Why Do We Remember Our Choices as Better Than They Were?

After choosing between two similar options, people often remember more positive features as belonging to the selected option and more negative features as belonging to the rejected one. This is choice-supportive misremembering. It need not involve inventing an entirely new fact. More often, a feature is remembered but assigned to the wrong source.

In experiments by Mather, Shafir and Johnson, participants chose between paired options and later identified their features. They were more likely to attribute positive features to what they had chosen and negative features to what they had rejected, including features that originally belonged elsewhere or were introduced later. The effect is not merely ‘I like it’; it becomes ‘I remember that it was better from the start’.

Suppose you compare two computers and select the one with longer battery life. Months later, you may correctly remember reading that one screen was brighter, yet wrongly attach that feature to the model you bought. The earlier decision then looks neater and more rational in memory than it did at the time, and that reconstructed account may shape the next purchase.

This does not show that every decision is rewritten. Much of the evidence comes from controlled option sets and delayed-memory tasks, and the size of the bias varies with whether the person made the choice, the delay, the similarity of the features and source-memory ability. My view is that memory here serves continuity of action more than neutral auditing. When a later review genuinely matters, a contemporaneous comparison is more reliable than ‘I remember why I chose it’.

https://pubmed.ncbi.nlm.nih.gov/11273420/
https://pmc.ncbi.nlm.nih.gov/articles/PMC5723021/