Thomson/ISI has issued their own statement refuting the JCB’s article claiming that Thomson could not provide the data used for determining a journal’s impact factor. According to Thomson, the data supplied to JCB matched the Impact Factor ratings to within 99.8% (note that the JCB claims this set of data appeared to be, “assembled in an ad hoc manner to create a facsimile of the published data that might appease us”). Thomson states that there is only one set of data, not two as claimed, and that this set of data is interpreted in different ways for different purposes.

—article continues—

It will be interesting to see if the JCB responds, or if any other publisher chime in (given the presence of the holiday season, one wonders if this will just blow over as everyone is on vacation). Note that CSH Protocols is not currently listed for an Impact Factor with Thomson/ISI, so I can’t comment on any personal interactions with the company or their services. In our experience, many listing services such as these aren’t quite sure what to do with methods journals–the articles published aren’t new research, yet they are of high interest to working scientists.

One consistency in both articles (and the previously cited PLOS Medicine article) is that the Impact Factor does seem to be, in some ways, more of an art than a science. As Thomson themselves note, the decision as to which articles to include is based upon “a detailed, journal-by-journal review of the presentation and labeling of articles in a journal, expanded by information provided by publishers regarding the content and structure of the journal, as well as key bibliometric characteristics.” So as PLOS Medicine notes, “rules are unclear…the company refuses to make public its process for choosing citeable article types.” While Thomson notes that these methods, “have proven effective across many years,” it is difficult to gauge their accuracy.