Deviant Case Analysis Definition
Deviant case analysis is a methodological procedure used primarily within qualitative social research, particularly in conversational analysis (CA), involving the examination of specific examples or instances that do not conform to the rules or patterns currently being proposed by the researcher. Unlike quantitative methods, which often treat anomalies as statistical error or noise, deviant cases are actively sought out and valued because they can illuminate underlying normative structures within social interaction. This approach is rooted in ethnomethodology, which posits that deviations from expected behaviour reveal the often-invisible, taken-for-granted rules governing everyday life.
Context in Qualitative Sociology
The focus on deviant cases stems from the work of Harold Garfinkel and his concept of “breaches” in the mundane social order. In this context, an exception or deviation is not merely an error but an opportunity to reveal the ‘seen but unnoticed background features’ of social life. This methodology seeks to analyse how instances that run counter to established patterns illuminate the normative expectations participants operate under.
Application in Conversational Analysis (CA)
In conversational analysis, deviant case analysis involves scrutinising talk-in-interaction to identify rules of conversational sequencing, such as adjacency pairs or invitation-refusal sequences. Crucially, these conversational rules are not observed as statistical regularities or causal connections; rather, they are understood as normative patterns that participants orient themselves to. Consequently, deviations do not require separate coding schemes but reveal the actual operation of these norms—for instance, observing an absence of an expected second turn reveals the underlying norm of reciprocation.
Uses and Outcomes
Deviant cases serve several important functions in CA. Stephen Clayman and Douglas Maynard identified three primary uses: first, deviant cases may demonstrate the same normative orientations as standard cases; second, they can prompt a revised analysis that accounts for both the anomalous and the main set of data; and third, a case might exhibit an entirely different order, suggesting it belongs to a separate collection. For example, in analysing sequences of telephone openings, identifying a single instance where a participant breaks the expected rule allowed researchers to develop a more sophisticated theoretical account, such as the ‘summons–answer’ sequence, which accounted for the observed variation across the entire dataset.
Methodological Status
Deviant case analysis is fundamentally inductive, aligning with what conversation analysts term ‘unmotivated looking’. It functions as an essential corrective to the tendency in research to seek only confirming examples. Its special relevance lies in its ability to reveal normatively oriented patterns of social interaction rather than focusing solely on frequent or causal relationships, making it a vital ingredient for understanding the structure of everyday social behaviour.

