A psychologist gives a group of participants 20 personality tests. He looks at the correlations between each test and each of the other 19 (a total of 180 correlation coefficients). He finds two correlations that are statistically significant-that is, that his statistical tests tell him would not occur more than five times out of 100. What should the psychologist conclude from his study?

Question

Quizplus · Accepted Answer

Given the large number of correlations calculated (180), finding two statistically significant correlations could easily occur by chance, especially considering the common alpha level of 0.05 for statistical significance. This means that, by chance alone, about 5% of the correlations could appear significant even if there's no real relationship.

A Psychologist Gives a Group of Participants 20 Personality Tests