The Three-Pass Approach
You have an essay coming up, and you want this one to be your magnum opus. Picking up an academic paper, with highlighters and a notepad on hand, you read it from the introduction straight to the conclusion. However, whenever you reach the end, you realize that the paper was not of any topical value. Your impressive notes, practically useless. How could this be?
Well, despite any impressive system of note taking or any decent reading routine, this way of reading simply doesn't work. That is not to say that you're a poor reader or that quality note-taking doesn’t matter, rather that Albert Einstein’s General Theory of Relativity should be read differently than Harry Potter.
Reading an academic paper cover to cover is a waste of your precious time, and in some cases—like the example above—it's even fruitless. This inefficient reading style could mean less time to form that perfect essay, and possibly a worse grade.
So how can this problem be alleviated? Introducing Nicole S. Keshav’s Three-Pass Approach!
The three-pass approach is self-explaining, you read the paper in three passes. Duh.
The First Pass helps you understand what the paper is about. Depending on the relevance of the paper, you may not need to read it past this stage.
The Second Pass is where you get to the meat of the paper. Here you want to understand the paper's arguments, evidence, and statistical methods.
It is also important to read the paper’s literature review here. This is where authors usually store background material, recommended readings, and— far more importantly—invaluable context on the problems that this particular study is trying to address.
The Third Pass is the most time-consuming, and it's plain awful. But if you want your argument to be comprehensive and correct, it might be necessary.
The First Pass
With this pass you're mostly just skimming through the article. Here you want to read the title, abstract, introduction, conclusion, and the headings. It may even be useful here to use a technique called the ADIRM method. Which is practically the PEMDAS of academic reading.
ADIRM stands for Abstract, Discussion, Introduction, Results, then Methods. However, for this pass’s sake, skip the latter two on the list.
Something about the ADIRM method should become increasingly clear, it's meant to foster critical thinking. Why should you read the discussion before the introduction? So you can judge whether or not the conclusions made can be logically inferred from the claims in the introduction.
Good writers usually lay out the foundations of their experiment in the introduction and summarize their findings in the conclusion.
Do not read papers from bad writers, even in an academic setting (unless your teacher says to), they take a longer time to read and are harder to get any value out of (thus why I would argue never to read Jean-Paul Sartre, or any philosopher who writes as bad as him.)
After your first pass you should understand the who, why, what, and how's of the paper. The four questions of understanding that have probably been pounded into you since preschool.
Of course, Keshav gives us a more elegant form of these questions: The 5 C’s.
Category; what type of paper are you reading—is it a natural experiment, a methodological design, a meta-study, etc.
Context; what papers, theories, and authors was this study built on.
Correctness; are there any problems with this study? The best way to answer this problem—if you're not a scientist—is to look for replies to this paper.
When browsing for more papers on the minimum wage I ran into a paper titled, “Minimum Wages and Employment: Comment”, where David Neumark and William Wascher claimed that the statistical methods used in “Minimum Wages and Employment” led to some underestimated results. This was later found out to be false, but the critique led to some other (damning) evidence that David Card and Alan Krueger’s (CK) paper was methodologically unsound; and—in the process—to another revolutionary study (that confirmed the CK’s results, but not their methods).
Contributions; what are the main contributions of the study? This could be some interesting findings, a new and improved method, or anything else.
Clarity; is the paper written well? This is less of a question of understanding and more of a is-it-worth-reading question.
If you can answer these questions then you're ready to move onto the more exciting part of the paper. Make sure that you go through any seemingly relevant paper at least once. After your first pass, decide if the papers are worth further reading. Do not be afraid to cull anything you deem boring, irrelevant, or too hard to read.
The Second Pass
During a paper’s second pass, you want to understand what makes it tick. In the ADIRM method, you're going to read the results and the methods; but also the literature review.
Most people downplay the importance of the literature review, heck it's not even listed in the ADIRM method, but it is the most important part of any paper.
You may be forgiven for thinking that the methods or the results were the most important—but you would be wrong. The methods section tells how the study was conducted, the results are, well, the results of using those methods; but the literature review gives us the evolution of those methods and why their results are important. It gives us the previous sources—the big names in the field or topic.
The citations give us something similar, but unlike the citations, the literature review tells which of the works cited are important. Rather than being instrumental, the citations page should be viewed as a shopping list—mark off, mentally or on paper, what you have already read and check out what you have not.
When you are reading through the literature review, and crossing off the cited works that you have already read, be sure to compile a document of all the papers relevant to your topic or field, and use the methods detailed in this post to go over all of them; culling what you're not interested in as you go.
Only the most important, most relevant, and most elegant papers should make it to the seventh hell that is the third pass.
The Third Pass
Keshav tells us to do two things in this pass:
Recreate the experiment.
List and critique every assumption in every sentence of the paper.
Now unless you had ten years of training, recreating the experiment is probably off the table. However anyone with a pencil and a hand can write stuff down, and anyone can think critically with a little bit of elbow grease.
Now you might think that this step requires you to write a thousand pages, however in actual practice most papers only have about 20-50 noteworthy assumptions in them.
Take this sentence, from a Nobel-prize-winning paper by David Card and Alan Krueger (CK), for example, “Contrary to the central prediction of the textbook model of the minimum wage, but consistent with a number of recent studies based on cross-sectional time-series comparisons of affected and unaffected markets or employers, we find no evidence that the rise in New Jersey's minimum wage reduced employment at fast-food restaurants in the state,” (pp. 792)
A lengthy jargon-filled sentence indeed, but what are the assumptions being made here?
Most sentences in this paper boiled down to a few similar assumptions: my methods are reliable, other papers confirm my results, other authors are correct, policy spill-overs don't exist or are rare, etc.
The true headache comes in the form of going through, one by one, and critiquing each of these assumptions.
Are these assumptions true? Not all of them, no.
Related papers confirmed CK’s finding, however his methods contained some serious flaws.
Going through this pass is important, especially if you're writing an essay or trying to become (assuming you’re not one) a scientist. Understanding the assumptions of these papers will allow you to make some contributions to the field/topic, in the form of addressing or reinforcing them.
Keshav noted that “Learning to efficiently read a paper is a critical but rarely taught skill,” (pp. 81) his paper attempted to address this problem. Sadly, however, Keshav’s methods are not well-known (rather, they are not well-known enough.), thus many people still lack the necessary reading skills.
This blog post seeks to address this issue by increasing exposure to the three-pass approach. The most efficient method for reading academic papers that I have found so far.
To end this post, let me give an example of its practical use: in writing a blog post on unemployment insurance, I started with a book on the subject. The book offered a set of interesting papers that, after reading the abstract, I implemented into a list on google docs.
I passed through each of these studies once, striking through the irrelevant or poorly written. Then I went for a second pass through the remaining few, where I gave a large amount of my time to the literature review.
Each study I read gave way to dozens more, until, finally, I was satisfied with my understanding of the subject. Using only a small fraction of what was read,, I formed my guide to unemployment insurance.
Of course, the only studies that warranted a third pass were the ones that resonated with me or the ones that I disagreed with, the latter more so than the former (A bias that I can undoubtedly improve on).
Nevertheless, without this approach I would have never finished my blog post within the timeframe that I had given it. Indeed, I probably would not have finished it at all; the introduction is a projection of my old reading habits.
Original article at Labor Studies (link no longer active).
Chase Stewart is a student at ACTC for his associates in arts, the author of Labor Studies and an aspiring economic researcher. Chase loves studying, speaking, and writing about politics, economics, and philosophy. He also plays chess.