| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.45 MB | Adobe PDF |
Orientador(es)
Resumo(s)
Composite indicators play a pivotal role across various fields, serving as powerful tools to condense information into a single, comprehensible metric. They function as a link between complex data and practical conclusions, which makes them valuable assets in diverse fields, assisting decision-makers. Since objectivity is vital for policy- and decision making, many researchers base their studies on the Benefit of-the-Doubt (BoD) methodology, which originated from the widely recognised Data Envelopment Analysis. This study provides a detailed comparison between the linear and multiplicative BoD approaches, incorporating both optimistic and pessimistic viewpoints to assess the performance of the Portuguese public hospitals. The linear approach is an additive linear programming model that constructs a piecewise linear and convex efficient frontier with the benchmarks. In contrast, the multiplicative approach constructs logarithmic curves instead of linear segments to define the efficient frontier allowing for greater flexibility by accommodating nonconvex and nonlinear shapes that better reflect the data distribution. As a result, the multiplicative approach achieves a tighter fit to the data, ultimately yielding higher overall scores. The results show that multiplicative scores in the pessimistic approach are lower than linear scores. Conversely, in the optimistic multiplicative approach, while expected to yield superior scores, certain entities lag due to non-compensatory elements. The implementation of the multiplicative BoD is remarkably simple, requiring only minimal changes when compared to the linear BoD. This raises questions about its low adoption and utilization compared to linear methods, despite its apparent advantages.
Descrição
Palavras-chave
Healthcare performance COVID-19 pandemic Composite indicators Multiplicative benefit-of-the-doubt Data envelopment analysis
Contexto Educativo
Citação
Editora
Springer