Proleader and antitrailer information distortion and their effects on choice and postchoice memory

• During choice, people distort information about the options’ attributes in two ways.
• Information about the leading (trailing) option is distorted positively (negatively).
• Proleader and antitrailer distortion have similar effects on choice.
• Predecisional distortion predicts postdecision memory for the original information.
• The memory bias is based on the tentatively leading option, not the chosen option.

In four experiments involving choices between apartments, we decomposed predecisional information distortion into positive distortion of information about the tentatively leading alternative and negative distortion of information about the trailing alternative(s). Proleader and antitrailer distortion were roughly symmetric, with similar magnitudes in all but one test. Additionally, proleader and antitrailer distortion predicted choice with similar coefficients in all but one test. Distortion predicted choice when we used participants’ own “true” preferences as the baseline for assessing distortion and when we considered only identical information items that did not distinguish between apartments (“true” preferences cancel out for such items). Finally, predecisional distortion of apartments’ attributes predicted participants’ postdecision memories for those attributes, with positive and negative distortions predicting corresponding memory errors. This effect appears to reflect the predecisional encoding of information about the leading and trailing alternatives rather than a response bias or other postdecision process favoring the chosen alternative.