Extended Data Fig. 2: Effects of interventions to reduce recall bias in study 2.
From: Historical narratives about the COVID-19 pandemic are motivationally biased
a–f, Linear regression predicting individual recall on the basis of past (December 2021) and present (January 2023) ratings and their interactions with vaccination status and experimental condition for infection probability (a), infection severity (b), trust in government (c), trust in science (d), mask wearing (e) and avoiding contacts (f) (n = 3,105). Each line visualizes directional bias and how past and present perceptions affect recall (at recall = 4) for a given vaccination status and experimental condition (see the Fig. 1 legend for details on how to read the figure).
