Non-statistical, substantive generalization: lessons from Q methodology

ABSTRACT

Considerations related to generalization of a study’s findings are often interconnected to researchers’ judgements regarding the ‘quality’ of the methodology and methodological pluralism. Too often, researchers consider generalization as only possible with respect to quantitative studies with large numbers of randomly selected participants (statistical generalization). Recently, Levitt suggests qualitative research can be generalized to phenomenon. Like others, she differentiates the type of generalization typically identified with quantitative research, which requires large samples and randomization so that findings can be generalized to the population, with the type of generalization that can be achieved with qualitative research. In turn, Levitt proposes the concept of qualitative generalization, an idea suggested by other researchers. However, her assertions about generalization in qualitative research resonate with the assertions of Thomas and Baas about generalizability in Q methodology. Q methodology offers a unique blend of qualitative and quantitative principles to study subjectivity and includes factor analysis. Yet Q findings are about developing theories, much like qualitative research. Q researchers may discuss substantive inference, connecting generalizations about phenomenon to a population, rather than statistical inference about a population. We discuss acceptance of two types of generalization within and their importance in relation to research equity.

KEYWORDS:

• Q methodology
• generalization
• mixed research
• qualitative
• quantitative
• external validity

Disclosure statement

No potential conflict of interest was reported by the author(s).