A new approach for considering a dual‐role factor in data envelopment analysis

The conventional data envelopment analysis models deal with dual‐role factor as non‐discretionary (uncontrollable) factor. However, there might be dual‐role factor which is under control of decision‐maker. In addition, despite the fact that there are several publications addressing dual‐role factors, it seems that their idea of classifying a factor as an input or an output within a single model has a limitation. They also do not consider non‐zero input and output slacks and cannot fully measure the inefficiency of decision‐making units. To resolve these limitations and to consider dual‐role factor as well, this paper proposes a slacks‐based measure model which does not consider dual‐role factor as a non‐discretionary factor. To compare the results of proposed approach with conventional model, a statistical analysis is run. A numerical example demonstrates the application of the proposed method.