Congenital Dislocation of the Hip

Our claim is that by this method we can obtain all that can be obtained by the use of the " Yes/No " method, but with greater precision. It appears from their statement, "For these purposes the simple Yes/No answer seems to work pretty well," that Drs. Fletcher and Tinker are willing to accept a low level of accuracy. Surely it is much sounder to gather data in the most accurate way practical, even though some may be discarded later. Drs. Fletcher and Tinker admit that their technique breaks down in "some borderline cases which are arbitrarily classed as negative," but when it is the patient who is left to resolve any ambiguities it is impossible to estimate how many decisions are borderline, so the amount of negative bias is unknown. Drs. Fletcher and Tinker suggest that their method is validated by a comparison of the results of two independent surveys on the "relationship between answers to questions and winter sputum volumes in two male populations" (Table VIII, May 27, p. 1495). By their technique the patient has three alternatives, " Not at all," " Part of -the time," and "All the time." It is apparent that the middle category contains all but the two extremes and so is much wider than the other two categories. Hence it is not surprising that the results of the two independent surveys show no disagreement for the two extreme categories but show a significant (p= 0.05) disagreement for the middle category. Had a continuous scale been used in place of the three categories by asking the question, " How much of the time do you bring up sputum ?" the results of the two surveys could have been compared with greater accuracy. With regard to the question of fallible memories, we fully recognize the dangers of retrospective surveys, but consider that any such objection applies equally to all questionary techniques. In the one investigation we have conducted in which it has been possible to estimate the importance of distortion due to the time elapsed between the events and their recall, it was found to be negligible compared with the main effects in which we were interested. However, there can be no general solution to this problem; it needs to be examined afresh in each investigation.-We are, etc., J. L. GEDYE. R. C. B. AITKEN. HELEN M. FERREs. R.A.F. Institute of Aviation Medicine, Farmborough, Hants.