Willem Wagenaar's study (chap. 10 of this volume) represents an excellent example of the value of carefully planned and meticulously collected naturalistic data. Because of the quality of his earlier study, he was able to readdress the data to consider whether his memory was biased against events that cast him in an unfavorable light. The data suggest that exactly the opposite is the case, and allow a range of alternative hypotheses to be firmly rejected. I myself am convinced by this analysis, even though the pattern of observations is exactly the opposite to what one might have predicted on the basis of common wisdom, and indeed considerable data (see Baddeley, 1990, pp. 379–406, for a review). I do, however, have some misgivings over the extent to which these findings would generalize, and even more concern over Wagenaar's theoretical interpretation of them.

As Wagenaar himself emphasizes, an inevitable limitation on single-case studies is the question of generality. Despite this danger, single-case studies in neuropsychology have typically proved to generalize, and have been justifiably influential. Is there any reason to suspect that the present study will be any less general? I think there may be. Many authors have suggested that there are major differences in the ways in which individuals cope with threatening or conflicting information, with some tending to minimize such evidence (repressors) while others tend to focus on the discrepancy (sensitizers) (Byrne, 1964; Davis, 1990). One would therefore like to see a number of equally well designed studies on other subjects before generalizing too widely from these findings. Such a concern is reinforced when one considers possible alternative mechanisms that might have produced the results obtained.