Mark Bucciarelli writes:
I’m interesting in applying the Bayesian ANOVA you described in your 2005 paper to some data I am analyzing.
Is your arm package for R the place to start? (I’m on Linux/Mac, so I’ll have to build OpenBugs and maybe update the RBugs package.)
Or is there a more direct path?
I’m analyzing the impact of ad attributes on the variance in click rates; e.g., product category, time of day, graphical vs. text, etc, etc.
My reply: For most of the computations in that article, I actually used a Fortran program that I wrote (and ran from Splus). There’s no way this could be recovered; it would be easier to start from scratch. For the last example (in Figures 6 and 7), I used Bugs; actually, I repeated this example in my book with Hill. For your example, I’d suggest Bugs if your sample size isn’t too large. Or you could try Doug Bates’s lmer() function which we use in our arm package. lmer doesn’t currently express uncertainty in the variance parameters but it’s a good start, certainly much better than trying to use R’s aov() function or similar procedures in other packages.
I did Banova (good word) for the first time for an engineering lab a while back, and I used WinBUGS (first time). As a template, I used a program out of Bayesian Methods for Ecologists that I found through google (http://arcue.botany.unimelb.edu.au/bayes.html). This worked out pretty well.
In R2.9-0 VarCorr(mcmcsamp(lmerModel, 10), "varcov") works (OK, with >10 iterations).
I'm about to listen to Martyn Plummer talk, so I feel obliged to point out that JAGS will also work. :-)
MCMCglmm could be another option.
The database community has policies about this. See http://tods.acm.org/Authors.html#ManuscriptPrepar…
In particular, they say "The solution is to reference your prior published work in the third person (just as you would any other piece of work that is related to the submitted paper). This allows you to set the context for the submitted paper, while at the same time preserving anonymity."
They also have thought about some refinements.
"Referring to any work (including but not limited to your own work) that has been submitted elsewhere for review (and hence is as yet unpublished), or that has been accepted for publication at a referred conference, referred workshop, or a journal for which proceedings (printed or online) will be made available after the submission of your TODS manuscript, requires a different protocol to ensure consistency with double-blind reviewing. In the body of your TODS submission, you may refer to such work in the third person as follows.
The authors have also developed closely related techniques for query optimization [Anonymous], but…
In the above example, the reference in the bibliography would then read:
ANONYMOUS, details omitted due to double-blind reviewing.
Note that you should mention neither the authors nor the title or venue of publication while describing anonymous citations like that above. Be sure to place all anonymous citations after the list of your regular citations. Of course, this does not mean you are not responsible to provide the details of such anonymous citations. You should disclose (for use by the Editors only) full details of each such anonymous citation in your cover letter, which will not be made available to the reviewer. Furthermore, you may be asked to submit a copy of one of these papers corresponding to such citations. An editor will contact you during the review period if this becomes necessary."
"Technical reports (or URLs for down-loadable versions) of your own work should not be referenced. Self-references should also be limited to only papers that are relevant and essential for the reviewing of the submitted paper. "