A Two-stage Evaluation Model for IT Investment Based on Interval DEA

Data envelopment analysis (DEA) is a very useful approach to evaluate information technology(IT) investment efficiency, but the original DEA approach can only measure the IT investment efficiency on one specific stage when a multi-stage business process is present, and it cannot solve the problems of imprecise data. Thus, the paper improves the two-stage model of Wang and Cheng, proposes a two-stage interval DEA model. The new model uses interval data instead of exact data, divides the investment into two stages and uses two different weights in two stages. The paper supposes the two stages give different contributions to the business performance, which reflects the decision makers' subjective wishes. The paper also shows that this non-linear programming can be treated as a parametric linear programming, and gives the solving procedure in detail. In the end, the approach is illustrated with an example taken from previous studies, and the results are more detailed and objective than the previous results.