Optimizing information-transfer rates for human communication

Abstract

This research is concerned with investigating the factors that contribute to optimizing information-transfer (IT) rate in humans. This issue has important ramifications for the design of human-machine interfaces in a wide variety of applications, including virtual environment and teleoperator systems as well as sensory aids for persons with impaired vision and/or hearing. Based on the results of several early studies from the 1950s, a general rule of thumb has arisen in the literature which suggests that IT rate is dependent primarily on the stimulus delivery rate and is optimized for presentation rates of 2–3 items/second. Thus, the key to maximizing IT rate is to maximize the information in the stimulus set. Recent data obtained with multidimensional tactual signals, however, appear to be at odds with these conclusions. In particular, these current results suggest that optimal delivery rate varies with stimulus information to yield a constant peak IT rate that depends on the degree of familiarity and training with a particular stimulus set. We discuss factors that may be responsible for the discrepancies in results across studies including procedural differences, training issues, and stimulus-response compatibility. [Research supported by Grant No. R01-DC00126 from NIDCD, NIH.]