Audiophiles, Significance Tests, Greenspun's Tenth Rule ctwardy | blog | SARBayes


My Weblogs
SARBayes blog
Prior Analytics
SCI 410 news
Blogger(?)

Other Weblogs
D4H
ISRID blog
Nuit Blanche
ACFR
Statistical Modeling...
Machine Learning Theory
pfh
njh'
Older njh
jpl-rpl
...More coming...


A great little post on Decision Science News.

A gem that packs lots of great quips together.

1. Greenspun's finding that people can tell high-quality audio cables from low-quality ones, but prefer high-quality ones only about 15 times out of 28. (See recent amusing post by njh, and coda.

2. To analyze his results, Greenspun rederived stats tests because "status books are virtually impossible to understand" and "I was able to explain what I was doing using simple probability laws that most of the readers would have understood." Similar reactions are common to the nice book Bayesian Statistics. Math people prefer probability.

3. Snarky programming quips.


Thanks to Statistical Modeling from whence I found it.




[æ]