Posted on June 15th, 2019 by Rafael Repiso
Author Rafael Repiso – Translation: Erika-Lucia Gonzalez-Carrion
- That the population studied is homogeneous. For example, that all elements of the ranking have had similar conditions, same time window to make merit, equality to make them, etc.
- That the elements of organization represent in a reliable way the concept by which the ranking is authorized.
- That there is a sufficient mass of data that allows ordering the elements of the ranking.
- That the element that you want to study has sufficient degree to distinguish between elements and do not accumulate elements with identical values and therefore positions.
The reality is that these conditions are given in few listings and yet we accept the rankings with their defects. When a complex and multidimensional reality is reduced to a single dimension it results in loss of information, despite everything, we like rankings so much! That is why it is important to analyze these products critically.
The present proposal compares two very different products that serve the same population, the Spanish community of researchers in Communication seen from In-Cites (WoS-International) and Dialnet Métricas (Dialnet Spain). While In-Cites WOS (SCIE, SSCI and A & HCI) represents the works published in the elite of international journals, Dialnet Métricas is a product that aspires to represent exhaustively the Spanish scientific production in Social Sciences in the national scope. That is, two different but complementary dimensions of the same reality. If we have given much importance to the international dimension, we should not forget that most of the research communities publish mainly in national journals, so this information is substantial to evaluate our system.
A search in In-Cites limited to authors with Spanish affiliation and in the category “Communication” returns 4456 different profiles, many of them duplicated and tripled results, of which the Top 10 is copied by number of citations. The problem in the rankings is usually in the queues, the headers usually result in the same actors. In-Cites generates a report for any researcher who works in a Spanish center and who has published in the category “Communication”. Therefore, there is no precise identification of the communication research community. It is striking that the two researchers with more citations are foreigners who work in Spanish centers, as well as the presence in the Top of appointments of professors from other areas, a bibliographer (Daniel Torres-Salinas) and a physicist (Juan Miguel Campanario). Another curiosity of the top ten is that a highly cited work positions its author, Ann Majchrzak as the tenth most cited.
The case of Dialnet Métricas is different, first of all, the Communication area is divided into two categories: “Journalism and Audiovisual Communication” that bring together 1266 researchers together. The assignment of areas has to do with the data of Dialnet profiles, so the selection is very precise. Only one researcher seems to be at the top of both rankings, Juan José Igartua. In the top only appear two women Carmen Marta-Lazo and Carmen Costa-Sánchez.
Link Journalism (https://dialnet.unirioja.es/metricas/ambito/13/area/1960) Link Audiovisual Communication (https://dialnet.unirioja.es/metricas/ambito/13/area/1961).
Advantages of In-Cites.
- It allows to restrict the reports to areas, combine them with other categories, show for periods of time, even perform searches by set of authors (names and individual identifiers).
- The results can be downloaded in tabulated formats.
- They reflect a very homogeneous scientific production, that of the elite of international journals.
- Provides multiple indicators.
Problems with In-Cites to order investigators:
- The names are not normalized, so there is more than one version of them, in this Top 10 the different profiles have to be unified.
- Some affiliations do not correspond to reality.
- Researchers from other areas can be introduced, as in the case of Daniel Torres and J.M. Belfry, who have published Documentation articles in “The Information Professional” magazine, which is in two different categories “Library & Information Science” and “Communication”. Therefore, the presence in more than one thematic category of journals used as a source can significantly alter the results.
- In-Cites’ own search and report creation system excludes articles published in other areas outside of “Communication” even though they are Communication or Communication professors jobs. In-Cites could take advantage of the profiles of Publons that are managed by the authors to assign more exhaustively the areas, but for that it is necessary that these profiles have a massive use. For example, Professor Igartua registers more than 20 articles published in journals indexed in JCR, but the report only shows those published in the “Communication” category.
Advantages of Dialnet Metrics
- They not only record articles, but also record other types of documents such as books, book chapters and theses.
- The transparency of the system and its connectivity with other types of data allow in-depth analysis of the communities.
- Very exhaustive control of the researchers, there are no duplications, at least they have not been detected in Communication.
- Visualize a reality where Spanish researchers participate massively, production at the national level, something that despite being the majority phenomenon in the disciplines of Social Sciences is ignored.
Problems with Dialnet Metrics:
- They omit the impact production published outside Spain that is perhaps the most significant for a researcher.
- Some of the area assignments are not correct, the assignment between Journalism and Audiovisual Communication / Advertising is not entirely clear, but this is an inherited problem …
- Some of the university assignments are not updated.
- Confuse conference proceedings with book chapters.
- Poverty in the indicators. Although it provides very interesting data such as the “age” of the researchers, it would be very important for the Ranking to include indicators that would allow the gross indicators to be relativized as the impact normalized by categories.
In short, these two products allow us to better approximate a set of data that is difficult to understand and also, in a combined way, we can obtain a very complete view of the communities of researchers in Social Sciences, however they use two very different approaches and in general They offer different data. We look forward to the growth of the Dialnet platform in number of areas as well as the correction of errors and the implementation of improvements. The interesting thing is that both Clarivate and the Dialnet Foundation have shown that continuous improvement is a daily goal, so we are very confident that we will soon be able to use both products with guarantees to better study the Spanish scientific communities.