Benchmarking scientific performance by decomposing leadership of Cuban and Latin American institutions in Public Health

• Published in: Scientometrics Vol. 106; no. 3; pp. 1239 - 1264
• Main Authors: Chinchilla-Rodríguez, Zaida, Zacca-González, Grisel, Vargas-Quesada, Benjamín, de Moya-Anegón, Félix
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.03.2016
• Subjects: Computer Science; Information Storage and Retrieval; Library Science; Normalized citation; Cuba; Scientific collaboration; Leadership; Public Health; Uncertainty; D8 Information; 94 Information and communication; Circuit; Knowledge; Latin America
• ISSN: 0138-9130; 1588-2861
• DOI: 10.1007/s11192-015-1831-z

Comparative benchmarking with bibliometric indicators can be an aid in decision-making with regard to research management. This study aims to characterize scientific performance in a domain (Public Health) by the institutions of a country (Cuba), taking as reference world output and regional output (other Latin American centers) during the period 2003–2012. A new approach is used here to assess to what extent the leadership of a specific institution can change its citation impact. Cuba was found to have a high level of specialization and scientific leadership that does not match the low international visibility of Cuban institutions. This leading output appears mainly in non-collaborative papers, in national journals; publication in English is very scarce and the rate of international collaboration is very low. The Instituto de Medicina Tropical Pedro Kouri stands out, alone, as a national reference. Meanwhile, at the regional level, Latin American institutions deserving mention for their high autonomy in normalized citation would include Universidad de Buenos Aires (ARG), Universidade Federal de Pelotas (BRA), Consejo Nacional de Investigaciones Científicas y Técnicas (ARG), Instituto Oswaldo Cruz (BRA) and the Centro de Pesquisas Rene Rachou (BRA). We identify a crucial aspect that can give rise to misinterpretations of data: a high share of leadership cannot be considered positive for institutions when it is mainly associated with a high proportion of non-collaborative papers and a very low level of performance. Because leadership might be questionable in some cases, we propose future studies to ensure a better interpretation of findings.