For all REDIB journals also indexed in WoS Core Collection, we will extract citation metrics over a rolling six-year period of time (each iteration analyzing citations from the latest full year to the previous five years). The data would be based on citations matched from indexed (or source) papers to indexed papers, and then those paper-to-paper citation links would be aggregated at the journal level. For all overlap journals, we would include citation counts and links with journals in the other segments of the WoS Core collection (Science Citation Index-Expanded, Social Sciences Citation Index, Arts --AMP-- Humanities Citation Index, Emerging Sources Citation Index) to create the most complete picture of journal influence and impact possible.
Measure of journal impact will include:
- Normalized Citation Impact Percentile(NCI Percentile)Relative citations per paper based on article by article analysis and aggregated to the journal level. Normalization will include consideration of category, document type and year of publication.
- Percent Cited Papers(Percent Cited Papers)The percentage of papers in a journal cited over the period surveyed (1 or more citations).
- Scaled Percent Category's Cites(Scaled Percent Category's Cites)Of all citations received by a given category in the timespan analyzed, the percentage for which a particular journal accounts. This will then be scaled so that the journal with the highest percentage has a score near 100.
- Top 10% Papers(Top 10% Papers)Percent of papers in a journal ranked in the top 10% within category. Percentile is normalized at the category, document type and year level.
- Mean Percentile(Mean Percentile)Mean percentile of papers in a journal, where percentile is again normalized at the category, document type and year level.
The measures chosen are primarily size and category independent, allowing for fair comparisons of journals that publish different numbers of papers, and in different categories. The exception is percent of total citations in a category, which is size dependent and an indicator of gross influence and stature or prominence in a field.
All of these scores will fall on a scale of 0 to 100. Therefore, for each journal, we will combine these metrics into an overall score by taking the average of across all scores. This average will be used to rank the journals. Not all journals will have data for the entire six-year window of the ranking. Those journals will still be included and have scores, but will be flagged to denote the fact that they did not exist long enough to amass data in all of the relevant years.
