I’d like to be able to tell you that reading academic research is an important life skill. But the fact is that few people outside a university read academic articles. That is why they are so hard to read – these are university professors writing for themselves and for other professors. But fortunately, most empirical research articles have a very specific structure that makes them easier to understand. In this video, I describe this structure so that you can learn to read these articles more efficiently.
First let me address why I am making you do this. If this isn’t useful outside of college, why bother? First, this class is about empirical research, and to understand it you need to read some of it. But more importantly, one of the most valuable things you learn in college is how to read more quickly. No matter what your future career, it will involve a lot of reading – reading the news and email and policy reports and on and on. A typical leader of an organization (CEO for example) reads around twice as fast as the typical high school graduate. Being able to process written information quickly allows them to do their jobs better and in around half the time of other people. Unfortunately, the only way to increase reading speed is to practice. A lot. And on a wide range of material. So academic articles are some of the hardest things you will need to read, but by expanding your vocabulary, your ability to break down a reading into its parts, and your ability to read graphs and tables, you’ll improve your ability to read everything. Because when you can read these, you can read anything.
So almost every academic article has seven parts. I am not going to read these slides – they are available on our Canvas page so that you can use them when reading articles. Instead, I am going to walk through a research article to show you its elements. The second part of the video, I will talk about strategies for approaching an article (what to read and what to skip, how to take notes, and things like that).
For this video, I have chosen an article that is a particularly good example of the structure of an empirical research article. Unfortunately, not every article is as clearly written, but seeing a good version of an article can help you find these same elements in a less well written piece. Citation to find the article: Potter, Rachel A. 2017. "Slow-Rolling, Fast-Tracking, and the Pace of Bureaucratic Decisions in Rulemaking." The Journal of Politics 79(3): 841-855. https://doi.org/10.1086/690614
Abstract
The first thing to look for in an empirical research article is the abstract. Sometimes it will be labeled Abstract, but sometimes it is just a paragraph set off at the top of the article. An abstract is a one paragraph summary of the entire article. A well-written abstract will tell you the author’s question, their proposed answer (theory/hypothesis), how they tested that answer, and what their results were. These are the essential parts of any article, and reading this abstract is going to tell you the key points you need to know and allow you to decide whether it is worth reading the rest of the article.
Introduction
Unfortunately, it’s not always possible to summarize an entire article in a paragraph (and some journals don’t require abstracts for their articles). The introduction is the first page or two of the article – everything up to the first header. An introduction is going to do everything an abstract does in a little more detail. In particular, it will have a summary not just of the research in the paper, but of why it is important. It will describe the different theories that try to explain the outcome the author is studying and explain how they came up with their hypothesis. And it will provide more information on the data the researcher gathered on the question. So I always read both the abstract and introduction of an article carefully.
Literature Review
The next section is a review of prior research on the topic the author is studying. We call this a literature review, but it is rarely labeled that in an article. The purpose of this section is to set up the hypothesis the author is testing – establishing that it is an important (or new and relevant) question, establishing that there is disagreement about the answer to their question, which then leads to a discussion of how the author’s research will help resolve that debate. A literature review typically categorizes existing research into a few groups and sets them up in contrast to one another. They can be great ways for you to learn the background on a topic.
Theory/Hypothesis
As noted, authors will then describe their answer to their research question. Sometimes they develop a full theory, sometimes they describe a series of tests of different theories – it varies depending on the author’s goals. It is always great when the author writes explicit and clear hypotheses, as this one did, setting her hypotheses off as H1, H2, and H3. This makes it so much easier to figure out what they are testing, but you should always be able to identify a specific sentence (or sentences) that says what the author is testing.
Data and Methods
The author will then describe what information they are gathering or analyzing as part of their study and explain and justify the method they will use to analyze it.
The data section may be separate from methodology if it is a particularly important part of the research, or it may be combined with the method if the author is mostly using pre-existing data. This part will vary depending on the type of data used. For a survey or experiment, this section will cover the types of questions asked or the way cases were selected. Text analysis will talk about how the author chose what texts to look at and what the author was looking for in them. For an article with statistical analysis, this section will describe what datasets were used and how new variables were measured and calculated. All can be helpful to know to evaluate the quality of the author’s research.
The methods section will be the hardest for you to understand, and I don’t recommend you get too caught up on the details here (this is what you learn when you get a PhD). Rather than stress about what a “nonparametric Cox model” is or whether that was the right choice, focus on understanding the overall category of research. Is it statistical analysis, interview research, a field experiment? In this case, this is statistics, so that will tell you to look at the statistical significance and sign of the results. We’ll spend the second month of class reviewing how to understand these, even though you aren’t going to learn what exactly a hazard ratio is. You’ll be able to get the overall picture, and don’t need to know the details.
Results and Analysis
After the author sets up how they will conduct the research in the data and methods section, they will present their results and comment on those results. You will read these sections differently depending on the type of research. For statistical research, you want to focus on the graphs and tables. For other types of research, you will need to read in more detail to understand their conclusions. But what you should focus on here is whether the authors actually did what they said they did in the introduction – did they actually demonstrate a causal relationship? Or is their analysis weaker than they claimed?
Conclusion
Finally, all papers have a conclusion. This section typically talks about the implications of the research and areas for future study. But as I will explain next, no one really reads the conclusions. The summary of this study is at the beginning of the article, not the end.
How to read a research article
So those are the sections of the paper, how can you read this quickly? First, let me be clear: no one really reads these articles from start to finish, and you should not either. The way academics can read dozens of these in a day is by focusing on the key points and what they need to get out of the article. I recommend you approach an article in three stages:
1) Get an overview
2) Read for content
3) Think and evaluate
The first thing to do when approaching an article is to understand quickly whether it is relevant to what you are interested in. Read the abstract and the heading titles, look at whether there are graphs or tables, and get an overall understanding of what type of research this is. When you are doing background research for your research design, you may find an article that has an interesting title, but reading the abstract you may realize it is not actually relevant to what you want to study (or that it similar to something else you have already read). So you need read no further – you can just stop here and move to the next.
If this does look like something you are interested in, then read for content. I always read the introduction carefully, but then I pick and choose what is most relevant to me. If I am new to a topic, I will read the literature review to better understand the issue. If I am interested in their new source of data, I might read that section carefully. As an academic, I read the methods section carefully, but as noted above, you should not do this. Then I read as much of the results as I need to understand their analysis. For statistical papers, you can usually just read the tables and graphs. For other research, I’ll skim their case studies or interview results or whatever.
The worksheet I prepared for you for your first article has the key questions to find in an article, so you can focus on finding the answers to those questions in any article you read. What is the cause? Effect? Alternate explanations? What data did they use? And so on. You will want to take notes on those questions on any article. And you can skim things that don’t answer those questions.
Finally, you should take time to think about the article and evaluate whether it is good research or could be improved. For me, the first step is to try to summarize the article in a single sentence. This is important for two reasons: 1) You are never going to remember everything about an article, but can keep track of a sentence, and 2) When you write a literature review, you need to be very concise in talking about existing research, so this helps with that.
Then you should think about whether you though this was a good example of research. Did it follow the methodology it said it would? Did its data measure what the concepts well? Did it have strong internal and external validity? We are going to cover how to think about good and bad research over the next few weeks (in other words, what validity means), so this will help you answer those questions.
Finally, I just want to urge you not to get discouraged the first few times you read this type of article. It is going to have words you don’t know and be hard going, but stick with it. Again, it gets easier with practice and time.