ResearchPublications

“In summary, many readers do not perceive major differences between the terms “predicts,” “increased with,” and “affects” with regard to causal attribution. Participants’ beliefs about the causal implication of a claim are also associated with prior beliefs about the relationship between the variables. Our study should be interpreted with an understanding of its limitations. We used a convenience sample, limiting the generalizability of results. We only used three contexts and cannot make more general claims about perceived causality; a conservative interpretation would be limited to these three specific scenarios. For feasibility reasons, we limited each participant to only 1 of 5 linking terms. Future studies with larger numbers would increase precision and might explore possible statistical interactions between linking words and context. We used a standardized format for all questions (e.g., linking terms always occurred after a clear and simple description of results). Other formats (e.g., leaving out a clear description of results) may have led to more heterogeneity. Future research should examine how background, training, and confidence in prior beliefs affect interpretation, and how different wording strategies might improve clarity.” (p. 3)