See How Easily You Can Avoid The Memory Bias

Will booking an expensive restaurant for the weekend make you happy?

To make decisions that will make us happy in the future we need to acknowledge how the memory bias can warp our predictions.

Today, making my plans for the upcoming public holiday sent my mind straight back four weeks. Then, heading out into the unseasonably warm spring weather, I had high hopes for a relaxing day off.

Unfortunately everyone else in London had exactly the same idea. The whole day my every plan for having fun was foiled: restaurants were booked up, bars were full to overflowing and when I finally did get sat down at a riverside restaurant, it started raining.

Perhaps this weekend I won’t bother. Or, then again, maybe I shouldn’t be too hasty…

When making decisions about the future, we naturally use events from the past as litmus tests. Our memories contain a huge database of experiences, all with emotions in tow, which help us work out what will give us pleasure in the future.

When trying to recall a past incidence of an event, people will naturally recall the worst instance, whether trying to or not. Unfortunately memory has all sorts of wicked tricks up its sleeve for deceiving us: it fades with time, can become blocked or be misattributed. Not only that, but psychological research reveals the type of memories we retrieve to make decisions about our future happiness are often biased to unusual examples that are either very positive or very negative.

Memory bias

A neat study by Dr Carey Morewedge from Harvard University and colleagues demonstrates how the memory bias works (Morewedge, Gilbert & Wilson, 2005).

Sixty-two subway passengers were randomly allocated to one of three groups. Each was asked to describe a time in the past when they had missed the train. But the question was asked in subtly different ways:

  • Free recallers were asked to describe any instance.
  • Biased recallers were asked to describe the worst instance.
  • Varied recallers were asked to describe any three instances.

Participant then indicated how happy or unhappy they were on those occasion(s).

The results showed that people in both the ‘free recall’ and ‘biased recall’ groups remembered equally depressing times when they had last missed the train. This suggests that when trying to recall a past incidence of an event, people will naturally recall the worst instance, whether trying to or not.

Participants in the ‘varied recall’ group, though, were more positive suggesting that out of the three events they had recalled, at least one of them was positive. Recalling more than one event, then, makes it more likely that at least one of them is more positive.

Predicting the future

After being primed with memories of past experiences of missing the train, participants were then asked to rate how unhappy they would be if they were to miss the train today. This was to test how the memory bias affected their prediction of their feelings in the future.

Surprisingly it was the free recallers made the worst prediction about how they would feel in the future, significantly worse than the varied recallers and the biased recallers. The reason it’s a surprise is that free recallers and biased recallers were both remembering past experiences that were equally bad, and yet the biased recallers made the lowest prediction.

Dr Carey Morewedge and colleagues explain that when people are explicitly asked to recall the worst event, they are then aware that it’s the worst event. In contrast, when people are allowed to recall any event they like (free recall) they do still recall the worst event, but don’t realise they’ve done so. As a result those in free recall make much worse predictions about how they will experience the same event in the future.

Bias emerges without prompting

Even when not specifically prompted to access past events people still display the memory bias.Two subsequent studies carried out by the same authors backed up these findings. In the first people demonstrated the memory bias when trying to predict positive events in the future. Given free reign people naturally recall an especially positive example of a particular event then go on to make much more positive predictions about the emotional effect on them of the same event in the future.

A third study extended the same findings to a more natural situation where one group weren’t asked to recall anything when making a prediction about how they would experience an event in the future. Nevertheless people still demonstrated the same memory bias for predicting future events. This tells us that even when not specifically prompted to access past events people still display the same bias.

Conquering the memory bias

This research suggests a straightforward way of conquering the memory bias. When trying to predict how you’ll feel at your friend’s party, eating at an expensive restaurant or that looming dental appointment, try the following:

  1. Explicitly recall more than one past event of that type. The events then average out, giving you a better prediction of how you will feel in the future.
  2. Be aware that if you only recall one past example of that type of event it is very likely to be either one of the best or one of the worst examples of that event. Simply realising this should be enough to negate the bias.

These two methods should bypass the memory bias and contribute towards decision-making that leads to greater overall happiness in the future.

[Image credit: Thomas Hawk]


Morewedge, C. K., Gilbert, D. T., & Wilson, T. D. (2005). How Remembering the Past Biases Forecasts of the Future. Psychological Science, 16(8), 626-630.

About the author

Dr Jeremy Dean is a psychologist and the author of PsyBlog. His latest book is "Making Habits, Breaking Habits: How to Make Changes That Stick". You can follow PsyBlog by email, by RSS feed, on Twitter and Google+.

Published: 23 May 2008

Text: © All rights reserved.

Images: Creative Commons License