The dispute is one which cannot be settled by the assertion of opinion, and, unfortunately, thoroughgoing investigations have not yet been made as to the frequency and extent of errors made by untrained or partially trained examiners. The only study of this kind which has so far been reported is the following:--[37] [37] Samuel C. Kohs: "The Binet Test and the Training of Teachers," in _The Training School Bulletin_ (1914), pp. 113-17. Dr. Kohs gives the results of tests made by 58 inexperienced teachers who were taking a summer course in the Training School at Vineland. The class met three times a week for instruction in the use of the Binet scale. During the first week the students listened to three lectures by Dr. Goddard. The second week was given over to demonstration testing. Each student saw four children tested, and attended two discussion periods of an hour each. During the third, fourth, and fifth weeks each student tested one child per week, and observed the testing of two others. The student was allowed to carry the test through in his own way, but received criticism after it was finished. Twice a week Dr. Goddard spent an hour with the class, discussing experimental procedure. The subjects tested were feeble-minded children whose exact mental ages were already known, and for this reason it was possible to check up the accuracy of each student's work. Kohs's table of results for the trial testing of the 174 children showed:-- (1) That 50 per cent of the work was as exact as any one in the laboratory could make it; (2) That in an additional 38 per cent the results were within three fifths of a year of being exact; (3) That nearly 90 per cent of the work of the summer students was sufficiently accurate for all practical purposes; (4) That the records improved during the brief training so that during the third week only one test missed the real mental age by as much as a year. Since hardly any of these students had had any previous experience with the Binet tests, Dr. Kohs seems to be entirely justified in his conclusion that it is possible, in the brief period of six weeks, to teach people to use the tests with a reasonable degree of accuracy.