논문

This paper concerns a multi-attribute value analysis procedure to rank
alternatives and to select the most preferred alternative. This procedure is
appropriate when there are multiple, conflicting attributes and no uncertainty
about the outcome of each alternative. To conduct a multi-attribute value
analysis, it is necessary to determine a value function, which combines the
multiple evaluation measures into a single measure of the overall value of each
evaluation alternative. The form of this function that is used in this paper is a
weighted sum of functions over each individual evaluation measure. Determining
a value function requires single dimensional functions and weights.

The case is on selecting a network strategy. An organization is deciding what
strategy to follow with respect to networking its personal computers. The
organization has selected four evaluation measures: productivity enhancement,
cost increase, computer security and user satisfaction. Cost increase is the net
present value of the increased cost for an alternative relative to the current
situation, while productivity enhancement, security and user satisfaction are
contrived to be constructed scale.

The best alternative is selected by means of value of single dimensional value
function and a sensitivity analysis was conducted to determine the impact on the
ranking of alternatives of changes in various model assumptions. Sensitivity
analysis that is used in this paper is on the weights. These weights represent
the relative importance that is attached to changes in the different evaluation
measures.