What do legal professionals need to know about statistics?

[sctyner]

Context A series of "Ten Simple Rules for..." articles in PLoS Computational Biology detail several "soft skills" needed by scientists to succeed professionally. This series inspired someone at ASA to ask Danica and Sam to propose articles for statisticians starting to work in forensic science applications, for forensic scientists starting to work with statistics, and for legal professionals encountering statistics.

Please provide feedback on the following list of possible rule topics for legal professionals encountering statistics

1. Statistics and probability are not equivalent.
2. Objective v. Subjective probability
3. Representation of statistics matters
4. Understand & question assumptions
5. Sources of error
6. All statistical results (estimates, models, etc.) inherently have some uncertainty.
7. Independence is rarely satisfied.
8. Understand Bayes theorem
9. At what level is the statistic being reported? (HoP)
10. Understand the scale of numbers
11. Interrogate the data

If you think this list missed something, please let us know.

So what? These TSR articles often have several authors, so it would be great to have collaboration on these topics. This is a great way to get our message out.

[sctyner]

Revised list of rules:

1. Interrogate the data
2. Expect and understand uncertainty (combine 5 & 6 above)
3. Scrutinize error rates
4. Do not equate statistics with probability
5. Question assumptions (combine 4 & 7 above)
6. Representation matters (combine 3 & 10 above)
7. Understand Bayes' theorem
8. Determine the hypothesis (hypotheses) being tested/questions
9. Don’t treat subjective probability as objective probability
10. Misuse of statistics has and will continue to lead to miscarriages of justice

Feedback is still welcome!

[sctyner]

@danica-ommen and I worked out a list. We're moving forward.