Identifying institutional relationships in a geographically distributed public health system using interlinking and co-authorship methods.

Autor/inn/en	Ontalba-Ruipérez, José-Antonio; Orduna-Malea, Enrique; Alonso-Arroyo, Adolfo
Titel	Identifying institutional relationships in a geographically distributed public health system using interlinking and co-authorship methods.
Quelle	In: Scientometrics, (2016) 3, S.1167-1191 PDF als Volltext Verfügbarkeit
Dokumenttyp	gedruckt; online; Zeitschriftenaufsatz
ISSN	0138-9130
DOI	10.1007/s11192-016-1839-z
Schlagwörter	Link analysis; Network analysis; Co-authorship; Public health; Institutional relationships; Spain
Abstract	Abstract Link analysis is highly effective in detecting relationships between different institutions, relationships that are stronger the greater their geographical proximity. We therefore decided to apply an interlinking analysis to a set of geographically dispersed research entities and to compare the results with the co-authorship patterns between these institutions in order to determine how, and if, these two techniques might reveal complementary insights. We set out to study the specific sector of public health in Spain, a country with a high degree of regional autonomy. We recorded all Spanish health entities (and their corresponding URLs) that belong to, and were hyperlinked from, the national government or any of the regional governments, gathering a total of 263 URLs. After considering their suitability for web metric analysis, interlinking scores between all valid URLs were obtained. In addition, the number of co-authored articles by each pair of institutions and the total scientific output per institution were retrieved from Scopus. Both interlinking and co-authorship methods detect the existence of strength subnets of geographically distributed nodes (especially the Catalan entities) as well as their high connectivity with the main national network nodes (subnet of nodes distributed according to dependence on national government, in this case Spain). However, the resulting interlinking pattern shows a low but significant correlation (r = 0.5) with scientific co-authorship patterns. The existence of institutions that are strongly interlinked but with limited scientific collaboration (and vice versa) reveals that links within this network are not accurately reflecting existing scientific collaborations, due to inconsistent web content development.
Erfasst von	OLC
Update	2023/2/05