Academics are often encouraged to write clearly and concisely, but that imperative may actually limit a paper’s impact on scholarship.
A new study out of the University of Chicago has found that papers with longer, jargon-laden abstracts are more likely to be cited in other academic works than are brief, clear abstracts, which researchers are typically taught to write. The scholars behind the study compiled 10 rules that academics are urged to follow in writing, but found that the abstracts of the most-cited papers do not heed those rules.
In a review of about one million abstracts, the three-person research team analyzed each one for clarity by evaluating metrics such as sentence length, parts of speech, and how much emotion was included in word choice. The researchers drew on that analysis to develop a mean score for each abstract. They then compared that score with how often each article was cited, theorizing that the most-cited articles were easily found through search engines.
The research team looked specifically at articles in the top journals in eight scientific fields over 17 years. They selected the top journals using Scopus, a database of abstracts and citations of academic articles.
- How Can You Fix Your Writing?
Do you have colleagues, grad students, or friends who could use a little straight talk about their writing? We’ve reprinted Steven A. Pinker’s manifesto as part of a handy booklet that’s designed to be shared. And we’ve got solutions to bad writing, too: The guide comes with advice from four experts about how to fix what ails you. It’s free; download it by following the link above.
Cody J. Weinberger, a graduating senior at the University of Chicago and the lead author of the study, worked with two professors in the university’s computational-biology lab, Stefano Allesina and James A. Evans, who were co-authors. Mr. Allesina had done research on how science is conducted, and this study followed up on his prior investigation of the effect of a researcher’s national origin on the prominence of the journals that publish his or her work.
The Chronicle spoke recently with Mr. Weinberger about the study and how the digital shift is changing academic writing. The conversation has been edited for length and clarity.
Q. What surprised you most about your findings?
A. It’s pretty much completely opposite of the common advice. They tell you, “Keep it as short and succinct, to the point, and have one main point per abstract.” And our findings are almost completely opposite, not only for one subject, but it was across fields. And even more than that, it was pretty uniform across all the journals in each field. So how unanimous the finding was, and the fact that it’s opposite what most people are told to do, is very surprising.
Q. Why do you think it is that academics are more likely to cite a denser article rather than a more understandable article?
A. One of the hypotheses we propose in the paper was retrievability. Most scientists, they’re not really flipping through the physical journal anymore. Instead, they’re going on the web, putting in the terms most relevant to their research, and looking at whatever comes up. So if you have a longer abstract with more specific jargon than people doing similar research, others are going to find that research first when they type in the search term.
Q. You mention in the article that most journals typically have strict requirements for their submissions. Do you think that journals encourage researchers to publish longer, more dense articles?
A. No, I think the opposite. I think traditionally the word limit was more because of print restrictions, and the cost of printing. But it’s kind of stayed as almost like a status thing for some journals, where you have super-short abstract requirements. It’s kind of our view that it’s not even unnecessary, but kind of restrictive. Because we’re finding scientists, they’re citing papers that have longer abstracts, which means they are probably easier to find.
Q. You also mentioned that search engines favor longer articles. How so?
A. We didn’t test this — this was one hypothesis — but most search engines look through all the words in the abstract and the key words of the paper, and that’s how you get a match when you search for papers. So it’s looking for how many words match that subset of words from the paper. So if you have a longer abstract, it’s just a larger sample you could potentially match. And if they’re more specific, then if you’re looking for relevant papers to what you’re doing, you’re going to type in the most specific things to your research first. So if a paper includes those key words, they’re going to turn up first.
Q. How are search engines becoming more prominent ways for researchers to find articles in the first place?
A. A lot of researchers are experts in their field, so they know the background, but when they’re writing up a paper, the introduction obviously needs to cite everything. So I think a lot of researchers already know what they’re aiming to write for the introduction, but just go to a search engine to find papers that substantiate that background knowledge. There’s no reason to flip through a thousand articles in a physical journal when you can just type in what you’re looking for.
Q. Do you think that it’s better for academics to write with the intention of being cited, or for being understood more broadly?
A. That’s the trade-off we’re getting at. I think definitely we want people to write clearly to communicate the ideas in their paper, and I don’t think it’ll get published if it’s not written that way, but there’s still kind of a disconnect between what’s going to get published. If it’s super-lengthy and terribly written, it’s probably not going to be accepted. But I think more the idea that we need strict word limits or even strict page limits on papers — that might not be as relevant as we think.