Scientific generalization in psychological inquiry: A concept in need of clarification.

Abstract

Although generalization is typically described in the methodological literature as both a goal and achievement within quantitative research traditions, qualitative research is often presumed to be incapable of and, in some cases, unconcerned with generalization. However, the concept of generalization is much less bounded than it might appear at first blush. In fact, it has multiple senses and applications, some of which are more applicable in some areas of research than in others. In this article, we present a conceptual analysis of generalization and related concepts. We begin with an analysis of ordinary language uses and then move to more "technical" uses within scientific discourse. We highlight accounts given in the general philosophy of science literature as well as prescribed and actual uses in psychological research within domains in which quantitative and qualitative methods, respectively, have historically been privileged. Finally, drawing from the different kinds of generalizations that feature within and across quantitative and qualitative research, we sketch some points that researchers might keep in mind when considering whether and how generalization plays out in their research.

Generalization Bias in Science

“…2 Specifically, the focus will be on generalizations that occur when scientists collect data from their study samples and extrapolate from them to broader populations of individuals or across contexts. These particular generalizations are pervasive in many empirical sciences and are one key part of scientific induction (Little, 1993;Slaney & Tafreshi, 2021). They will be the sole referent of the term "scientific induction" here.…”

“…While much has been written on scientific induction (e.g., Achinstein, 2010; Claveau & Girard, 2019; Cohen, 1970; Feeney & Heit, 2007; Little, 1993; Slaney & Tafreshi, 2021; Sloman & Lagnado, 2005), one fundamental assumption has gone largely unquestioned. It is that scientific induction is an “act of reasoning” (Polit & Beck, 2010, p. 1451), a voluntary process that scientists fully control rather than an automatic tendency that may often operate by default and affect scientists’ research and conclusions against their epistemic goals (e.g., the formation of true beliefs).…”

“…The account does not imply that scientists’ inductive generalizations are always beyond their control. Scientists make significant efforts to assess the generalizability of their results and commonly tailor their conclusions accordingly (Engel & Schutt, 2013; Kukull & Ganguli, 2012; Slaney & Tafreshi, 2021). The generalization bias account rather holds that scientific induction involves a default but modifiable extrapolation tendency that can and often does drive scientists to unintentionally overgeneralize.…”