In the study of differential abilities in learning and in the diagnosis and remediation of learning disabilities, much attention has been paid to the mode in which material is presented or becomes available (input), and to the form in which response is made (output). Following common practice, the input modes used in this study are Auditory, Visual, and Haptic (tactile-kinesthetic), and the output modes are Vocal and Motor. Different investigators have stressed either the input modes or the output modes as critical in the development or amelioration of learning disabilities. Reports reaching the writer from teachers and clinicians have described learning problems that seem unexplainable in terms of such input or output effects. This study advances and tests the proposition that, important as input and output modes may be separately in accounting for children's performances, there are children for whom input and output modes operate interactively instead of (or as well as) independently. The task chosen for the experiment was one of memory for object sequences. It involved presenting, under each of the three input conditions separately, sequences of familiar objects increasing in number from two to eight. The subjects were 90 Grade I Vancouver, B.C., children. Following each sequence, every subject was required to indicate the objects that had been presented. This was done independently under both vocal and motor output conditions. To increase reliability, three sequences were presented at each sequence-length. Each child was given two replications on all input-output combinations, with an interval of two weeks. Scoring was designed to take account of both the length of sequence that a child could recall correctly, and the number and kinds of error that he made under each combination of input and output modes. The resulting scores were viewed as entries in a four-factor experiment having two fixed factors (3 levels of Input and 2 levels of Output) and two random factors (90 subjects and 2 replications). Standard ANOVA procedures reveal, as hypothesized, a highly significant Subjects x Input x Output interaction. There is also a highly significant Subjects x Input interaction and a less impressive but nonchance Subjects x Output interaction. Estimates of variance components associated with each of these effects show the S x I x 0 interaction to account for about as much variance as the total of the two-way interactions, giving a clear indication of the potential importance of Input x Output combinations in learning diagnoses. The scoring patterns of individual children were analyzed. Certain children were found to have performed particularly well or badly under specific combinations of input and output modes that seemed to be unrelated to whatever input or output strengths or weaknesses they had. The implications of this interactive role of input and output modes were explored and resulted in recommendations for teachers and clinicians, as well as for further research. / Education, Faculty of / Educational Studies (EDST), Department of / Graduate