The following memorandum is a paper that was read at the convention of the American Psychological Association in September 1952. It reports an experiment in which the analysis of covariance method was used to determine the effects of different sets of directions for identical sets of test items. Two pairs of tests were compared, using the same population and predictor test. In the first pair of tests, Identical Blocks, the differences in test directions did not produce any observable difference in the score distribution. In the Paper Folding tests, however, the score distributions indicated that substituting pictures for verbal explanation made the test about one raw-score point easier. Four tables are added in an appendix to this memorandum in order to record the data from which these conclusions were drawn.