I have discovered myself to be in need of a statistical tool I do not possess. I am confident that a frequentist formula exists, based on the nature of the task to be executed, but it occurs to me that there may be people who would like to prove some point about Bayesianism vs. Frequentism - so here's a challenge for you all:

I am a mechanical engineer - numerate, literate, and reasonably intelligent - educated to the extent of one college course in basic probability and statistics. I have also been reading EY's essays for years, and am familiar (approaching level 3) with the introductory Bayes Law material he has written.

What I want is a handbook - preferably available from the University of Maryland, College Park library edit: [catalog link] or the Montgomery County, Maryland public library system edit: [library system homepage with link to catalogs; but necessarily available for less than $30 U.S. (exact cutoff negotiable) - which is likely to include a procedure I can use to analyse my data and act. Optimally, it should be sufficiently clear that I can use my results to justify a course of action to someone else, if necessary. (Feel free to assume I am eloquent for purposes of this additional requirement.)

Should I have both Bayesian and frequentist methods available, I will employ both and report my results in summary form with analytical details in a separate post.