Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: comparing groups with categorical variables

From   Austin Nichols <>
Subject   Re: st: comparing groups with categorical variables
Date   Tue, 23 Mar 2010 13:33:00 -0500

Ariel Orson <> :
Make dummies for each of the three categories, and tab versus each.

On Tue, Mar 23, 2010 at 1:29 PM, Ariel Orson <> wrote:
> Thanks to you both. It's my fault that I wasn't being clear. It's
> actually more complex than that.
> Rephrase: comparing three groups on the dimension of not one, but
> multiple attributes: media preference, political preference, religion
> etc. All data are collected through questionnaires and replicated
> bootstrap weights are used in tabulations. I tabulate each attribute
> separately. At the "svy:tab" command for each attribute, Stata outputs
> tabulations with uncorrected/design-based Pearson. But my aim is to
> find whether among the three groups (one of the groups is the base
> group), the two other groups each are different from the base group.
> Thanks in advance,
> Ariel
> On Tue, Mar 23, 2010 at 12:15 PM, Jeph Herrin <> wrote:
>> In addition to Maarten's hint, you might try
>>  -help tabulate twoway-
>> hth,
>> Jeph
>> Maarten buis wrote:
>>> --- On Tue, 23/3/10, Ariel Orson wrote:
>>>> How do I compare groups with categorical variables?
>>>> Say I have three groups: men, women and children. I have
>>>> data on whether their favorite choice of media is book,
>>>> newspaper, TV or film.
>>>> How do I test whether their tastes are statistically
>>>> different from each other with Stata?
>>> Sounds like a homework question to me, so I will
>>> only give a hint:
>>> -help mlogit- in combination with -help test-

*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index