What the Numbers Mean

In SSCI Journal Citation Reports, the Institute for Scientific Information (ISI) summarizes the bibliographic citations in social science journals. These summaries include impact factors. An impact factor is the average number of citations received in one year by the articles that appeared during the two previous years:

Citations in year t to articles published in t-1 or t-2
Articles published in years t-1 or t-2

For example, the 2001 impact factor for Psychological Review is the average number of citations during the year 2001 to articles that the Review published during 1999 or 2000.

The lists posted here are not intended to be exhaustive. The lists focus on the 704 journals that are most central to business as indicated by citations. The listed journals had been cited at least eleven times by at least one of the business journals between 2000 and 2007. For this purpose, the initial “business journals” were the 176 journals that ISI classified as falling into business, business finance, industrial relations, or management.

The numbers are estimates based on the data from 1998-2007. Because the raw data include random errors, smoothed data are generally more accurate. Also, the most recent figures are a year old, and strong trends have likely continued. Thus, I have used exponential smoothing both to remove some of the random errors and to extrapolate the data up to 2009. That is, the numbers in the table of forecasts are the citations that each journal will have at during 2009 if the 1998-2007 trends continue.

Exponential smoothing gives more weight to more recent data. I varied the smoothing coefficients while comparing the actual outcomes for 1990 with estimates generated from the 1981-89 data. The smallest average errors occurred when estimates of the current values give 50% weight to the most recent observation, 25% weight to the next most recent, 12.5% weight to the third most recent, and so on. However, the estimates of trends give only 25% weight to the most recent observation, 19% weight to the next most recent, 14% weight to the third most recent, and so on.

Because the raw data have a lognormal distribution, I made all computations with logarithms.

What Do Citation Rates Mean?

Citation rates measure visibility. They do not measure quality or intellectual rigor as such. Journals low on the list may publish high quality articles or uphold rigorous standards, but people rarely cite the articles published in these journals. Quality is, of course, a highly subjective attribute. One of the journals I personally find most interesting ranks 125th.

Journals with larger circulations tend to receive more citations. However, if one controls for circulation, journals published by professional societies receive fewer citations than do independent journals. It seems likely that this is because journals published by professional societies are mailed to more people than actually read them.

American journals tend to get more citations than foreign ones, especially ones in foreign languages. These lower citation rates afflict even British journals that have high standing in the United Kingdom, such as the British Journal of Psychology, Organization Studies, or the Journal of Management Studies.

Citation rates incorporate systematic biases arising from the bibliographic practices in different fields. In 1999 and 2000, articles in business finance cited around 31 references on the average; articles in management cited around 49 references each; and articles in the general business journals cited around 54 references each. If articles cited only other works in their own category, articles in general business journals would receive over half again as many citations per article as those in business finance. Actual impacts do not, however, exhibit such large differences. Articles in general business journals averaged .83 citations, those in management .82 citations, and those in business finance .69.

The number of relevant journals has increased about 40% in three years! A major factor in this increase has been that business journals have discovered modern technology and the natural environment. Business journals are making many more citations to topics relating to computers, information, and environmental concerns. As well, ISI is beginning to take account of more journals published outside the US and more journals that are not in the English language.

These trends seem likely to continue. For three decades, publishers have been launching many new journals. Although each of these new journals attracts only a tiny fraction of the total readership and citations, the combined effect of so many new journals has been dramatic.

Updated 9/23/08