From the inbox:
Dr. Deidre McCloskey,a Distinguished Professor of Economics, History, English, and Communication, University of Illinois at Chicago, will give a talk titled "Use and Misuse of Statistical Methods" on Thursday, April 9 at 3 pm in room 118 in Anne Belk Hall. This is a talk that is accessible to all, without background in statistics.
This event is free and open to the public and sponsored by Appalachian Spring Conference in World History and Economics (further details can be found here: http://history.appstate.edu/news-events/appalachian-spring-conference-0), for which Dr. McCloskey is the keynote speaker.
The Department of Economics was one of the sponsors.
The talk was on statistical vs. economic significance. The former is misuse and the latter is use. I totally agree with the thesis if all the researcher does is consider t-values and p-values. I still referee the occasional paper that ignores effect sizes. I also still referee the occasional paper that discusses a coefficient estimate that has a standard error such that the p-value is greater than .10. I routinely scold the author about how this coefficient is not statistically significant and shouldn't be discussed. As McCloskey said yesterday this is nonsense because p = .10 is an arbitrary cutoff. A large effect at p = .11 might still be economically important (also, a one-tailed test might be more appropriate for things such as price [and scope] effects, but economists seem to disdain one-tailed tests). I'll try to to better with this in the future.
Afterwards I introduced myself and thanked her for coming to Boone. She asked me what was my field, I said something like "environmental, we don't have much of a problem with ignoring effect sizes since it is mostly policy driven." She nodded in agreement and said something like "policy makers don't care about your p-values."
I mentioned that I always read the "Other Things Equal" articles in the Eastern Economic Journal way back in the 1990s. These were short pieces that appeared at the end of each issue. The ones I still remember and recommend to others are:
- Economical Writing: An Executive Summary [pdf] - This is one I ask seniors to read before they commence writing their papers in the seminar course. It is short enough that undergraduates won't begrudge you for assigning it and heavy enough that, if taken seriously, can help readers of their papers.
- To Burn Always with a Hard, Gemlike Flame, Eh Professor? [pdf] - To avoid burnout (and I began this profession at the age of 26, so I've always worried about burnout) you need to work on the things that most interest you (e.g., don't read the AER just to try to figure out how to publish in the AER).
- How to Host a Seminar Visitor [pdf] - It is about promoting your department just as much as it is about learning something from the speaker.
These, and almost all of the others, are worth a re-read.