The REDIB Ranking analyses the citations of indexed articles to indexed articles, and then adds these article-to-article citations for each journal.
The REDIB Ranking goes much further than a simple calculation of the citations made in the given period and dividing this figure by the number of articles. Rather, it contextualises the citations in each subject area, taking into account certain factors (number of publications, expected average impact, etc.) The REDIB Ranking differs from other rankings of similar document bodies in the sense that it is the only one that calculates expected citations or which compares its journals with the rest of the world, outside its own corpus.
The ranking does not analyse the publications as a whole. Rather, the data provided about each of the journals are obtained by aggregating the data gathered through analysing each of the articles and then contextualising them, as mentioned above.
With this in mind, the following indicators are developed for the overlapping REDIB/WoS publications, the combination of which results in the position of each journal in the annual REDIB Ranking.
- NORMALISED CITATION IMPACT PERCENTILE
For each scientific subject or discipline, there is a unique communication dynamic between scientists. The abundance of citations that feature in the cited bibliography, the persistence of citations over time and the time it takes for a piece to have an impact on subsequent related works are examples of factors which differ greatly from one subject to another. It therefore makes a lot of sense to treat each subject separately in the impact calculations, as it isn't a good idea to apply the same yardstick to all of them. Following this criteria, it is possible to calculate the average citations that a document may have (Expected Citations per Article) for each subject, publication year and document type. Each document published in REDIB journals that overlap with the WoS Core Collection during the years covered by the ranking (2013-2018 for the 2019 edition, which was published in 2020) is extracted. Firstly, the performance of each document is analysed in comparison with the expected citations. The expected citations correspond to the average number of citations received for each document published in the WoS Core Collection with the same characteristics, taking into account the year of publication, the topic and the document type. The number of citations received by a document is then divided by its expected citations. This generates a document's Normalised Citation Impact. This is added on a journal-by-journal level, as per Crown's rules. The Normalised Citation Impact (NCI) is calculated in the following way: For each article published by an overlapping REDIB-WoS journal on a given subject during the six-year period covered by the ranking, a calculation is made of the number of citations received by the set of journals of its subject area in the WoS Core Collection corpus published in the year of the ranking. For each topic, the number of cases received is the numerator and the expected citations per article is the denominator. This division produces the NCI for each article. By adding them together, you get the NCI for the publication being analysed. Once the NCI has been obtained for each journal, percentiles are allocated. A value of 100 is given to the highest NCI. Using this as a reference, the percentile of each journal can be calculated. This 100 is then adjusted to a value very close to 100. This uses the same formula as the one used for JIF percentiles.
- PERCENT CITED PAPERS
Not all articles in a publication have the same impact. In fact, some of them are never cited. This indicator allows us to distinguish those publications for which citations are evenly distributed across its articles from those that improve their results by publishing high-impact articles that cover up for other articles that have had zero impact. More balanced and consistent journals are given greater weight in the rankings.Rather than the number of citations, this percentage is calculated by counting the articles that were cited in the ranking year from all the articles published by the journal being analysed in the six-year period, compared to the total number of articles published by the journal in the period analysed. The data is presented as a % to three decimal points.
- SCALED PERCENT CATEGORY'S CITES
In contrast to disciplines that distribute their citations across a solid number of publications, there are others that, for various reasons, concentrate their impactful publications in very few journals. This indicator provides information on the concentration of citations in a given journal within a certain subject area and, of course, during the period of time being analysed. To calculate this indicator, the number of citations received by each overlapping journal during the period being analysed is counted, with this figure then being divided by all citations received in the same period by all WoS Core journals within the same subject area. The indicator shows the weight of citations of each WoS Core journal. All results are allocated a percentage, with the top result being allocated 100.
- PERCENTAGE OF MOST CITED ARTICLES (TOP 10%)
Indicates the percentage of a journal's articles that are in the top 10% most cited articles within a certain subject area (or areas).For each WoS Core subject, the articles in the top 10% are identified. Among these, those that feature in REDIB journals are identified. For each REDIB journal, the total number of 'top 10' articles is divided by the total number of documents in the journal. The result is given as a % to three decimal places.
- MEAN PERCENTILE
This is the journal's percentile with regards to its subject area, found by aggregating the percentile of its articles. Each article published during the six-year period receives, in the year of the ranking, a number of citations. This number allows us to order all the articles of a certain subject area, and this order is applicable to each category. The non-cited documents will be given a 0 by default (the lowest score); the most cited article in its category, meanwhile, will be given 100. The average percentile for each journal is obtained by calculating the average percentile obtained for each article published in the same journal. Journals are ordered by the their average percentile, which is expressed as a number between 0 and 100.
- OVERALL SCORE
The five variables given acquire a value between 0 and 100. This allows an average of the five variables, each weighted equally, to be calculated for each journal. This is the Overall Score, which is specific to each journal and year of ranking. It is the best parameter for evaluating the journal's evolution over time, and is the value that determines the journal's position when filtered by each of the fields that are not partial indicators (subjects or countries) and the overall classification of the REDIB Ranking.
To conclude, it should be noted that the way in which the results are presented allows the scholar to order the results by each of the five indicators independently, giving him/her an accurate view of each journal and its position according to the different criteria.
More information
