Linda Skitka recently shared a great exercise for helping students improve their writing.*
This “reverse engineering” exercise sounds helpful. I can feel when writing is executed well and when it is not; however, I have never tried to verbalize why the writing is executed well. Thus, before asking my students to complete this exercise, I thought I would do a test run. Also, rather than outlining the paper like Skitka recommended in her tweet, I merely tried to identify what features of the paper worked well.
Below is my first attempt at reverse engineering the writing of an academic article. The purpose of this exercise is not to engage with the arguments of an article per se. Rather, the purpose is to identify features of the writing that I believe worked well, to abstract a lesson from those features, and to put those lessons in my writing toolbox. In short, the purpose of this exercise is to improve my writing voice by honing my reading ear.
You can think of this post as me trying to “think out loud” while I am reading an article.
Selecting an article
I can certainly bring to mind several articles that I feel are written poorly. And I can certainly bring to mind a handful of articles that I feel are written well. An article that is on my “well-written” list is Simons (2014). I assign this article in my undergraduate social psychology lab courses both because it is relevant to the course objectives and because the writing is crisp and economical. For these reasons, I thought this was a suitable article for a test run of this exercise.
Here are my thoughts about why I think Simons (2014) is well written.
Strong start and strong finish
Simons (2014) starts and finishes strong. Here is the first paragraph.
The article could have started with the second sentence: “The idea that direct replication…”. But notice how many little pieces of information are actually in the second sentence: There is a claim that “direct replication undergirds science,” there is support for that argument, and there are many concepts such as “robust,” “competent,” and “statistical power” that a reader must parse. If the article started with the second sentence, it would take the reader a moment to gain her bearings and understand the direction she is headed.
However, Simons (2014) starts with the brief and strong declaration that “Reproducibility is the cornerstone of science.” This first sentence boldly grabs the readers attention and says “we are starting right here!” It is a firm and unambiguous beginning.
Here is how Simons (2014) ends the introductory section.
Without even reading the content of this passage, it is visually obvious there are three main points being made. And having each point on a separate line implies these are important points; so important that the author does not want them to be lost in the jumble of a “normal” paragraph. The format of this paragraph screams out “hey, there is some important information right here!”
If you look at the full-text of the article, you also will notice there is a separate section devoted to each one of these points. This combination of three bullet points that prelude three corresponding sections provides a strong logical structure to the entire manuscript. To see why this works well, imagine if there was a fourth bullet point that did not have an accompanying section later in the article. The manuscript would appear incomplete or sloppy to the reader.
Finally, here is how Simons (2014) ends the article.
This article ends with an equally brief and strong declaration as the first sentence: “Direct replication is the only way to make sure our theories are accounting for signal and not noise.” This last sentence strongly and concisely summarizes the arguments that were made in the article. Also notice the use of “our” when describing theories. This subtly helps the author end the article on a positive note. This is not an author lecturing a group of “others,” but this is a person who is communicating information to a group of readers that includes himself.
Strategic use of levity
Simons (2014) includes this passage when discussing “hidden moderators.”
Understanding the hidden moderator argument is not important here. The important point is that Simons (2014) uses rather absurd examples such as phases of the moon and how much corn participants ate as a child. There is a bit of playfulness in this passage that is rare in the typically-dry academic articles. Authors can risk appearing not serious if they overused such absurd examples. However, this is the only place in the manuscript this sort of levity is employed, which makes it feel as if it was strategically placed rather than intellectually shallow filler.
The use of absurd examples also is a calculated choice. If plausible examples were used, no matter how explicitly they were labeled with disclaimers of being “just examples,” there would be some readers who infer those were actual examples that were referring to actual arguments made by actual people. The reader would then expend mental effort engaging with these plausible examples, which would detract from the general point that is being made. The use of absurd examples allows the reader to easily dispense with the specifics of these examples and to keep their full mental effort onto the heart of the general argument.
Simons (2014) also had a choice here to completely omit the highlighted text. It would have been grammatically correct to put a period after the word “infinite” and no intellectual information would have been lost. However, providing concrete examples, even absurd ones such as phases of the moon, helps the reader understand the abstract concept of “an infinite number of possible moderators.”
Address the arguments head-on
The final notable feature of Simons (2014) is that he does not dance around arguments that are counter to his thesis. Nor does he set up easily-defeated straw men. He fairly articulates a challenge to his central thesis and addresses it head-on. Here is a paragraph where this is nicely executed.
Notice that Simons (2014) points out areas of agreement (e.g., “Cesario is right…”) and uses the entire first half of the paragraph describing an argument in a way that is fair, free of mockery, and is not a caricature. Then, once this argument is laid out, the paragraph pivots on the word “But…”. Not only does the second half of the paragraph directly address the first half, the word “But” provides a clear logical relationship between the two halves: The first half is an argument and the second half is a counter argument. It is critical that the points of the argument is laid out fairly and that the counter-argument directly addresses those points because it allows the reader to feel as if the take-home message is due to an actual engagement with the ideas and not due to an unfair and one-sided framing of the arguments.
Why is Simons (2014) well written? My opinion is that the writing is plain, it has a strong start and finish, it uses helpful and strategically-placed examples, and it addresses the main arguments and counter-arguments head-on and logically.
The other conclusion from this exercise is that it was fun to introspect as I read a well-written article and, based on my experience, my students will now be doing this exercise this semester.
* This post is about trying to improve writing skills. Any comments about how my writing is less than exemplary is merely pointing out information I already know. I just hope that something (anything) I say is helpful to one other person.