The Web of Science database already provides the tools for ranking history journals. If journals in history of science and technology were overrepresented in this database during the early 1990s, that is no longer the case. There are now 233 history journals in the database (see here, here, here, and here), and only about a fifth of the titles are categorized as history and philosophy of science. The database provides two pieces of information which are critical to figuring out the ranking of history journals: 1) the number of articles that have been cited at least 20 times, and 2) the average number of times each article has been cited.
Here is a quick preview of what this method yields:
| History Journal Title | Average number of times each article has been cited | Number of articles that have been cited at least 20 times |
| AMERICAN HISTORICAL REVIEW | 10.32 | 42 |
| JOURNAL OF AMERICAN HISTORY | 7.55 | 35 |
| COMPARATIVE STUDIES IN SOCIETY AND HISTORY | 7.26 | 35 |
| JOURNAL OF AFRICAN HISTORY | 5.76 | 26 |
| JOURNAL OF INTERDISCIPLINARY HISTORY | 6.07 | 24 |
| JOURNAL OF SOCIAL HISTORY | 4.38 | 25 |
| JOURNAL OF MODERN HISTORY | 4.18 | 22 |
| ARCHIVE FOR HISTORY OF EXACT SCIENCES | 3.43 | 16 |
| ANNALS OF SCIENCE | 3.08 | 15 |
| SOCIAL SCIENCE HISTORY | 3.2 | 14 |
We can draw a few conclusions from these numbers and methods. The Web of Science database contains about half of the nearly 400 English-language history journals in the AHA's database. So this method is fairly good at capturing and analyzing the range of peer-reviewed journals in history. Significantly, it does not compare the "impact factor" of history journals with those of science journals. This method also solves the outlier problem. (What I mean is that the above journals publish as much as one or two articles a year that will eventually be cited 20 or more times in subsequent articles.) In other words, part of this method will tell you whether a history journal occasionally includes highly- or moderately-cited articles that skew averages, and the other part factors this out by calculating the average citation rate for each article that has appeared in the journal. This means we can determine which journals consistently publish highly-cited articles, which journals sometimes publish moderately-cited articles, and which journals rarely publish articles that are cited in other journals. And don't forget that Web of Science captures citations of articles in both other history journals and non-history journals. So, in effect, Web of Science is simply telling us how historians and other academics are voting on the articles that appear in history journals. As I see it, this is a pretty good measure of journal quality.
There is something more I must say. The ranking of scholarship in history will never be complete until we include both articles and monographs. Now that the API for Google Books is being updated, we are on the verge of creating text mining algorithms that can scan the endnotes and bibliographies of books to figure out who is citing who. This information could then be merged with citation data from Web of Science to create a more comprehensive picture of citation patterns in history scholarship.
Posts
5 comments:
I completely agree - the potential for using Google's API for such studies is absolutely fascinating. It would not only provide the kind of "rankings" of prestige and respectability referred to here, but could also reveal differences in areas of study. Do social historians have a wider scope of citations than political historians? Which fields are the most narrow in focus? The possibilities are endless...
Would you mind detailing your methodology for pulling up the citation rates in Web of Science? Just curious. We only have access to the WOS file back to 1983. Did you use the full file (back to '56)? Did you include book reviews in your averages? Thanks!
Jennifer Duncan's questions are good ones. It's also not clear why "citation" should be regarded as synonymous with "cited in another article." Are not citations in published books and essays in edited collections no less legitimate indications of scholarly impact? (Though, if so, are we to distinguish further between titles cited the notes and those appearing only in the bibliography?)
Really helpful data, lots of thanks for the post.
Great information it really helps so much and this topic is so interesting
Post a Comment