Significance-testing the validity of idiographic methods: a little derangement goes a long way.

We present an experimental design for validating idiographic data and the mathematical basis for subjecting this to statistical hypothesis testing. The method involves matching descriptions of four or more objects to the people from which they arose. If four or more can be matched, this is unlikely (p < 0.05) to have arisen by chance alone and the null hypothesis that the data communicate no information can be rejected. This criterion score of 4 is sufficient for statistical significance regardless of the total number of objects on which matching was attempted. The mathematics of this somewhat counterintuitive result is given, and we argue that this method is of value in validating idiographic data. Its relationship with the 'method of mismatched cases' is discussed.