turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- A question about Tukey test

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

08-26-2011 02:20 PM

Hi all,

When doing Tukey's HSD, SPSS gives the results of a homogeneous subset (ie., categorizing different groups into different subsets). Just wonder if the same analyses can be done in SAS? If so, how to write the syntax for this?

Thanks,

Chester

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to PsycResearcher

08-26-2011 04:20 PM

I'm not certain, but I think you are looking for the LINES option. Take a look at: http://udel.edu/~mcdonald/statanovaunplanned.html

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to PsycResearcher

08-29-2011 03:00 AM

Yes. proc glm can do it. If you want "multi-compare".

But I am not sure whether it is you find.

Ksharp

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Ksharp

08-29-2011 08:59 AM

Hi Ksharp,

I am not sure what is "multi-compare". What I want is "homogeneous subsets", which look like the last table in the webpage:

http://webpub.allegheny.edu/dept/psych/SPSS/SPSS1bANOVA.html

The reason I ask for this kind of output is that I encountered the situation like this: I have five groups of people (A, B, C, D, and E). B differ from every other group. A does not differ from C, D, and E. However, C differ from D and E, and D do not differ from E. I have lots of analysis results like this, and found it extremely difficult to report them. Therefore, if I can use homogeneous subset (or other possible solutions) then that will be useful.

To art297: I will try your method and see if it works! Thanks!

Chester

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to PsycResearcher

08-29-2011 11:31 PM

Yes. If it were ANOVA. The more information please check the documentation about proc glm.

proc glm;

class A B;

model Y=A B A*B;

means A B / tukey;

means A*B;

run;

Ksharp