Impact factor, IF, is a metric that has generated a lot of controversies in the scientific community. Journals now use this as a medium for promotion. Recently, some journals that were well ranked with high IF in JCR dropped to the bottom of JCR table by losing their impact factors. This was due to fraudulent acts by the journals to gain IF through citation cartel (co-operating citations and self-citations). Though the journals in question might have shown excesses, the question is 'how many journals do not promote citation of their published articles?' Some people believe the journal office is at fault, while others blame it on the IF metric.
What is your opinion? Is IF a true measure of journals/articles/scientists' values? Is IF necessary?
http://scholarlykitchen.sspnet.org/2012/04/10/emergence-of-a-citation-cartel/
Interesting question. I think (but I may be wrong) that there is a consensus among true scientists to consider that IF is NOT a measure of the scientific value of a searcher. However, most of them would consider that IF is a measure of the "value" of a journal. Well, let' s see the IF list of ISI. At the top is "CA: a Cancer Journal for Clinicians", whose impact factor is FOUR TIMES that of "Nature". It is likely that there is nothing fraudulent there, and also that this not the journal where would be published scientific breakthroughs in the first place (I do precise that I am in the cancer field). This just indicates that IF is just related to the number of readers, and that considering it as a measure of "scientific" value is a matter of blind faith, not science.
Interesting question. I think (but I may be wrong) that there is a consensus among true scientists to consider that IF is NOT a measure of the scientific value of a searcher. However, most of them would consider that IF is a measure of the "value" of a journal. Well, let' s see the IF list of ISI. At the top is "CA: a Cancer Journal for Clinicians", whose impact factor is FOUR TIMES that of "Nature". It is likely that there is nothing fraudulent there, and also that this not the journal where would be published scientific breakthroughs in the first place (I do precise that I am in the cancer field). This just indicates that IF is just related to the number of readers, and that considering it as a measure of "scientific" value is a matter of blind faith, not science.
It's a pity that we cannot edit our own message anymore. Please read where, not were.
Hi Daniel, You should definitely be able to edit your own messages, at least I can edit mine, it's simply not so obvious as it once was. There's a "hidden" down arrow in the upper right hand corner of your messages. You need to move the mouse over your message and find the arrow. Then you click on it to choose whether to edit or delete. Good luck.
I largely agree with the comments made in the question as well as Daniel's answer. We can use IFs to compare which journals in our own field have the highest impact, which is also a fairly good indicator of how selective the journals are. I think most of us do this routinely when deciding where to send our next manuscript. If it is describing a major breakthrough of wide scientific interest, you want to go with an high impact journal. Long before IFs were calculated, scientists knew quite well the relative prestige in publishing in different journals - it's a part of the competitive research world we live in.
IFs should not be used, however, to evaluate individual researchers, as they may well have published great science in low impact journals. Here, citation frequency (and H-index) is a much more appropriate bibliometric parameter.
Also, a lot of strange things happen with the IFs of individual journals, as indicated both by Roland and Daniel, so one should be aware of this. An interesting read is http://blogs.scientificamerican.com/information-culture/2012/05/07/understanding-the-journal-impact-factor-part-one/ outlining e.g. how Acta Crystallographica suddenly jumped from IF 3 to 54 due to a single paper which has been cited over 26 thousand times. I think that tells us that we should be fairly familiar with the journals we publish in and their reputation, and not go blindly by IF. The IF of Acta Crystallographica is back to 2-3.
To reiterate the points above....many papers (often a majority) of papers in a given journal will have fewer citations than the IF "average". Many will have 0. The flip side of the Acta Crystallographica example above is all the papers that were not being cited while its IF was high - probably just as many as before. If citations are a measure of quality, at best the IF tells you something about the journal but next to nothing about and individual paper. And as we have seen it may not give a good overall view of the journal either, since it is based on means which can be skewed by extreme values. The IF is probably better understood as a measure of 'impact' in any case - this will (hopefully) correlate with quality but its not the same! But it is impact of the content of the journal as a whole, not individual papers. QED the IF of the journals you publish in may say little about the quality of the science you are publishing!
The extent to which it measure the journal's impact is likely to vary across fields - some field most scholarly activity is in journals and so this might be a fair approximation. Even then the measure of impact is limited to the scholarly community.
Hi Bjorn, thank you for the tip. Actually, I think that evaluating a paper or a journal by its number of citations is like evaluating a book or a publisher by its sales. It has no meaning, except if you don't read books. To go back to "CA: cancer journal for clinicians" its high impact factor is sustained over several years. This means that it is not related to a single article, but that the journal is read and cited by clinicians.
Hi Daniel. I must admit that I really don't understand your comment on book sale information saying "It has no meaning, except if you don't read books."? Bestseller book lists are useful in the way that a lot of people use them to pick out books to read. They carry certain information, i.e. a lot of people have bought this book before you. So either it's a very popular author or is a book which has received a lot of good reviews or both. Thus, it's not meaningless, but just as the IF it's just a number and one has to take it as such.
If you think the IF is meaningless, then that's your opinion of course, but the fact is that the IF is a widely used bibliometric parameter and that's why we're having these discussions on RG. Cheers!
Hi Bjorn, what I mean is that judging from book sales is not a way to evaluate literature. Otherwise, Nobel prizes in literature would go to authors of cookbooks, biographies of actors, politicians and sporstmen, and so on. It's the same for science. Recognition helps in boosting the sales, that's all, but to evaluate a book, you must read it, there is no other way.
I think we all agree that the IF is not a good estimator of journal 'impact', whatever that is (cf. the difficulties within the social sciences to give an unambiguous definition of the term 'power'). The shortcomings of the IF was pointed out in a letter by Sture Hanson to The Lancet already in 1995, when the 'art of impact manipulation' probably was less wide-spread than what is highlighted by the first link of this thread. (I take the liberty of violating the copyright laws by adding Hansson's article to this posting, as I think it's a comprehensive paper which merits a much wider spread than what have been the case so far. The paper also contains some references to early publications on this topic.)
Briefly, by analysing bibliometric data Hansson found a relationship between 'citation half-life' and IF. His conclusions were "IF is predicted by field-associated habits to cite articles published within 2 years, favouring research areas that generate many short-term studies. IF creates a tendency to treat clinical journals as less important. Quality plays a small part in determining IF." Challenging to the notion of even the 'proper' IF as a useful tool for assessing general scientific 'impact' of a journal, and devastating to the belief that evaluation and ranking of a researcher's scientific endeavor necessarily is made objective by adding bibliometrics estimates.
I suggest that scientists should compare the 2 years impact factor of their own papers to what they consider their "value". They would consider that there is no correlation at all. In my case, it is also the same for the number of total citations. For my own papers, the only correlation I see is between late citations (after 15 years) and "value".
About Journal Impact
Impact Factor - What is it?; Why use it?
The impact factor (IF) is a measure of the frequency with which the average article in a journal has been cited in a particular year. It is used to measure the importance or rank of a journal by calculating the times it's articles are cited.
How Impact Factor is Calculated?
The calculation is based on a two-year period and involves dividing the number of times articles were cited by the number of articles that are citable.
Calculation of 2010 IF of a journal:
A = the number of times articles published in 2008 and 2009 were cited by indexed journals during 2010.
B = the total number of "citable items" published in 2008 and 2009.
A/B = 2010 impact factor
- Badges
- Science topic
- Similar topics
- Education
- Higher Education
- Academic Writing