, , ,

Like many academics, I spend a fair bit of time reviewing other people’s work.  Whether this is manuscripts from collaborators or students, submitted articles from journals, or grant applications, there are few weeks in the year where I don’t spend some time thinking about what suggestions I can make to authors so they can improve their writing, their science, and their science communication.

What’s the first thing you do when you read a paper? Chances are, you read over the title, perhaps the author(s’) name(s), and the abstract.  After that, you probably flip to to figures and tables, since they’re likely to give you an indication of the kind of analysis, and a glimpse of the results.

But if there’s one constant in all the reviewing I do, it’s pointing out that the figure and table legends most of us use in scientific papers are terrible, awful, machinations of verbiage.  This is something I only really became aware of during my PhD when a committee member pointed out what has got to be the single paper I point authors to in my reviews:

Kroodsma, D.E. 2000. A quick fix for figure legends and table headings. Auk, 117: 1081-1083.

More often than not, we just describe the contents of the table or figure.  Anyone can see that you’re reporting the length of eggs, or the annual survival estimates, or frequency of death by hippopotamus over time.  “So what?” asks the reader. “What is the author trying to tell me here?”

Use table and figure legends to convey the message you want, not just describe the contents.

Table 2 in Don Kroodsma’s paper gives wonderful examples of how to do this.  For example:

  • Descriptive: “Relationship between territory size and distance from roads for 21 Ovenbird territories …. “
  • Informative: “Territory size of 21 Ovenbirds decreases with distance from roads…”

In the descriptive (and published) caption, we’re told there’s a relationship, and not much else.  By revising the caption to contain the message, we get what the relationship is, and the take home message (i.e., why the figure or table was included in the first place).  If you’re giving an oral or poster presentation, you don’t just point to a graph on your poster or the projector screen and say “And this shows the relationship between the number of people hailing a cab, and those who have attended at least one bat mitzvah“.  Are cab-hailers more or less likely to have attended bat mitzvahs? Are people who go to bat mitzvahs more or less likely to hail a cab? Sure, in this example it’s obvious, but that’s not always the case.

Science communication is not just about communicating science to the general public, but to the scientific community as well.  I’ve argued before that the way we write affects how well people understand it. Figure and table captions are no different.

Characteristics of roads in Saskatchewan. | Roads in Saskatchewan tend to be straight and flat

Characteristics of roads in Saskatchewan. | Roads in Saskatchewan tend to be straight and flat.