niques for determining validities of test j items is the Vincent Overlapping method.1 By this method the validity value for any item is the per cent of those failing the I item who have higher criterion scores than | the median criterion score of those who pass the item. Consider, for example, a test | administered to 100 students, with 30 fail ing a particular item. If, of these 30, 6 have criterion scores higher than the median ? score of the 70 who get the item correct, the Vincent validity value for the item will be 6/30 of 100, or 20. The smaller this j measure of overlapping the greater is the I discriminating power of the item, and so the higher is its validity. Usually the critei rion score is simply the total score on the test of which the item forms a part. j The Vincent method makes rather inade
[1]
E. Lindquist,et al.
Experimental Procedures in Test Evaluation
,
1933
.
[2]
D. F. Votaw.
Graphical determination of probable error in validation of test items.
,
1933
.
[3]
T. Lentz,et al.
Evaluation of methods of evaluating test items.
,
1932
.
[4]
T. G. Thurstone.
The difficulty of a test and its diagnostic value.
,
1932
.
[5]
Carl Brigham,et al.
A study of error
,
1932
.
[6]
P. Symonds.
Choice of items for a test on the basis of difficulty.
,
1929
.
[7]
E. L. Clark.
A method of evaluating the units of a test.
,
2022
.