A major problem which confronts the behavioral sciences is the lack of a unifying set of dimensions for describing human tab]'-, performance. The absence of such a system limits the ability to relate human performance observed in one task to that observed in similar tasks. There is a need for a well-defined task-descriptive langnage for use by those who must apply the results of research to operational tasks. This report describes one of revcral approaches under development as part of a larger

more »

... rogram: the approach is concerned with developing a task classification system based upon known parameters of human performance. The human abilities, upon which this system was based, were derived primarily from the reported factor analyses of human performance in the cognitive, psychomotor, physical, perceptual, and sensory areas. Definitions of the abilities were developed together with rating scales for each ability. A series of pilot studies then were undertaken with the objective of producing an instrument which would have high reliabilit/ in classifying human tasks. During these exploratory studies, the initial set of human abilities was modified, the definitions of the abilities were revised, and the rating technique was improved upon. In addition, the studies examined various methods of analyzing the reliability data, and compared two methods of anchoring the rating scales. The results of this pilot research indicated that it was possible to develop a set of reliable, ability-based scales for classifying tasks, although more work will be needed. Future research on a human ability approach to classification will continue with the investigation of the problems of scale reliability and will initiate research on questions of the validity of the class if icatory instrument. ■ 111 ■ i TABLE OF CONTENTS Section