[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: Identify Categorical/Dichotomous and Continuous Variables

From   "Verkuilen, Jay" <>
To   <>
Subject   RE: st: Identify Categorical/Dichotomous and Continuous Variables
Date   Mon, 6 Oct 2008 13:48:57 -0400

Martin Weiss wrote:

>Also, note that the distinction would not be easily achieved (how does
>go about this business?): 

This is one of those areas I warn students---many of whom feel the call
of the Dark Si^H^H, I mean SPSS---about. 

The answer is, SPSS "just does it" and if frequently does it wrong. 

Example: I gave students a design matrix in my categorical class to fit
the Bradley-Terry-Luce model for the tennis players given in Agresti's
book. The course was written around Stata but I didn't require it and
some people were using R, SAS and SPSS. Someone volunteering to use R or
SAS I don't worry about; SPSS users worry me and in this case I was not
disappointed: SPSS took the design matrix I gave them and made it a
"factor," which meant the model they were supposed to fit got turned
into something else which was not statistically equivalent. It did it
without warning. 

Stata won't be suggested in the future... it will be required.


*   For searches and help try:

© Copyright 1996–2023 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index