Towards better titles for academic papers: a hermeneutic approach from a blogging perspective

Standard

I think a lot of academic article titles are pretty bad. What do I mean by bad? The title doesn’t really tell you what the paper is actually is about. It could be buried in jargon, or overselling an idea, or focuses on details that most of the intended audience won’t care about.

Does the title of a paper affect how it gets read and cited? Probably. In what way? That’s not so simple, based on my short browse of some scientometric findings.

In medical journals, the longer your title is, the more you get cited. And there also a positive effect of using a colon before a subtitle, as in the title of this post. If you’re specific about a particular country, that hurts your citation rate.

In the journal Functional Ecology, you get cited more if you don’t mention the kind of organism you’re working with. If you have a subtitle after a colon, then you’re less likely to get a negative evaluation from a reviewer or an editor.

In the PLOS suite of journals, titles with colons were cited less. Articles with questions in the title were downloaded more, but cited less.  Articles with longer titles had a slight disadvantage compared to ones with shorter titles.

A study of a broad variety of journals in all kinds of fields found that journals with very high impact factors demonstrate a positive relationship between title and citation length, but this effect didn’t occur in journals with more typical impact factors.

Because titles typically aren’t that good, it’s hard to know what kind of results you really will find in a paper from reading the title. This phenomenon gets real old. The timing on this was really funny, as I am writing this post, I looked over at twitter and saw a link to this paper:

“Environmental change explains cichlid adaptive radiation at Lake Malawi over the past 1.2 million years”

accompanied with a note of dismay that, upon reading the paper, there are no frickin’ data about cichlids. How can you explain an adaptive radiation without actually testing questions using data from the animals in question? (Now, since the National Academy of Sciences doesn’t want everybody to readily access the papers they publish, I don’t have a copy of this paper to read at this writing.) Based on the abstract, at least, my gosh they’re talking about lake sediments. The historical changes inferred from these sediments, interpreted in light of the extraordinary diversity of cichlids, can help you sell up a paper from Hydrobiologia to PNAS.

The way I read the abstract, I see this paper is not about cichlids. It’s about mud. Yes, I know they’re studying the mud because of the fish, but the paper isn’t based on fish data. The mud is interesting because of the story it implies about the fish, but the data are about the mud in the lake, right? I’m not entirely alone in this interpretation.

I get it. It’s sales. I know I have to read journal titles with the same skepticism to be employed while reading clickbait headlines. We can’t trust our colleagues — or even ourselves — to call a spade a spade, because the metrics matter. Even a high quality product still needs marketing. Academic moneyball is a thing, even if we want to overlook it or pretend that it’s not.

I’m not claiming my own titles are good, but I don’t think they’re misleading. In some cases, I’ve chosen titles to increase marketability to editors and reviewers. In some cases, like this one, I’ve had absurd reviewer demands suggestions make a title bad. In some cases like this, I tried to make it as broad as possible for sales. But if you spell out too many results in a title, like in this one, it gets weighed down. And in this title, I used the title just to show off a new phrase I coined. And this title is weighty merely because fewer caveats in the title probably would have pissed off the wrong people in review. This is an example of a recent title of mine that I like, and I’m fond of this one too. They’re straightforward, tell you what it’s about, not too cutesy, and aren’t trying hard to be more important than they are.

I think what’s more criminal than a poor title is actually choosing to write a whole clickbait academic article. One that’s about something people feel like they should care about even though it’s not going to make a difference. I would place this one of mine in that category. By packaging the paper with fashionable fluff instead of a more nuanced story with a lot more detail, I sold the paper higher than it would have if I chose to write what I thought was a better paper.

How do you figure out what your titles are? Do you have any guidelines that you follow?

17 thoughts on “Towards better titles for academic papers: a hermeneutic approach from a blogging perspective

  1. I have a (gut) feeling that if I tell in the title the exact species I’ve studied then the reviewers tell that I should use broader names (e.g. pied flycatcher vs. migratory passerine). And also the other way round, if I use a broad name (e.g. insectivorous birds) then reviewers complain that the species names are not mentioned in the title. So, it’s good to prepare few alternative titles already in the beginning. In general, I try to keep my titles as short as possible. No extra jokes or word games.

  2. In behavior research it is notorious to take an experiment performed on one species and then write a title claiming that “males do this, while females do this” leaving out the taxon, subtly implying that the results of this one study are somehow universal. Drives me up the wall, this particular practice.

  3. I’m in the midst of a literature review now, so this topic is near and dear to me. lol. Please, write titles that reflect the purpose of the paper! ugh. I had no idea about the statistics regarding how a title format affects citations… interesting to consider when I formulate a title, but when I research, I always compare the title with the abstract, if they don’t match, or it’s misleading – it just might get tossed off my list on principle alone.

  4. Tangential, but possibly interesting: my husband and I used to have a weekly competition to spot the most strikingly odd title in that issue of the journals we read, and a “first sentence” contest from NATURE. Much hilarity ensued, but it did encourage us to attempt articles beyond our actual expertise. (“What do you suppose THAT’S about?” “I dunno–I’ll take a stab at it…wait…it’s not that difficult…”)

    So bad titles can be good titles if they increase the readership, but some bad titles are just lead-lined coffins that sink your paper forever because no one can fight through them.

  5. I am notorious for colons in paper titles. 2 of my 5 first-author papers have titles with colons. haha. I’ve gotten away from it lately — none of my papers currently in review have colons, primarily upon feedback from my advisor, who hates long titles and also colons. Very interesting statistics on correlation between length/colon and citation rate.

    I’ve also had discussions with my advisor over including vs. not including my organism name in the title. I included it in the title of my paper with a more narrow focus on phylogeography of that species, but after discussion I removed it from the title of a paper currently in review that uses the same species but has a broader focus on use of a particular type of genetic data for salamander studies, and the species itself is a case study. Instead I included the species binomial as a keyword. The thought process included several things you mentioned: paper might be cited more if title isn’t specific to organism, and title is shorter without organism name.

  6. Punchy titles do indeed increase your acceptance and citation rates but might also disappoint potential readers. I own up to doing this myself on occasion – this was published some years ago in an ecological journal – “Sub-lethal plant defences: the paradox remains” . I am 100% certain that if I had used the title that described what the paper was about it would not have been published in Oecologia – e.g. The effect of two lodgepole pine (Pinus contorta Douglas ex Loudon) seed origins (South coastal and Alaskan) on the growth, survival and development of larvae of the pine beauty moth, Panolis flammea (Denis & Schiffermuller) in the presence and absence of predators in a Scottish field site. 🙂

  7. I noticed a while ago that I had a (subconscious) penchant for alliterative titles:

    “Adaptations to arid environments in the Asclepiadaceae”

    “Reconciling ecological processes with phylogenetic patterns: the apparent paradox of plant-pollinator system”

    “Sunbird surprise for syndromes”

    “Multiple meanings and modes: on the many ways to be a generalist flower”

    I’ve stopped now…

  8. Nowadays, given that social media is becoming part of the science “game,” I try to keep my paper titles as short as possible, so that they can fit into a Tweet. Really.

  9. Maybe this is my bias as a paleoecologist here, but I don’t think you’re being fair to the Malawi paper. The context of the paper is the previous arm-waving about paleoclimate’s influence on cichlid radiations. We haven’t had the appropriate data to really test that. This is a poster-child example of adaptive radiation, and it finally has a well-dated, robust environmental context. They don’t claim in the paper to be providing the cichlid data — they’re providing the environmental data that supports an existing hypothesis. Would it not be okay for me to publish a paper saying “dinosaur extinctions caused changes in floristic traits” if I only focused on pre- and post- extinction vegetation, and didn’t also include new dinosaur fossil evidence?

    From a paleo perspective, that’s just not practical. Yes, a lot of our work is descriptive compared to other fields, but it’s also incredibly time and labor intensive to generate those records. And this paper comes with a really nice conceptual model (explained in the abstract!) that underlies the mechanism for the contemporary pattern and phylogenetic work done on cichlids. It’s a cool paper that has direct implications to cichlids, and I’m kind of baffled by the negative response.

  10. A couple of other relevant studies for you:

    Jamali, H.R. & Nikzad, M. Scientometrics (2011) 88: 653. doi:10.1007/s11192-011-0412-z
    http://link.springer.com/article/10.1007/s11192-011-0412-z

    Habibzadeh, F. & Yadollahie, M. (2010) Are Shorter Article Titles More Attractive for Citations? Cross-sectional study of 22 scientific journals. Croatian Medical Journal, Vol. 51, No. 2, pp. 165–170.
    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2859422/

    Sagi, I. & Yechiam, E. (2008) Amusing titles in scientific journals and article citation. Journal of Information Science, Vol. 34, No. 5, pp. 680–687.
    http://jis.sagepub.com/content/34/5/680

    Research Trends (Heading for success: or how not to title your paper (Internal Research): https://www.researchtrends.com/issue24-september-2011/heading-for-success-or-how-not-to-title-your-paper/

  11. I really like nearly all of your posts, Terry, so I’m sorry I’m getting pedantic about one particular post. But this recent blog post of yours has, in my opinion, a poor title: https://smallpondscience.com/2016/10/03/open-source-software-doesnt-necessarily-mean-well-have-better-stats/. I’ll offer an analogy to illustrate my issue with the post: suppose you wrote a post titled “Buying a ruler from local hardware stores doesn’t necessarily mean you’ll get better measurements”. The content of the post, however, was about how you found one local hardware store in particular to be inconvenient and had rulers with instructions that weren’t all that easy to follow. Instead, Wal-mart rulers were a much easier experience all around. The point wasn’t about all local hardware stores at all, but about your issues with one in particular. Further, it wasn’t about the measurements (because, all rulers should provide the same measurements) but about the process required to get to those measurements. This is, admittedly, a caricature of your post (and aligning you with Walmart is unkind) but I’d argue that the analogy is useful to make my argument.

    Pedantry about that post aside, I do think the larger issue with titling articles (or blog posts) is the fine line we all walk when making inductive inferences from our experiences. For many of the students I teach, writing a conclusion is a real challenge because they so often just want to restate the results section. I have to remind them while the results are about what you actually measured or observed, conclusions are about what the results mean. Thus, we are more interested in, for example, how moisture affects the distribution of all oak trees and not just the specific saplings you measured under drought and moist conditions. Of course, there are huge caveats in this process (latent variables, limitations of the timing and duration of the study, etc). These caveats also belong in the conclusion section (but not necessarily the title). Most of the complaints I hear about titles are at the other swing of the pendulum – that the contribution of an individual study is overstated. Many titles that I like provide a succinct statement that describes the main conclusion from the study.

  12. Christopher, there’s no character limit on comments. So there’s no need to resort to caricatures when trying to create an appropriate analogy for a post of mine that came out a few weeks ago. You can take all the real estate you wish to be more clear with your views. I don’t consider you to be pedantic in your point, because pedantry is about taking pains to emphasize accuracy and precision, perhaps with a hint of arrogance. I don’t think your remarks were accurate enough to be considered pedantic.

  13. I apologize for the tone of my comment. I still think the title of the post in question was about one subject, but the content was about another. But you are free to disagree with my interpretation. Here is a link to (someone else) making a point about that post (in 140 characters!) that I basically agree with.

  14. I like how The Journal of Clinical Investigation (JCI) does it. They require that the title is simply the main conclusion of the paper with an allusion to the species. Something like “Protein A causes effect B in the mouse model of disease C.” Pretty much tells you everything you need to know. When I was younger, I thought clever titles with colons and word play were great, but as I get older I just find it irritating. I don’t have time for that stuff.

  15. Great post! I recently had a paper published in a journal that had so many rules on the title (number of characters, no questions, no colons, no complete sentences etc. etc.) I spent half a day just trying to get the title right. My pet peeve is generalised causal statement titles for correlative/exploratory ecological studies from a single system. I think prefer short titles – makes them so much quicker to write out for refs too! Interestingly, I’ve noticed that older ecology papers (e.g pre-2000s) tend to have shorter, simpler titles than current literature.

Leave a Reply, Comment, Opinion, or Remark. You can leave a comment anonymously just by leaving every field blank.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s