Abstract
The rapid advances in technology with declining costs of hardware is expected to widen the market for computers. One of the interesting balances for vendors to achieve is keeping users happy with increases in price performance, usually with lower unit costs. One way to achieve this objective is by satisfying users and meeting their needs. Users have many different needs, however, and it may be very difficult to prioritize them since many of the needs are partially mutually exclusive. For example, the promotion of one need, low response time, would be at the expense of another need, low cost. Therefore, mini/micro vendors and users could use a model that will prioritize needs and measure how satisfied users are with specific computer systems as compared to others. Similarly, the same information may be very useful to current and future computer users for system selection. Accordingly, the main purpose of the present study is to provide vendors and users of mini/micro computers with a discriminant model. This model classifies correctly 93% of the successful systems. Thus it can be used for system selection by users and for system design by vendors.
