Reading papers
Reading papers is an acquired skill.
At the start of graduate school, one of my professors told us that the point of first-year classes is to equip you to read a run-of-the-mill journal article. By that token, the purpose of second-year courses is to use your newly honed technical skills to digest and master the existing literature relevant to your research interests. But reading a paper effectively and efficiently is an acquired skill. Here are some tips.
Spend time on the introduction.
I do a lot of my thinking about a paper before proceeding to the body of the text. Is the question interesting? What kind of model do the authors have in mind? Is that model appropriate to the empirical setting? Is the identification strategy sensible? What kinds of threats should we worry about? Are the data up to the challenge, or do they have known limitations that will undermine the analysis? The more you think these questions through in advance, the more you'll pick up as you read the rest of the paper. The introduction is a contract with the reader: by the time you move on, you should have a general sense of what it will take for you to believe the paper's conclusions.
Understand the "story".
A good paper is trying to tell a story, and understanding the basic contours of that story will make it much easier for you to follow the zigs and zags of the analysis and to notice anything that smells fishy. The story will inform the sample construction, choice of specification, sensitivity analyses, bounding arguments, and the sequencing of results. Keep the big picture in mind: don't lose sight of the forest for the trees.
Know what you want to get out of a paper.
Read some papers in great detail, both to develop expertise in your field and to learn how to write good papers. In other cases, skim, but do so thoughtfully. If your interest is in the model, focus on the model. If your interest is in the data, focus on the data. If you're trying to understand how a paper fits into the broader literature, focus on the introduction.
Don't get frustrated if you're a slow reader.
Lots of "straightforward" papers will take a long time to read at this stage of your careers, and many important papers are long, slow slogs no matter how experienced you are. Be patient, and don't fret if it takes a long time.
Keep a searchable library of the papers you've read.
Use a reference manager (like Mendeley or Zotero), or Dropbox, or just a directory on your computer. Wherever you store it, a well-curated library makes it easy to look back at papers you dimly remember, to store notes, and to create bibliographies for papers and proposals.
Keep a reading list.
Strike a balance between your own field and other fields, and between more and less technical papers. Read papers when you're in the mood to be working but not in the mood for brainstorming or writing code. A diverse portfolio helps ensure you'll find something you're in the mood to read. The Journal of Economic Perspectives and Journal of Economic Literature offer great primers on unfamiliar topics.
Use the references to decide what to read next.
If you keep seeing the same paper cited by papers on your topic, it's probably worth reading. Reading paper B's characterization of an earlier paper A is a great way to understand A better.