Reevaluating the empirical support for a lens-based account: dissecting comparative fit and reconstructive category guessing

The lens-based account of intersectional stereotyping (Petsko et al., 2022) proposes that perceivers put on social “lenses” for categorizing their environment that exhibit comparative fit with observed behavior in a given situation. Support for this claim comes from studies using the Who-Said-What paradigm (Taylor et al., 1978), in which category membership was manipulated to covary with speakers' pro versus contra positions in a discussion along one of two crossed social dimensions. However, as noted by Klauer et al. (2014), this paradigm is vulnerable to a critical limitation: when category membership can be inferred from statement content, apparent categorization effects may reflect reconstructive category guessing rather than genuine category memory. Across three experiments ( N = 764), we tested whether comparative fit modulates social categorization beyond such reconstructive processes. We replicated strong comparative fit effects when category membership was inferable from discussion statements, yielding large effects comparable to those reported by Petsko et al. (2022). Crucially, when participants responded to discussion-unrelated statements that no longer allowed inference of category membership, these effects disappeared, indicating that prior findings were substantially inflated by reconstructive guessing. In a final experiment, we introduced an alternative source of situational fit by spatially aligning speakers with opposing discussion positions during encoding of discussion-unrelated statements. Under these conditions, a significant but markedly smaller comparative fit effect emerged. Our findings suggest that comparative fit can enhance social categorization along the fitting dimension beyond reconstructive guessing. However, this situational lens is rapidly “put off” once the fit is no longer present.