Few studies of psychological tests which have been shown to differentiate between psychiatric groups have attempted to follow-up the patients originally tested so as to determine what the predictive power of the test might be in the light of subsequent events. One of these few studies has been reported by Walton (1958) and commented upon by Inglis (1959) who was able to show that the test Walton described was better able to predict the outcome of illness after a follow-up period of two years, than was the original diagnostic label. The importance of this kind of study has been pointed out most clearly by Payne (1958). His argument is that, “Description is only one implication of the diagnostic label. Can the test score aid the doctor in making a prognosis? This need not be the case. Let us consider the original validation of the test again. The doctor who diagnosed the standardization group of patients might well be able to give a more or less accurate prognosis for this group of patients. In fact there might be a significant relationship between the presence or absence of the label he assigns, and prognosis. We also know that there is a significant correlation between presence or absence of his label and the test scores. This does not prove, however, that there is any relationship whatsoever between the diagnostic test score and prognosis. Two things which correlate with the same thing do not necessarily correlate with each other, unless the correlations concerned are greater than ·7” (1958, p. 27).
The power of a test for seasonality.