Outcome proportions, numeracy, and attribute-framing bias

Objectives: An object presented positively is often judged more favourably than the same object described negatively even when the descriptions are logically equivalent. This difference, termed the attribute-framing bias, has been shown to be affected by numeracy, such that less numerate individuals are more susceptible to the bias than highly numerate individuals. This article examines the hypothesis that less numerate individuals are less attentive to numerical information than highly numerate individuals; hence, their judgements rely more heavily on positive and negative words that elicit the bias. Method: In two experiments, participants’ numeracy was measured, and they were asked to rate different scenarios while attribute framing was manipulated by presenting logically equivalent scenarios described with either positive or negative outcomes in three different proportions. Results: Significant attribute-framing effects were found in both experiments. Critically, less numerate participants were as sensitive to different outcome proportions as highly numerate participants. Nevertheless, numeracy affected attribute framing as hypothesised: Less numerate participants were typically more susceptible to attribute-framing bias than highly numerate individuals. Conclusions: Individual differences in sensitivity to numerical information did not modulate the effect of numeracy on attribute framing. We discuss the implications of these findings to understanding the cognitive processes underlying attribute-framing bias.