topic Re: Regarding Mutilpe Response Analysis in SAS Programming

Re: Regarding Mutilpe Response Analysis

Reeza — Fri, 04 Jan 2013 17:04:32 GMT

I typically do something like the following. But as BallardW has indicated, your results don't match your question.

data have;

input Slno Q1_1 Q1_2 Q1_3;

cards;

1 1 0 1

2 0 . 0

3 1 1 0

4 1 . 0

5 1 1 1

6 0 . 1

7 0 1 1

8 1 0 1

9 1 . 1

10 1 0 0

;

run;

data step1;

set have;

array ques(3) q1_1 q1_2 q1_3;

do i=1 to dim(ques);

response=ques(i);

question=vname(ques(i));

output;

end;

run;

proc means data=step1 n sum mean;

class question;

var response;

run;

proc freq data=step1;

table question*response/list missing;

run;

tadiboinak — Fri, 04 Jan 2013 17:23:32 GMT

hi Reeza,

The solution array given by you worked perfectly but unfortunately is not the solution as the observations are increasing as a result of the array,

For example i have 30 respondents. when i am using an array it increases to 90 respondents.

I am attaching a sample files for your perusal. Request you to kindly go through the file and suggest a soultion.

Thanks,

Saikiran

Reeza — Mon, 21 Jan 2013 00:05:23 GMT

It will increase to 90 observations, # of respondents *# of questions - in this case, 30 * 3= 90.

You basically have Person*Question level data instead of person level data now.