What Qualitative Methods Are For

A survey can tell you that 42% of young men in a given neighborhood report having been stopped by the police; it cannot tell you what the stops feel like, how they reshape the walk to school, or how they enter into the moral world through which a sixteen-year-old interprets his own future. A regression can estimate the effect of having an eviction on a tenant's court record on their probability of future homelessness; it cannot show you what it is to pack a household's possessions into black garbage bags on a Milwaukee sidewalk in January, or how a landlord's decision is reached across a kitchen table five minutes before the notice is served. These are not shortcomings of quantitative methods — they are simply questions that operate in a different epistemic register, and that register is what qualitative methods are for.

The register is one of meaning, mechanism, context, and process. Qualitative methods ask what a given practice means to the people engaged in it; how a patterned outcome is actually produced at the micro scale where human action occurs; what surrounding conditions make the mechanism run as it does; and how the phenomenon unfolds through time rather than in a static snapshot. These questions are not ornamental additions to a quantitative account. They are often prior to it — we cannot sensibly measure what we have not yet learned to see — and often posterior to it, because an estimated effect has to be interpreted through some substantive account of what is happening to the people being counted.

The intellectual inheritance is old. Max Weber, writing in the early twentieth century, argued that sociology must aim at Verstehen — interpretive understanding of social action from the standpoint of the actor (Weber 1922, pp. 4-9). Social action, for Weber, was action oriented to the subjective meanings of others, and to study it was to reconstruct those meanings with enough discipline that the reconstruction could be checked. Verstehen was not a retreat into empathy; it was a method of causal explanation. One cannot explain why a Calvinist merchant in seventeenth-century Amsterdam ploughed his profits back into the business rather than spending them on feasts without reconstructing the religious anxiety that made methodical accumulation a sign of election (Weber 1905/2001, pp. 102-125).

The Chicago School of urban sociology, formed around Robert Park and Ernest Burgess in the 1910s and 1920s, institutionalized the fieldwork tradition. Park, a former newspaper reporter, famously told his students to 'go get the seat of your pants dirty in real research' — to spend time in rooming houses, dance halls, and street corners and write what they saw (Park and Burgess 1925, pp. 1-18). The resulting monographs — Anderson on the hobo, Thrasher on gangs, Cressey on the taxi-dance hall — set the template for generations of ethnographic work.

A useful clarification at the outset: 'qualitative' is not synonymous with 'non-numerical,' nor with 'exploratory,' nor with 'soft.' Qualitative researchers routinely count, measure, and cross-tabulate, just not as the principal form of evidence. They test hypotheses, discard hypotheses that the data will not support, and produce findings that can be and are contested on empirical grounds. The difference from quantitative work lies in the kind of evidence that does the primary work — recorded speech, observed practice, documentary text — and in the corresponding analytical techniques, not in any lowered standard of inference.

One more framing matters. Qualitative methods are plural. Ethnography, in-depth interviewing, and textual or archival analysis have overlapping commitments but different data, different skills, and different characteristic pitfalls. The next pages take each in turn, then consider what rigor means across the family, and finally how qualitative and quantitative tools fit together in contemporary practice.