Memory & Attention · Quick Test
Hindsight Bias
Do you remember what you answered before you knew the answer?
The experiment
Make a guess for each question, even if you don't know. The right answer appears right after.
What happened?
You made guesses, saw the right answers, and then tried to remember your guesses. In the table, the “Memory” column shows whether each memory matched your guess, moved closer to the answer, or moved further away. In studies, wrong memories tend to land closer to the right answer: people think they guessed better than they did. With only a few questions, a single result can go either way, because memory sometimes errs by chance.
Why does your brain do this?
Often the exact memory of a guess is not stored. When it fails, the mind does not hand back a blank: it rebuilds a plausible number from what it knows now. And what you know now includes the right answer, which acts as an anchor and pulls the rebuilt number toward it. Since the rebuilt number feels like a memory, you don’t notice it was influenced.
When does this shortcut work well?
For almost everything in life, knowing the correct information matters more than remembering what you used to think. A memory that updates with what you have learned leaves you ready for next time, without carrying wrong ideas around. Some researchers see the bias as a side effect of that learning.
Where does this show up in daily life?
“I knew it would turn out like this,” after an election, a game, or a decision that went wrong. The bias makes an outcome seem more predictable than it was, and it can make us unfair to people who decided without knowing how things would end, such as doctors, judges, or the coworker who bet on a project. Writing down your prediction before you know the outcome is the simplest way to check later.
Nerd Mode
Open Nerd Mode
Baruch Fischhoff opened up the topic in 1975. In one study, people read a description of a historical episode and rated the chance of each possible outcome. Those who were told which outcome had happened gave it a higher chance than those who were not told, even when instructed to ignore that information. That same year, Fischhoff and Beyth asked people for predictions about President Richard Nixon’s 1972 trips to China and the Soviet Union and, after the trips, asked them to remember what they had predicted. Their memories landed closer to what actually happened.
This page uses the design of Wolfgang Hell and colleagues (1988), with numerical estimates. In their study, 59 students answered 88 questions, got the right answers to half of them right afterwards and to the other half a week later, and then tried to remember their guesses. In 35.1% of cases, people remembered their exact guess. The size of the bias was measured with an index: 0% when the memory matches the guess and 100% when it matches the right answer. The medians per condition ranged from 1% to 22%, and asking for a reason for each guess reduced the bias.
The effect also shows up within a single session. Daniel Bernstein and colleagues (2011) tested 194 people aged 3 to 95 in a one-hour session, with other tasks and a countdown between the guesses and the memory test. Every age group showed the bias, which was stronger in young children and older adults. A 2004 meta-analysis (Guilbault and colleagues) pooled 252 effects from 95 studies and found an average size of d = 0.39. In the meta-analysis, studies that tried to reduce the bias found no significant reduction; warning people about it has also had little success.
Studies compare questions whose answer was shown with control questions whose answer was not, to separate the bias from ordinary forgetting. This version has no control questions and only a few questions, so one person’s result is an illustration, not a measurement. The percentage is the median of Hell and colleagues’ index over the questions where the guess was not exactly right.
Technical terms
- Memory
- Hindsight
- Numeric estimation
Discovery
Described by Baruch Fischhoff in 1975.
Level of evidence
Robust evidence
Replicated in multi-lab studies or consistent meta-analyses.
Basis: Guilbault, R. L.; Bryant, F. B.; Brockway, J. H.; Posavac, E. J. (2004). A meta-analysis of research on hindsight bias.
References
- Original studyFischhoff, B. (1975). Hindsight is not equal to foresight: The effect of outcome knowledge on judgment under uncertainty. Journal of Experimental Psychology: Human Perception and Performance, 1(3), 288–299.
- Original studyFischhoff, B.; Beyth, R. (1975). “I knew it would happen”: Remembered probabilities of once-future things. Organizational Behavior and Human Performance, 13(1), 1–16.
- Original studyHell, W.; Gigerenzer, G.; Gauggel, S.; Mall, M.; Müller, M. (1988). Hindsight bias: An interaction of automatic and motivational factors? Memory & Cognition, 16(6), 533–538.
- ReplicationBernstein, D. M.; Erdfelder, E.; Meltzoff, A. N.; Peria, W.; Loftus, G. R. (2011). Hindsight bias from 3 to 95 years of age. Journal of Experimental Psychology: Learning, Memory, and Cognition, 37(2), 378–391.
- ReviewGuilbault, R. L.; Bryant, F. B.; Brockway, J. H.; Posavac, E. J. (2004). A meta-analysis of research on hindsight bias. Basic and Applied Social Psychology, 26(2-3), 103–117.
- ReviewBlank, H.; Musch, J.; Pohl, R. F. (2007). Hindsight bias: On being wise after the event. Social Cognition, 25(1), 1–9.
Review
This text has not been reviewed by a volunteer yet. Want to help review it?
Found a mistake? Write to [email protected]