Press "Enter" to skip to content
- What question do the authors claim they are addressing?
- Why? Is it an interesting question?
- What’s new about this question? Does it use theory that extends beyond what we have studied?
- Data?
- look at Table III Summary Statistics
- big data set?
- lots of “control” variables?
- what can’t they “observe” (important but not in their data)
- empirical methodology – read but skip the really technical stuff
- Difference-in-differences? look at two groups before and after some change
- Dummy variables? Shareit = αi + β0 1[Post]t … indicates a yes/no variable coded as 0 or 1.
- scan initial sentences in sections that appear unduly technical
- example: read only the first sentence of “Section VII: Robustness Checks” (which includes Table XI that at first glance is part of the “Section VIII: Conclusion”)
- results:
- count the asterisks!
- graphs: two groups and before/after. can you eyeball?
- with lots of ups/downs, letting the computer “see” for you is surely more reliable? or do you see what the computer sees?
- What do the authors claim they find?
- Does their analysis really lend itself to their conclusion
- What do they leave out? what qualifications do they make? etc. Do they believe their conclusion?
