Comparing a single case with a control sample: correction and further comment.

Crawford, Garthwaite and Howell [Crawford, J. R., Garthwaite, P. H., & Howell, D. C. On comparing a single case with a control sample: An alternative perspective. Neuropsychologia, in press] correctly point out that the t-test I suggested for asking whether a single case belongs in the population from which a control sample is drawn is incorrect. In fact, the corrected formula can be interpreted either as addressing this question, or as asking whether the single case and the control sample are drawn from populations with the same mean. Extension to analysis of variance is recommended for factorial experiments, where single case and control sample are tested under different experimental conditions. It is not recommended for studies in which different tests are administered and treated as different levels of the same dependent variable, as in tests for neuropsychological dissociations.