Tips for Law Students Who Are Doing Empirical Research?

I'm also looking for tips that I can pass along (again, in Academic Legal Writing) to law students who are doing empirical research (of whatever kind, and I realize there are many kinds). There, too, I already have quite a few, but it would be great to have more. If you have any suggestions, I'd love to hear them. Thanks!

The Reference Manual on Scientific Evidence is something every student and practitioner who relies on statistical evidence should be familiar with. The Manual bridges the gap between statisticians and scientists on the one hand, and judges and lawyers on the other. At first glance, this might seem like its utility is restricted to practitioners, but it gives simple, useful explanations for presenting and verifying scientific arguments.
6.14.2007 8:24pm
byomtov (mail):
I don't know what the manual fffff recommends contains, but I think that learning statistics - maybe even (shudder) taking a class - would be very valuable. I'm not a lawyer but I've read a law review article or two, and the statistical analyses are often quite poor.
6.14.2007 8:36pm
Passing Through 01:
Consult statistics professors. They can provide tips on how to collect empirical data. Further, when it comes time to interpret the data, professors can help (i) challenge "obvious" conslusions the student has drawn and (ii) offer alternative methods of analyzing the data.
6.14.2007 8:39pm
frankcross (mail):
I would note that you don't have to know statistics to do meaningful empirical research. Simply counting cases or facts can provide valuable information. One needn't be inordinately ambitious.

I would recommend checking Epstein &King's Chicago Law Review article from a couple of years back. It nicely sets out criteria for empirical research
6.14.2007 8:49pm
First things that come to mind:

1) Sample size is crucial.

2) Often, people make conclusions that go beyond the data. Don't report what you think the data means unless you have good support for that conclusion based both on the findings and on prior literature. Report the data and let it speak for itself. If you must speculate, remember that your speculations are tentative - and make it sound that way (liberal use of the word "may" is helpful - e.g., "These data may suggest that ..."). Remember that there are always multiple alternative explanations for your findings, including the most likely of all - that your research was flawed.
6.14.2007 9:00pm
byomtov (mail):
I would note that you don't have to know statistics to do meaningful empirical research. Simply counting cases or facts can provide valuable information. One needn't be inordinately ambitious.

True. But then the author needs to be very cautious about stating conclusions.
6.14.2007 9:07pm
Sarah (mail) (www):
Be careful about empirical research that involves human subjects -- there's a reason many (well, at leas three that I know of -- the only three I have any experience with) require a whole class on working with human subjects before letting undergrads or graduate students do any studies. I had to sit through two class sessions in our statistics class on the topic, and probably no more than one (on the outside, two) of us ever even tried doing a simple survey. I know I didn't -- mostly because you had to get departmental approval, waivers, and signed consent from every participant. And I was in political science.

Also, be careful to clearly state your expectations, and avoid letting them unduly affect the way you conduct your research. Consider doing a small scale trial before spending time and (hopefully someone else's) money on a question that might not be a question. Says the girl who lost an entire weekend to tracking down a (phantom) gender bias in crack cocaine sentencing, which turned out to just be a database which was not collected in anything approaching a uniform manner (turned out that sometimes they just skipped over the bit with the jail term, and only wrote down how long the driver's license would be suspended for. Sigh.)
6.14.2007 9:22pm
Former Law Review Editor:
Please don't.

That's basically the best advice I can give.
6.14.2007 9:26pm
Thaddeus Mason Pope (mail):
Two tips:

1. I would tell readers that they should consult subject-specific advice. The tips for law &psychology may differ from those for environnmental empirical research. One highly-regarded discipline-agnostic source is Fred Pyrczak &Randall R. Bruce, Writing Empirical Research Reports (5th ed. 2005).

2. I would refer readers to the growing scholarship on empirical legal research itself. e.g. (collecting links)
6.15.2007 12:34pm