Last week, I talked about H-Index which is a metric for measuring the impact an author has in his field.  This week, I will discuss Impact Factor (IF) which is one way to measure the impact a journal has in a field.

Impact Factors for a journal are available through Journal Citation Reports which is produced by the same company that produces Web of Knowledge.  The simple impact factor for a journal in any given year is the average number of citations in that year of articles from the previous two years.  So if a journal has an impact factor of 2 for 2011, it means that on average all the articles published in 2010 and 2009 in that journal have been cited two times.

Just like an H-Index has more meaning in context, an Impact Factor should also be viewed in context.  Some disciplines have longer times to publication so a 5-Year Impact Factor may yield more about the journal than the standard 2-year Impact Factor.  Also, journals should be compared to other journals in the same field rather than to other fields or all journals.  One can do this using an Aggregate Impact Factor (AIF) which is calculated from all the journals in a field or the Median Impact Factor (MIF) which gives the median value from all the journals in a field.

The trick is figuring out what to use for the field.  Biology (AIF=3.146, MIF=1.339) is listed separately from different different sub-disciplines of biology and medicine (agronomy, biophysics, microbiology, etc).  Likewise there is a general listing for environmental science (AIF=2.496, MIF=1.560), geology (AIF=1.773, MIF=1.049), mathematics (AIF=0.735 , MIF=0.584), neuroscience (AIF=3.914, MIF=2.783), psychology (AIF=2.579, MIF=2.075), and statistics (AIF=1.140 , MIF=0.948), but none for chemistry, computer science, or physics.

It is important to keep in mind that Impact Factor should not be the only way a journal is evaluated.  Especially as there are ways for journals to manipulate to a certain extent their impact factors. In addition because Impact Factor is an average, one paper with a large number citations can skew the average.  If a journal has 10 papers and 9 of them are cited once and one of them is cited eleven times, the Impact Factor is 2 even though the median and modal citation number is 1.

Librarians do consider Impact Factor when subscribing to journals.  An expensive high impact journal or an inexpensive low impact journal but not an expensive low impact journal may be worth subscribing to, but local usage because of research interests of faculty and students may be a more important factor.

Authors often consider Impact Factor when deciding what journal to submit their article to, but the research focus of a journal is also important.  Sure every geologist might want to publish a paper in Nature or Science especially since they publish very few geology articles it’s even more prestigious.  But it might be more effective to publish in Geology since its primary readership are geologists and the primary readership of Nature and Science are non-geologists who will probably just skip or skim the article.