Comparing source coverage, citation counts and speed of indexing in Google Scholar and Scopus, Henk F. Moed, Judit Bar-Ilan, Gali Halevi (Submitted on 17 Dec 2015) arXiv:1512.05741 [cs.DL]

“An analysis of 1,200 target articles in 12 journals in 6 subject fields and of 7,000 citations to 36 top cited articles found in virology and chemistry a ratio of Google Scholar (GS) over Scopus citation counts between 0.8 and 1.5, in Chinese studies between 1.8 and 2.8, in computational linguistics between 2 and 4, and in political science journals between 3 and 4. Open access journals show higher ratios than their non-OA counterparts. Unique GS sources come from Google Books and/or from large book publishers, and from large disciplinary and institutional repositories. Unique Scopus sources are mainly books and Chinese journals. There is a huge dispersion in GS source titles and web links. The citation impact of documents in surplus sources covered in GS but not in Scopus and vice versa is some 80 per cent lower than that of documents in sources indexed in both. Pearson R between GS and Scopus citation counts at the article level are in all 12 target journals above 0.8, and for 8 journals above 0.9. Effect of double citation counts due to multiple citations with identical or substantially similar meta data occurs in less than 2 per cent of cases. In GS, the trade-off between data quality and indexing speed seems to be in favor of the latter. A median Scopus indexing delay of two months compared to GS is largely though not exclusively due to missing cited references in articles in press. Pros and cons of article-based and concept-based citation indexes are discussed.”