Propuesta de un modelo de análisis redinformétrico multinivel para el estudio sistémico de las universidades españolas (2010)
Enrique Orduña Malea
Resumen (Inglés):

The University, as ancient institution, has nowadays an unquestionable influence in Society, both active (in its formative aspect of future professionals and citizens; generation of new knowledge and research; and transfer, development and innovation), and passive (due to its huge financing needs).

This influence marks the need to establish mechanisms and procedures to analyze their performance, efficiency, and effectiveness as an institution, as well as tools to properly display this performance, all within the context of a Society marked by the management and massive transfer of information through communication networks.

Therefore three complementary research lines are glimpsed: the analysis of the institution performance (explicated by the digital trail it generates, and the corresponding existing tools to quantify), and the visualization of this performance, where the ranking technique is widely extended, given the impact that this format generate in users. In addition, the systemic nature of the university determines, complicate and narrow each of these areas.

This thesis therefore aims to explore the capabilities that Cibermetrics provides to analyze, from a systemic point of view, the universities of the Spanish system, in order to obtain new knowledge about the performance of these to allow the future construction of systemic university rankings.

To do this, first provides a state of the art focused in the main three research lines (university, rankings, and Cibermetrics), with the aim of providing a comprehensive and critical framework.

After the introduction chapters, a netinformetric multilevel analysis model of universities is proposed in order to facilitate the acquisition of structured information, and which allow their subsequent use in the design and development of web university rankings. This analysis model is based on the definition of three levels (institutional, external, and satellite), and two sub-levels (contour and internal).

Once the model is proposed, the work argues to validate it; first through a descriptive analysis of the Spanish university system in the Net, during 2010, and later with a performance analysis, through the implementation of various netinformetric indicators at each level of the model proposed in order to identify patterns of distribution and dispersion of each indicator in each university at all study levels, and their evolution over time.

The descriptive analysis starts with a systematic study of the Spanish universities to obtain, depending on their missions, the different internal units of the university. After identifying the type of entities (institutions and products), we proceed to identify, retrieve, and analyze, in each of the 76 universities in the Spanish system, all available items, and the corresponding URLs.

The performance analysis starts with the selection of a set of information sources (search engines, web tools, etc.), and indicators (mainly size, linkage, and audience, among others) to then apply them to each URL (university or university unit), in four quarterly samples during 2010.

The results show that the proposed analytical model, based on levels (institutional, external, and satellite), and sub-levels (contour, and internal) is simple, independent of technique, and provides structured information that allows a complete analysis of each institution. Furthermore, the separation of the conceptual model from its application allows their continuous updating without changing its overall structure.

Similarly, it is concluded that netinformetrics (as a technique through which the model is applied in this study) provides information of interest for the description of universities as an information science method, and its use allows systemic analysis, which are shown to be necessary for a complete description of the universities, as objects of study themselves. It also provides information not obtainable by other methods and even in such orders of magnitude that it cannot be ignored.

The interpretation of these web indicators clearly shows that academic Spanish web space is exceeding complex due to the excessive practices of multi-domain (of 7,467 universities’ entities identified, 13,941 valid URLs are located), and web redirection (found in 19.14% of all universities).

However, despite the advantages of the model, it has a number of limitations: it is too limited by the instrumental netinformetrics, which causes their results should be interpreted with great caution; the institutional level, internal sublevel is only applicable to the Spanish case, where the academic functions and structure are clearly legally spelled, and not have to coincide with that of other university systems; finally, the model does not include all diversity categories and attributes.