Estimating the outcome of a pregnancy test: women's judgements in foresight and hindsight.

Abstract

Previous research on judgement under uncertainty has suggested that, when we know the outcome of some event, we perceive that outcome as more likely than when we do not have outcome knowledge. That is, in comparison with judgements made in foresight, judgements made in hindsight are biased in the direction of the outcome the judge believes to have happened. While the effect appears to be robust in the laboratory, it has very seldom been tested in real life. This experiment therefore went outside the laboratory, and examined women's estimates of the outcome of a pregnacy test. It was predicted that those who knew the result of their test (hindsight) would perceive that outcome as more likely than those asked to make the estimate before they knew the result (foresight). The prediction was supported only for women whose result was positive and, furthermore, the positive group made consistently higher estimates than the negative group, both in hindsight and foresight. The findings were therefore less marked and more complex than in previous laboratory research, and support the argument that experiments and materials must be constructed with salience for the subjects. The findings are interpreted in the light ot Tversky & Kahneman's (1974) work on heuristic rules of thinking.