This paper is concerned with the problem of “How can an organization that develops computer products ensure good human factors in those products?” Thomas provides his definition of “human factors” and outlines reasons for concern. He points out that there are tradeoffs associated with good human factors, and that the tradeoff with cost today is often based on last year’s costs. This frequently means that systems are less friendly than they might be in order to save a few dollars, a few bytes, or a few machine cycles. The author stresses the need to develop systems for the way people are, rather than the way system designers wish them to be. The author suggests that good human factors will cost an organization less in the long run and will aid it in turning out a better product. Good human factors design will probably increase the design time and cost, but will also increase the value of the product and reduce redesign. When the primary user of the system is computer naive, good human factors may mean the difference between successfully using the system and abandoning it. This is important because many nonDP professional users can make discretionary use of a computer system. Thomas emphasizes the need to test a new system on representative users before submitting the system to general use.

Thomas’ comments and ideas are pragmatic and could be of use to those who design or modify computer systems. The paper provides good background reading on the problem. Tighter editing and fewer repetitive examples would have been better. The paper is a bit too much of a sales pitch for human factors, and the author is too evangelical in his presentation style.