On 6 Jan 2004 08:25:53 -0800, [EMAIL PROTECTED] (Kea) wrote: > Hello, > I have two question regarding the definition of "odds": > > 1. Which statistical technique do the "odds" correspond to? Is it [ snip, about Bayes; probability; who introduced it ... ]
Gamblers have been talking about "odds" for centuries. I think. The *modern* statistical usage is the one with "log-odds". The log-odds were encouraged by the development of additive, log-linear models for multi-way contingency tables. Bioassay was talking about logits, decades ago, partly because they were more convenient to compute than "probits." (Neither was very convenient, before computers.) Also, a few decades ago, heightened interest in health (tobacco, asbestos, industrial exposures) meant that "survivorship curves" grew in popularity, along with several related concepts including relative-risks. However, true "relative risks" are relatively intractable in math, compared to odds-ratios, so the latter have been given a bigger stage. And it sometimes happens that Odds ratios are mislabeled Relative risks, because the latter name seems more 'natural.' -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html "Taxes are the price we pay for civilization." . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
