0

For a while ResearchGate provided something called impact points. It was calculated as the sum of the impact factors of all the publications an academic had. For example, if an author had been an author on four papers with impact factors 0, 2, 4, and 3. They would have 0 + 2 + 4 + 3 = 9 impact points. They've recently stopped reporting this value because supposedly they no longer believe impact factors are valid indicators of an individual article (of course, as a side note, they continue to report their own mysterious ResearchGate points which in addition to publications also weights things largely irrelevant to the research community, but relevant to encouraging behaviour on the site that they desire).

If you also knew the number of publications someone had, you could also quickly determine their average impact factor per publication.

I found the combination of average impact factor and the sum of impact factors to be a really useful metric when getting a quick feel for an academic's publication track record. In addition, I think that impact points seemed to provide a reasonable starting point for discussing some of the trade-offs in publication strategies between quality/significance and quantity.

Given that this site includes many who are mindful of the problems with metrics for evaluating academic output, allow me to justify why I like the combination of average impact factor per paper and the sum of impact factors (i.e., impact points):

  • Person-specific citation based metrics (e.g., h-index, total citations, and so on), which are often cited as preferable, are heavily influenced by time. Citations accumulate over time. Thus, a young researcher a couple of years out of their PhD may have been publishing high quality work in top journals, but may have very few citations. In contrast, a researcher may have been publishing lots of publication at the mid-tier level for many years, and may have a lot of citations. This connection with time is more than the linear increase in publication output you might see given a research with consistent output each year. Instead, In a simplistic model, it is a multiplicative effect of average number of publications per year, time since first started publishing at that level, and average time between publications and now.

  • While journal impact factor is field dependent, you can start to adjust for this mentally, if you know your field. For example, I'm in psychology, and it tends to have lower impact factors than psychiatry due to various citation practices. As an aside, it would also be useful to use other journal impact metrics to form the average or sum that are less field dependent (e.g., the SNIP or SJR).

Probably the biggest issue with impact points is that some authors have publications with many more co-authors, or have more or less first-author papers, although this is less of an issue if you focus on average impact factor per paper.

So my question is, given that ResearchGate has stopped reporting impact points, is there an alternative provider where you can quickly obtain the sum or average impact factor of the publications of a given academic?

  • 5
    Incoming comments about how useless bibliographic metrics are in 3.. 2.. 1... – Cape Code Jul 20 '16 at 7:23
  • 4
    if an author had been an author on four papers with impact factors 0, 2, 4, and 3. — What does it mean for a paper to have an impact factor of 2? Impact factors are usually defined as average citation counts over all publications in a particular journal in a particular time window. By that standard definition, individual papers do not have impact factors. – JeffE Jul 21 '16 at 18:38
  • 1
    @jeffE For the purposes of calculation, papers get the impact factor of the journal they are published in. – Jeromy Anglim Jul 21 '16 at 23:37
  • 5
    I'm sorry, but that's just silly. If you insist on counting something, why not count citations to the individual paper, using Scopus or Google Scholar, for example, instead of an aggregate statistic to which the paper only makes a minuscule contribution? – JeffE Jul 21 '16 at 23:49
  • 1
    But if you want to get a bibliometric measure of quality then I think that in the short term, something like the impact factor of the journal (or some other indicator of quality) is more useful than citations. – Jeromy Anglim Jul 21 '16 at 23:55
3

I'm not aware of any specific tools like ResearchGate which provide an average impact factor for a researcher - maybe others know of something? However, you could fairly painlessly generate the average SJR/SNIP value for a researcher through SciVal in Scopus.

If you add a researcher in SciVal, then click on the benchmarking tab, and then click on View list of Scopus Sources for the selected Researchers and Groups this will bring up a list of journals the researcher has published in. You can then click on Export and download the list of journals, with their SJR/SNIP values into excel, where you can then just use the =AVERAGE formula to calculate the average. I just did this and it took me 3-4 mins max.

Back to your specific question, you could use InCites to follow a similar process as above to calculate a researchers average Journal Impact Factor (Thompson Reuters) however it's not as straightforward as SciVal. As far as I know, you can't generate an XLS list of journals a researcher has published in within InCites (so you would need to get this list somewhere else). You can however export a list of journals from InCites with their impact factors but you would then need to link your researchers list of journals with the incites list to collate the relevant impact factors. you could use excel VLOOKUP to do this and then calculate the average. This is a lot more effort though, so it would depend why you want this average measure of research impact in the first place and whether it might be enough to just use the scopus impact metrics.

  • Thanks. Yes. SciVal works. They also provide IPP which is similar to impact factor. – Jeromy Anglim Jul 22 '16 at 0:17
1

An alternative to looking at impact points that indexes a similar concept is citations in the last full year. This can readily be obtained for anyone with a Google Scholar profile. Author search on Scopus also shows it.

E.g., see the 2015 column below (e.g., go to this example, hold mouse over to see value): einstein scholar citation

Impact factor is essentially the mean number of citations per article per year in a given journal (where for example if the year for gathering citations was 2008, the publication years would be 2006 and 2007). But there is also a five year impact factor which extends the number of years articles are considered (e.g., 2003-2007). Depending on the life cycle of citations in a given field and journal, expanding the window of time for considering articles may alter the estimate of mean citations per year. My main point is that the underlying construct that is being measured is mean citations per article per year.

Thus, in a rough sense, the impact factor provides a rough guide to how many citations per year an author can expect from a given article published in a journal with a given impact factor. Of course, the distribution of citations in articles is positively skewed, but as we aggregate over many papers, the central limit theorem will begin to kick in.

Thus, citations per year in the previous year is in some senses addressing the same concept as impact points (i.e., the sum of the journal impact factors of the papers that an author has published in). In particular, it doesn't double load on years as an academic (i.e., total citations double loads for both the time you had to publish more articles and time that those publications have had to accrue citations). There are of course a few differences:

  • It is based on how much people cite the particular author rather than the journals that they publish in. In many respects this is a positive, because it is more aligned with the authors achievement.
  • It is typically based on a smaller sample size than journal impact factors. So there is greater scope for outliers to skew the distribution. I.e., one or two papers with hundred or thousands of citations may distort the underlying pattern.
  • It uses the full history of articles by the academic. Thus, depending on the citation patterns of the field and how old the authors articles are, this may introduce particular distortions. For example, older academics may have articles that have stopped being cited particularly in fast moving fields. In fields with citation half lives that are often over 10 years (like mathematics, psychology, and the social sciences), this should not be a problem for most active academics.

As an aside, all citation based metrics including impact factors are contingent on the database or articles used to specify articles and used to source citations. Google Scholar is quite inclusive. ISI Impact Factor has greater quality control and is less inclusive. So, from casual observation, It seems like Google Scholar picks up a multiple of between 2 and 5 times more citations than ISI or Scopus. Thus, any comparisons need to be mindful of that.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.