In this study, we develop an approach based on analytic
network process (ANP) and data envelopment analysis (DEA)
to evaluate the performance of Research and Development projects.
We identify a set of criteria and subcriteria having interdependencies
among themselves. Interdependency is treated using a hybrid
ANP model consisting of both hierarchy and network. The interval
pairwise comparison matrices are constructed in a group decisionmaking
process. ANP is extended to obtain interval weights from
the interval pairwise comparison matrices. Interval weights are formulated
as assurance region constraints in a super-efficiency DEA
model, through which the project ranking is obtained. The superefficiency
DEA model is also modified to handle missing data. The
approach is applied to a real case study on performance evaluation
of the ongoing projects in a Defense Research and Development
Institute.