Monday, 18 April 2011

Psychologists like to cite themselves

In a striking case of the experts falling foul of a phenomenon studied by themselves and their colleagues - the self-serving bias - it turns out that psychologists have a tendency to over-cite their own research papers.

Marc Brysbaert and Sinead Smyth analysed one recent issue of Psychological Science and the Journal of Experimental Psychology: Learning, Memory, and Cognition and two recent issues of the Quarterly Journal of Experimental Psychology and the European Journal of Cognitive Psychology.

For each of the articles in these journals, Brysbaert and Smyth used the 'find related records' function on the ISI Web of Science to find the article out there in the wider literature with the greatest overlap in the references it cited, but which was written by a different set of authors. This way the researchers ended up with a list of original target articles, each one paired with a second comparison paper by a different research group, presumably on the same or a highly similar topic (hence the overlap in the reference lists).

To check for a self-citation bias, Brysbaert and Smyth simply looked to see how many times the authors of a target article cited themselves compared with how many times they cited the authors of the comparison paper (and vice versa). For target articles, the average number of self-citations was 4.1 (11 per cent) compared with 2.3 citations of the comparison paper's authors. For the comparison papers, the average number of self-citations was 9 (10 per cent), compared with 1.8 citations of the authors of the target article.

The researchers summed up: 'A typical psychology article contains 3 to 9 self-citations, depending on the length of the reference list ... In contrast, cited colleagues in general receive 1 to 3 citations. This is what we call the self-citation bias: the preference researchers have to refer to their own work when they [supposedly] guide readers to the relevant literature.' The finding adds to past research that's shown academics are biased towards citing other researchers from their own country, and towards citing the work of the editor of the journal their research is published in.

Brysbaert and Smyth believe that psychology researchers indulge in biased self-citation practices not because their own past papers are always necessarily useful to the reader, but because it's 'good for the researchers' esteem, by means of self-enhancement and self-promotion.'

If that's the case, does it work? The evidence for this is mixed. A 2006 study in the field of economics found that papers with more self-citations were no more likely to end up being cited by other research groups. However, another study published in 2007 (pdf), which involved the analysis of over 64,000 Norwegian journal articles, found that authors who self-cited more also tended to receive more citations from others. 'So, although self-citations may not increase the likelihood that a particular article is cited, they do increase the chances that a particular author is cited,' Brysbaert and Smyth explained.

So, what to do about this self-citation bias? One option proposed by Brysbaert and Smyth is for journals editors to impose a cap on self-citations, particularly for journals, like Psychological Science, that have a cap on the total number of references allowed per paper - articles in this journal tended to have the highest proportion of self-citations. What do you think?

ResearchBlogging.orgMarc Brysbaert, and Sinead Smyth (2011). Self-enhancement in scientific research: The self-citation bias, Psychologica Belgica. In Press. [pdf via author website].

Post written by Christian Jarrett (@psych_writer) for the BPS Research Digest.


Anonymous said...

Well, duh! Two very simple reasons. 1) Self-citing increases the number of citations your papers have. This is stupid, but in some realms, it matters.
2) Studies don't happen in isolation - they're designed to test the next step in a particular theory, most commonly the one that the author has been working on for a while now. In which case, it's often justified to mention the previous steps in the story (as summarised in the author's past work).

Anonymous said...

I particularly agree with point 2) that the anonymous commenter above raised - that studies often follow on from earlier findings. unless you hop from field tofield of course you are going to cite your own theories and papers! As long as you are not only citing yourself (therefore being stuck in a box where your knowledge of what else is going on in the field is stuck) then it shouldn't be seen as a prblem. The quality of individual pieces of work should shine through, whether the author self-cites or not :)

Lewin said...

I also agree with the above points. When writing a paper that stems from my theory and builds on my previous work, of course I'm going to cite myself!

One interesting point from the paper's introduction is that becoming and editor increases one's citations. I've certainly been involved with papers where we especially tried to cite the editor (when it was relevant). Not to ingratiate per se, but to avoid irritating him/her by missing a relevant citation (or one that the editor might see as relevant).

Neurobonkers said...

This problem isn't limited to psychology but is an indicator of malpractice in all disciplines.

See the "literature review" that the FDA cite as their single reference for their proposition that MDMA is neurotoxic and you will find the authors or the paper cite themselves ten times including papers that were rapidly discredited.

Review here:

Jess B said...

I completely agree. How can you not cite yourself? Why wouldn't you?

I can understand that citing yourself does not mean you should not cite others. But it is essential in building a theoretical picture to link back to your previous work.

Anonymous said...

Re- self serving bia. This sounds very much like 'self cherishing' which often pops up in modern Buddhist texts .

Hafodia said...

I agree with the points made by Anonymous above - which means there is often good reason to cite your own work if you are building on past work.
Moreover, we are all more familiar with our own studies than others so this may increase the tendency when looking to cite related information.

Samantha Garratt said...

Although I agree with your comment, I disagree that it is stupid that the number of citations a paper has matters. The reason it matters, particularly for say, undergraduates, is that it gives an initial indication of how seminal the paper is, how much use has been derived from it and in some ways can give a rough indication of the quality and validity of the paper. Now this is not to say this is a perfect tool to find these things out but when you need a large number of references it is a quick way to sort through the tons of articles out there to then look over in detail the papers selected on this basis.

Anonymous said...

Is there any particular reason to do anything about this kind of bias?

Anonymous said...

I think the underlying thematic that has been missed by certain comments is - self citation often limits the vista within a field of study. Which maybe amplified by low levels of research productivity in the UK. As such, literature reviews may miss the obvious errors within a study

Post a Comment

Note: only a member of this blog may post a comment.