Friday June 30, 2006
This TV-B-Gone uses a 9V battery to send its signal through a matrix of 20 IR LEDS. This extends the working range of the device to about 90ft (line of sight). Using this in a regular sized room you are pretty much guaranteed to kill the TV no matter where you point it.
read moreÂ |Â digg story
Comments Off on Ultimate TV-B-Gone – turns off TVs from 90 feet away
Wednesday June 28, 2006
The Supreme Court today ruled on the “Texas redistricting case” (actually a bunch of cases, consolidated…): League of United Latin American Citizens et al v. Perry, Governor of Texas, et al. A few political science colleagues (Gary King, Andrew Gelman, Bernie Grofman, and Jonathan Katz) wrote an amicus brief over Christmas (shameless self-promotion alert: while we were in Australia on vacation, so I missed out on participating on this, but it was nice to be asked…).
The basic position argued in the amicus brief is that partisan bias in an electoral system can be measured/estimated, via the partisan symmetry concept that undergirds work on seats-votes curves that goes back to Tufte and that (through some very nice statistical modeling) Andrew Gelman and Gary King have made operational in the case of data from a single election. A majority of the court didn’t go along, either not addressing the question or stating that a reliable, implementable standard for assessing partisan gerrymandering is still not available. Still, my quick read is that Stevens was the single biggest fan of the symmetry standard, and that at least Souter, Ginsburg, and Breyer are interested, with a little more skepticism from Kennedy.
The best line in the opinions is actually in a Stevens footnote:
proponents of the symmetry standard have provided a helpful (though certainly not talismanic) tool in this type of litigation.
Uh-huh. Thats political methodologists for you: helpful, just not talismanic…!
Thursday June 22, 2006
Jan de Leeuw put me onto PathFinder. A bunch of us (well, Andrew Martin and I) spied that he had a very slick enhancement to Finder running on his desktop, based on what we could see on the “big screen” in the AudioMax at TU-Wein. Said Jan:
That’s PathFinder, which is not an Apple product, but shareware from
http://www.cocoatech.com/. It has many Finder enhancements, some more
useful than others. I start it at login time and then ask it to close
I’ve got the demo running, nice… Might well keep it (i.e., buy it). But one wonders what Finder in 10.5 will look like.
Comments Off on PathFinder, alternative to Finder on Mac OS/X
last Sunday we put 6 of our PhD students over the top, including Matt Levendusky and Sarah Anderson, pictured below. Matt’s off to UPenn and Sarah has got a gig at the Hoover Insitution at Stanford.
My blue robes are Rochester colors: Stanford bought me my robes when I was promoted to tenure a few years back (nice touch).
And, I’m assuming, my last commencement as Director of Graduate Studies.
Comments Off on commencement 2006
This one comes up a lot, so I thought it deserved a blog entry.
On Jun 19, 2006, at 8:07 PM, Tianji Cai wrote:
I have a data set which has 200 clusters; however, 10% of clusters are singletons, which mean that any one of those 20 clusters only has one case. Any one in the left 90% of clusters contains 10 cases. How could I modify my previous code to fit this data?
Thank you for your help!
[my reply below]
Comments Off on unbalanced data, BUGS, Bayesian modeling
Friday June 16, 2006
useR! has been a great conference, the first time I’ve felt “older”: at 40, I am the oldest of about 12 polisci folks here, we are the best represented social science here by a long shot, its really quite exciting to see this much push into a mainstream stats meeting by younger polisci folks. wow. all the major presses represented here (Wiley/Springer/CRC) and talking up pushes into soc-sci methods etc.
Comments Off on Wien, wrap up
Thursday June 15, 2006
Jan de Leeuw on Psychometrics. Lots of disciplines, generically discipline foo, have Foometrics. Classic Jan. Most psychometricians use SPSS, some SAS; not a lo of interactino between psychometrics and statistics (“…historical reasons, most of them silly”). Psychometric software distributed by incorporating it as modules in standard packages (SPSS, SAS, Stata); guarantees good distribution, some money, but certainly not efficient computation (e.g., CATEGORIES for CA in SPSS, CALIS for SEM in SAS, gllamm in Stata). Standalone: SEM packages like LISREL, EQS, M_PLUS, AMOS, or MLA packages like HLM, ML-WIN (standalone companies, by psychometricians, who work for universities, and so are poor, and are trying to make some money). Black-box, proprietary; machinery completely hidden (because it is proprietary); but models/programs complicated, many parameters, complicated optimizations, doubtful standard errors (simple stuff in SAS/Stata/SPSS, hence harder stuff in commerical products). Hence, R in psychometrics has advantages:
- distance to academic statistics becomes smaller
- software is more transparent, driven by interpreted code. Reproducible results are more likely
- one can teach with R. One teach SAS, but one cannot teach with SAS, or LISREL
- Software should be free
psychoR; “let a thousand flowers bloom”. http://www.cuddyvalley.org/psychoR. Special issues of Journal of Statistical Software.
- simple and multiple correspondence analysis: MASS, FactoMineR, homals. ade4, PTAk; homals to become gifi. distaccoc, scalassoc, singlepeaked, logithom (IRT variants).
- ltm simple Rasch model (and extensions). mprobit. logistic IRT in VGAM. simple unifying algorithm for doing lots of latent variable models, including most IRT models.
- factor analysis: factanal in stats. MCMCpack (!) ordinal and mixed factor models. Related to IRT. See also homals
- three-mode analysis: PTAk, various form of k-mode component…
- SEM: sem (R package) using the RAM specification. Needs a lot of add-ons to compete with the stand-alone SEM packages.psychoR does least squares SEM.
- multidimensional scaling: non-metric MDS in MASS, labdsv, ecodist, vegan and xgobi/ggobi. Kruskal-type least squares loss functions. Classic Torgerson metric MDS, and principal coordinate analysis (Gower) in …. psychoR has metric and non-metric least squares multidimensional scaling.
- HLM and LogLin not discussed because they are mostly outside Psychometrics…
Comments Off on Wien, useR!, Day 2
so weird, here we are on a beautiful day in Vienna, its a public holiday and the streets are pretty much deserted, but there are 200+ of us in this huge auditorium at WU Wein to talk about R. International geekery in full effect.
and I count a lot of people from U.S. political science departments. how cool. Kevin Quinn, Andrew Martin and me, Luke Keele, Micah Altman, 3 or 4 Harvard students (including Ben and Olivia), and Jake Bowers.
Jan de Leeuw was in the 3rd row as I walked into the auditorium; a new MacBook open, running Parallels, with debian linux and XP running: “many versions of R running at once”, said Jan.
Comments Off on Wien and UseR!
Thursday June 8, 2006
I’m enjoying the following:
GPGMail (nice integration of GPG crypto/decrypto into Mail).
letterbox (Mail goes into 3 column mode, using as much desktop as want to give it; great productivity enhancement given the amount of e-mail I/you/we deal with…)
I am also going to take a look at Mailtags at some point.
Comments Off on enhancing Mail.app (Mac OS/X)
Wednesday June 7, 2006
Kids says the darndest things. This morning, at home:
Simon (me, Father): “Tom! Don’t hit Jo-Jo, she’s your friend.”
Josephine (daughter/sister): “He’s not my friend, he’s my brother…”
Comments Off on Art Linkletter moment