cases appear naturally, most qualitative methods advocate that researchers actively search for disconfirming evidence. This search includes seeking outlying data that could disprove the working hypothesis or confirm an alternative one, intervening variables that might refute assumed causal relationships, as well as collecting new data from additional sources. Miles and Huberman also suggest seeking out a friendly skeptic to review your working hypotheses and data. It is only through actively seeking to test and refute one's findings and to explain not only the consistent, but also the inconsistent data that one can truly come to a final, rigorously defensible understanding of one's research findings.