Why Lengthy Titles Will Not Necessarily Secure High Citations for an Article, but Field-Specific Scientific Acronyms Probably Help, Along with Suitable Keywords and a Statement-Like Structure…
A recent article by Jacques and Sebire(1) in the Journal of the Royal Society of Medicine suggests that particular features of article titles have a significant impact on the number of times the articles are cited. Most striking are the findings that title length, presence of a colon (:) and presence of field-specific acronyms are positively correlated with the number of citations to date.
Having a personal interest in article titles, as an Editor in Chief, I thought I would attempt a comment on the paper; to do that credibly, I needed some ammunition (data) of my own. Not having much time, I took a look at the top 10 and bottom 10 cited papers (citations to date) in BioEssays (review articles only) each year from 2001-5. Then I reviewed the 25 top and bottom cited articles (research articles only) in the Journal of Biological Chemistry from 2005 (numbers comparable to the Jacques and Sebire study). I can say with confidence that in both datasets (BioEssays and JBC), there is no positive correlation between title length and citations – in the case of JBC, it might even be negative. Bypassing the colon analysis in the interests of time, I went on to analyze the presence of acronyms, and there (BioEssays only), I do see quite a pronounced positive co-segregation: the most cited papers clearly have more acronyms in their titles than the least cited, on average.
What are we to make of that? Well, the difference in the results for titles could be some kind of field-specific one: Jacques and Sebire used medical journals. But they explain their results in a way that should be applicable to most, if not all, scientific literature these days: ‘findability’ via electronic search algorithms (be they in literature indexes or Google-like search engines). Intuition indeed predicts that a longer title would have more scope for accommodating crucial keywords than a shorter one (and a colon might create space for more keywords by making an implicit connection between two concepts); but the length argument doesn’t seem to hold in general. An intuitive explanation for title acronyms aiding citability (if they truly do, in a correlative sense) could be that scientists tend to search for highly specific concepts, with very consistently spelt acronyms (in effect keywords), if they exist, rather than more linguistically variable search strings. And yet there is also something else happening here, because we in publishing know that there is no straightforward relationship between downloads and citations.
The highest cited article from BioEssays in 2002 was The intestinal epithelial stem cell – a short title, with no acronyms. What are we to make of that? As Jacques and Sebire concede, there are many other influencers of citability that they do not cover. The most likely reason for citation of a paper from the review literature is that it is an authoritative summary of a great deal of prior knowledge in a particular field: hence, a short title containing a suitable keyword(s) is likely to be enough to do the paper justice in terms of citability. But that’s not all: the last author of the above review paper happens to be a very established, key, scientist in his field, with a good publication record. A similar article with a more ‘attractive’ title, but from a lesser scientist, might well be downloaded more, but it will probably be cited less; it would certainly stand less chance of making it onto the reference lists of highly authoritative papers from key figures in the field. And that is where the real business of citation starts: authoritative citations make more citations.
It is estimated that no more than 20% of citations of prominent papers involve the citer actually reading the papers in question: they are merely copied from the reference lists of other authoritative articles [2]. In that context, a title is everything: a title that contains good keywords, and expresses the finding explicitly, rather than ‘The influence of…’ or ‘The role of…’ etc. is likely to be picked up and re-cited often.
In conclusion, title/citation analysis cannot per se be developed into a strategy to increase citations: those are overwhelmingly driven by other things. However, it can be useful to optimize a title in order to help an article gain its deserved reader attention, or to catch a substandard title that unnecessarily falls far short of communicating the nub of the paper, e.g. even to attract peer reviews in the first place, see Moore(3). Authors and editors owe it to the papers on which they work to avoid underselling them with a weak title. That is the most valuable role of keyword optimization, suitable use of punctuation, placement of key concepts early, and explicit statement of findings where possible.
References:
- Jacques SJ and Sebire NJ. 2010. The impact of article titles on citation hits: an analysis of general and specialist medical journals. Journal of the Royal Society of Medicine 1:12 DOI: 10.1258/shorts.2009.100020
- Simkin MV and Roychowdhury VP. 2003. Read before you cite. Complex Syst. 14, 269-274 (http://arxiv.org/ftp/cond-mat/papers/0212/0212043.pdf)
- Moore A. 2010. What’s in a title? A two-step approach to optimisation for man and machine. BioEssays 32:3, 183-184
