An information processing approach was applied to the development and validation of a test battery intended for personnel selection, classification, and guidance; and design of training programs. The approach specifies that test should measure specific cognitive processes and basic abilities, rather than prior experience. Tests should be short, easy to administer, statistically independent, and reliable. Taken as a whole, the test should not be designed to predict ability or interest in a particular job. The following tasks were selected from the literature on memory, psycholinguistics, and visual information processing for possible inclusion in the battery: (1) letter classification; (2) lexical decision making; (3) graphemic and phonemic analyses; (4) short term memory scanning; (5) memory scanning for words and categories; (6) linguistic verification; (7) recognition memory; and (8) semantic memory retrieval. The tasks were administered twice to 54 college students. Results replicated previous experimental findings with similar paradigms. Statistical data on reliability, construct validity, practice effects, and response distributions established the adequacy of all eight tasks as measures of individual differences. All were selected for inclusion in the battery. (Sample items from each task are appended.) (CP)