"Hedges, Mottes, and Baileys: Causally Ambiguous Statistical Language can Increase Perceived Study Quality and Policy Relevance"
Co-authors are Assistant Professor David William Braithwaite (Florida State University), Hugues Lortie-Forgues (Loughborough University), and Melody Moore, Mayan Castro, and Associate Professor Elizabeth Martin, all from the UCI Department of Psychology & Social Behavior.
Alvarez is specializing in Human Development in Context (HDiC) for her doctoral studies. Her research interests include fadeout of educational interventions, child development, mathematics learning, and increasing the representation of minority groups in STEM fields She is advised by Bailey.
Wan, who is advised by Bailey and Distinguished Professor Jacquelynne Eccles, also is specializing in Human Development in Context. His research interests include academic motivation and achievement, educational intervention, individual and gender differences in STEM fields, mathematical and spatial cognition, and career development.
Bailey's research focuses on understanding the longitudinal stability of individual differences in children’s mathematics achievement and on the medium- and long-term effects of educational interventions.
There is a norm in psychological research to use causally ambiguous statistical language, rather than straightforward causal language, when describing methods and results of nonexperimental studies. We hypothesized that this norm leads to higher ratings of study quality and greater acceptance of policy recommendations that rely on causal interpretations of the results. In a preregistered experiment, we presented psychology faculty, postdocs, and doctoral students (n=142) with abstracts from hypothetical studies. Abstracts described studies’ results using either straightforward causal or causally ambiguous statistical language, but all concluded with policy recommendations relying on causal interpretations of the results. As hypothesized, participants rated studies with causally ambiguous statistical language as of higher quality (by .48-.59 SD) and as similarly or more supportive (by .16-.26 SD) of policy recommendations. Thus, causally ambiguous statistical language may allow psychologists to communicate causal interpretations to readers without being punished for violating the norm against straightforward causal language.