The IBM Answer Sheet as a Major Source of Variance on Highly Speeded Tests

THE EXPERIMENTAL design of a current research project called for the production of four highly speeded tests, each pair of which had some process in common. One of the tests required the examinee to read a clock dial (hour hand only) and to mark an IBM sheet; another test required writing in the answer. Two other tests required the examinee to read pairs of numbers, one of which was to be either marked on an IBM answer sheet or to