by Tim Thurman
In the United States and Canada, most phone numbers are seven digits long (without their area codes). Purportedly, this length was chosen because most people can only hold about seven objects in their short-term memory. I’m not sure if this is true: the history of phone numbers in North America looks pretty complicated, and there may be other reasons why seven digits were chosen. But it is nice to think that the seven-digit phone number was thoughtfully designed in accordance with the average person’s working memory.
This (possibly apocryphal) anecdote relates to a pet peeve of mine that I frequently run across in journal articles: authors often use naming schemes (for sampling sites, populations, or experimental replicates) that are hard for readers to hold in their working memory. I don’t want to single out any papers, so I’ll create a hypothetical example to demonstrate what I mean.
Suppose some researchers were studying the per-capita growth rate of populations of unicorns. Perhaps they studied four different populations of unicorns, which the researchers had chosen because they varied in density and habitat type. The rainbow plains (RP) population is in grasslands and has high population density. Sparkle creek (SC) is a low-density forest population. The stardust valley (SV) population is also in grasslands but has low density, while the magic ridge (MR) population is a high-density forest population.
Having read that paragraph once, can you remember which populations correspond to which ecological conditions? If the hypothetical results noted a difference in growth rate between the RP and SV populations, do you immediately know (without double-checking above) what ecological factor may be driving that difference?
Perhaps you have no trouble recalling those details, and this blog post is just an embarrassing admission of my own lack of short-term memory (I made up those populations a minute ago, and I’ve already forgotten). But, my guess (my hope?) is that other people have trouble with this issue as well. Reading a journal article, particularly one about an unfamiliar study organism or research system, requires loading a lot of new information into your working memory. Authors are, of course, very familiar with their own research systems. They’ve spent years studying unicorns at sparkle creek and stardust valley, of course they don’t have any trouble remembering which is which. It can be easy to forget that readers may struggle to keep everything straight.
I’ve often wished that authors would recognize this and pick more reader-friendly naming schemes, instead of using geographical place names or alphanumeric site numbers. In the example above, we could use aliases for each actual location: e.g., the high-density grassland population of rainbow plains is shortened to something like GRA-HI, while the low-density forest population of Sparkle creek is shortened to FOR-LO, instead of SC. Now if I read that there’s a difference in birth rate between GRA-HI and GRA-LO, the possible ecological cause is clear (density, not habitat type).
To be sure, this is a minor point to consider when writing articles. After all, it only takes a few seconds for readers to go back and double-check which ecological information a particular abbreviation or site name corresponds to. And I can think of some instances where this type of naming scheme may be undesirable. Studies which re-visit populations or sites from other studies may wish to be consistent with previous work. In exploratory studies which examine a wide range of ecological variables, post-hoc renaming to focus in on the few factors that were statistically significant may give the false impression of a direct hypothesis test.
Nevertheless, I think it is important for authors to write with their readers in mind. This may require some extra effort on the part of the author. In particular, it may require an extra table in the supplementary material to make explicit the relationship between the reader-friendly site names used in the publication and the site names that the authors used internally during their work (and perhaps in their code and raw data). But this little bit of effort will be worth it: even relatively small changes, like choosing informative site names and abbreviations, can make it easier for others to read, understand, and enjoy your work.
So, the next time you’re writing or presenting results and you find yourself comparing site 187D to site AH4, or discussing the differences between Lake Lapinjärvi and Lake Lappajärvi, remember the lesson of the seven-digit phone number and consider choosing a naming system that your audience will be able to easily remember.
Tim Thurman is a PhD student in McGill’s Neotropical Biology program, co-advised by Rowan Barrett at McGill and Owen McMillan at the Smithsonian Tropical Research Institute. He studies the genetic basis of adaptation, working with Heliconius butterflies in South and Central America and Anolis lizards in The Bahamas. You can find more about him on his website.