Some statistical methods for election auditing

My former Berkeley colleague Phil Stark has written a series of articles on election auditing which might be of interest to some of you. Here they are:

Stark, P.B., 2009. Auditing a collection of races simultaneously. Working draft.

Miratrix, L.W., and Stark, P.B., 2009. Election Audits using aTrinomial Bound. Submitted to IEEE Transactions on Information Forensics and Security: Special Issue on Electronic Voting.

Stark, P.B., 2009. Risk-limiting post-election audits: P-values from common probability inequalities. Submitted to IEEE Transactions on Information Forensics and Security: Special Issue on Electronic Voting.

Stark, P.B., 2009. CAST: Canvass Audits by Sampling and Testing. Submitted to IEEE Transactions on Information Forensics and Security: Special Issue on Electronic Voting.

Stark, P.B., 2008. A Sharper Discrepancy Measure for Post-Election Audits. Annals of Applied Statistics, 2, 982-985.

Stark, P.B., 2008. Conservative Statistical Post Election Audits. Annals of Applied Statistics, 2, 550-581.

Phil has an interesting background: he got into statistics after working on inverse problems in geology. The methods he uses are based on exact error bounds, really much different than the Bayesian stuff I do, much more focused on getting conservative p-values and the like. As a result, the things he does in his papers are nothing at all like what I would do in these problems.

In a larger sense, though, I believe in methodological pluralism, and I’m glad to see a researcher such as Phil, who’s working from such a different statistical framework as mine, work on these problems.

2 thoughts on “Some statistical methods for election auditing

  1. "As a result, the things he does in his papers are nothing at all like what I would do in these problems."

    The idea of replacing mathematicians led to the field of automated theorem proving and although these techniques haven't and will likely never replace us, they have led to interesting insights.

    Does the above quoted remark mean that any attempt at automated data analysis is doomed to be impractical? Does it seem utterly impossible to try to automate the work of a statistician? And I'm not talking about machine learning here. When you sit down with new data and fire up R and do your poking around, how much of that creative process can be automated?

  2. "When you sit down with new data and fire up R and do your poking around, how much of that creative process can be automated?"

    I would argue that the question needs to define its terms vastly better.

    But as to the difference in approach between Professors Gelman and Stark, I adore the many things Professor Gelman has taught me in his many written works and expect to learn many more. At the same time, my mindset is much closer in perspective to that of Professor Stark, and I'd expect I'd have a lot of sympathy with that approach. My reasons aren't some deep-rooted suspicion of Jaynes, who I also respect hugely, but, rather, some notion that I want to make as few assumptions as possible when I model, and assumptions about distributions and priors count against that score when embraced.

Comments are closed.