The profit-maximization is always the main target for enterprises. For those not-for-profit organizations, such as public sectors and research institutions, although profit is not the key objective, it is still important for decision makers to know its scale-efficiency in order to maximize the utility of limited resources. DEA is a non-parametric method to measure relative efficiency of Decision Making Units (DMUs) with multiple inputs and multiple outputs. And DEA provides information about scale efficiency changing, especially for those not-for-profit institutions. In this paper, new DEA models are presented in which some limitations of standard DEA models have been overcome and decision makers' value judgment can be adopted more directly. The results for scale efficiency of research groups of one research institute are presented as a case study.