Sobre estatística de dados bibliométricos em grupos de pesquisadores : universalidade e avaliação / About statistics on bibliometric data of researchers¿ groups: universality and evaluation

AUTOR(ES)
FONTE

IBICT - Instituto Brasileiro de Informação em Ciência e Tecnologia

DATA DE PUBLICAÇÃO

2012

RESUMO

Research financing agencies, research centers, universities and the scientific community are frequently seeking for improvement and enhancement on the quality of researchers¿ work. Therefore, it is necessary to provide optimized tools and methods to get consistent evaluations. Several methods have been proposed over the years and different forms of evaluation are used by agencies such as Coordination for the Improvement of Higher Level Personnel (Capes), although some issues have been overlooked. With the invention of the h-index (Hirsch), that binds quantity with quality by measuring a set of papers according to their citation distribution, many studies have been considered with several different statistical methods. In the current literature, it is not found an analytical expression for h-index distribution over a researchers¿ group, nor the proof of universality in this kind of distribution for different groups and databases. This master thesis discusses primarily a study about h-index distribution and citations in three distinct scientific fields: physics, biology and computer science, which has different publication and research practices. Also, it is proposed a novel metric for ranking based on successive h-index, named as s-index, which makes possible to compare researchers¿ groups with different sizes, providing for example, a comparison in terms of productivity on higher and smaller groups of researchers, reflecting their skills on scientific production. A correlation study is conducted in order to compare the s-index, developed in this thesis, with the classification of post-graduation courses performed by Capes, using data from postgraduation researchers in physics and biology. The approaches presented can be used to classify researchers¿ groups through a quantitative view, by eliminating some qualitative evaluations that are hardly generalizable and replicable.

ASSUNTO(S)

researchers¿ analysis recuperacao : informacao h-index ontologias bibliometrics

ACESSO AO ARTIGO

Documentos Relacionados