الفهرس | Only 14 pages are availabe for public view |
Abstract This thesis builds on the extensive literature of the benefit of the doubt (BoD) methodology to set weights for composite indicators (CIs). The proposed methodology, meta - goal programming benefit of the doubt (MGP - BoD), proved to overcome some of the BoD shortcomings and enhance its performance. MGP- BoD belongs to the family of common - weights BoD models. It comprises of two sets of goals and two meta - goals. Among other merits, results prove two additional benefits of the MGP - BoD over older BoD. First, it enhances BoD discriminating power by eliminating all ties in CI values and, hence, country ranks. This high discriminating power is achieved in only one totally endogenous step. Second, MGP - BoD weights add up to one. This makes weights more insightful, interpretable, comparable to weights from other weighting systems, and easier to interpret compared to BoD. These additional merits favor MGP - BoD over previous weighting methods. In the meantime, the proposed method preserves a significantly high correlation with previous methods results as shown a set of pearson and spearman correlation tests to compare it to various previous methodologies |