Researchers downloaded the data from Wikipedia on January 30, 2010 that contained 19 million pages. They then extracted all PubMed journal article references from this data. They then took citation information scores from Faculty of 1000 (F1000) and PubMed Central (PMC) for PubMed journal articles published before 2010. To then also compare articles referenced in Wikipedia and those that are not, they calculated the probability of getting citation counts or scores that are similar by randomly drawing articles from F1000 and PMC. They took their data from Wikipedia and compared it to journal article citation counts and F1000 scores. They used PMC to find the journal article citation counts for about 19000 articles referenced in Wikipedia, and F1000 to find scores for 5000 articles referenced in Wikipedia.