It may seem irrelevant to dwell upon anatomical data, when the purpose of this chapter is to deal with mental aptitudes. The pertinence of the data cited, however, lies in the fact that if any sex difference in physical variability could be established, this would suggest (though it would not prove) the existence of a sex difference in mental variability also. No experimental studies have ever been made for the express purpose of determining whether there exist sex differences in mental variability. Such scattered data as we possess have come incidentally from studies made with some other chief purpose in view. Such data were collected and summarized in the _American Journal of Sociology_ for January, 1914. There was at that time very little evidence that could be cited on this subject, but such as there was gave no ground for maintaining the existence of any sex difference in variability. Since 1914 Trabue's experiments, with "completion tests," performed on about 1,300 school children, have been published; the Courtis arithmetic tests on several thousands of school children in New York have been made public; Terman has tested 1,000 unselected children by the Binet-Simon tests; and Pyle has undertaken his study in the measurement of school children. The evidence from these extensive experiments is in all cases that there is no sex difference in mental variability, as thus measured.
It is necessary also for the reader to bear in mind that there is as yet much controversy among those best equipped to understand the problems of variation, as to the proper methods of measuring comparative variability. The mathematical considerations involved need not be rehearsed here. But until it has been definitely determined just how comparative variability can be scientifically measured, it would seem premature to make any final statement as to sex differences in this respect.
We can therefore answer our second question thus: There is little or no agreement among those best qualified to speak, as to what constitutes the scientific method of measuring comparative variability. But according to the methods now deemed the most reliable, and according to those studies wherein presumably correct methods of measurement have been employed, there is no reason to suppose that there is any sex difference in variability, so far as the numerous traits tested are concerned. There has never been an experimental study made in which the sampling from both sexes was large, random, equal, and from groups of equal homogeneity socially and racially, that showed any reliable sex difference in variability. If we adhere to the literature of fact, we must conclude that, so far as we know, human females differ from each other as much as do human males in abilities and aptitudes.