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
- /
- Generating a list of all combinations of 12 subjec...

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
- Highlight
- Email to a Friend
- Report Inappropriate Content

05-06-2010 10:06 AM

The original data is:

Placebo

8.62

1.48

8.93

9.57

2.65

Vitamin A

0.06

1.72

2.19

7.32

7.53

I would like to test Fisher’s (sharp) null hypothesis of no treatment effect on these 12 subjects. I want the potential outcomes under the null. I need a way to generate all combination of these 12 subjects placed in 2 groups of equal size (6 each). The data has to be output in a form that I can perform a test on each assignment.

I tried PROC PLAN combination function but I only got the 12:C:6 (924 outcomes) for the treatment group. I did not know how to put the remaining outcomes into the control group. I am not familiar with macros in SAS.

Placebo

8.62

1.48

8.93

9.57

2.65

Vitamin A

0.06

1.72

2.19

7.32

7.53

I would like to test Fisher’s (sharp) null hypothesis of no treatment effect on these 12 subjects. I want the potential outcomes under the null. I need a way to generate all combination of these 12 subjects placed in 2 groups of equal size (6 each). The data has to be output in a form that I can perform a test on each assignment.

I tried PROC PLAN combination function but I only got the 12:C:6 (924 outcomes) for the treatment group. I did not know how to put the remaining outcomes into the control group. I am not familiar with macros in SAS.

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

05-06-2010 05:15 PM

If you are looking get a permutation-adjusted t-test, you can use PROC MULTTEST which will do the permutations internally for you and give you a permutation-adjusted p-value computed from the p-values of all the permutations. For example:

data a;

input Placebo VitA;

Grp="Placebo"; Y=Placebo; output;

Grp="VitA"; Y=VitA; output;

keep Grp Y;

datalines;

8.62 0.06

1.48 1.72

8.93 2.19

9.57 7.32

2.65 7.53

;

proc multtest data=a perm;

class Grp; test mean(Y);

run;

See the MULTTEST documentation for details on the computations and examples:

http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/multtest_toc.htm

data a;

input Placebo VitA;

Grp="Placebo"; Y=Placebo; output;

Grp="VitA"; Y=VitA; output;

keep Grp Y;

datalines;

8.62 0.06

1.48 1.72

8.93 2.19

9.57 7.32

2.65 7.53

;

proc multtest data=a perm;

class Grp; test mean(Y);

run;

See the MULTTEST documentation for details on the computations and examples:

http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/multtest_toc.htm

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

05-06-2010 09:48 PM

Thanks. I may ultimately have to do that as well. But I still need the output for various combinations.

Can this procedure do that as well?

Can this procedure do that as well?

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

05-07-2010 10:14 AM

Use the N= option to control the number of resamples generated (default is 20,000) and the OUTSAMP= option to create a data set containing all of the generated samples. See the MULTTEST documentation:

http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/multtest_toc.htm

http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/multtest_toc.htm