topic Re: PROC CORR? in SAS Procedures

PROC CORR?

robertrao — Mon, 14 Oct 2024 03:58:09 GMT

Hello,

I have the data in this format shown below and i have given examples for 2 ID's.

Each domain may have several questions and each ID may be either Established or NEW.

How do i find if there is a correlation of scores between Established and NEW.

I tried the following:

/*FOR ESTABLISHED*/
proc transpose data=Have(where=(Appointment_Type="ESTABLISHED")) out=test_transpose_Established prefix=EST_;
by ID question;
var score;
id DOMAIN;
run;

/*NEW PATIENTS*/
proc transpose data=Have(where=(Appointment_Type="NEW")) out=test_transpose_NEW prefix=NEW_;
by id question;
var score;
id Domain;
run;

/*Both Together*/

data new_EST;
set test_transpose_Established test_transpose_NEW;
run;

proc corr data=NEW_EST;
run;

ID DOMAIN QUESTION SCORE APPOINTMENT_TYPE

101 Med Practice: Care Provider Med Practice: CP Concern 100 Estblished

101 Med Practice: Care Provider Med Practice: CP Explanation 100 Estblished

101 Med Practice: Care Provider Med Practice: CP Discuss treatments 100 Estblished

101 Med Practice: Care Provider Med Practice: CP Efforts 100 Estblished

101 Med Practice: Access Med Practice: Ease of contacting 100 Estblished

101 Med Practice: Access Med Practice: Ease of scheduling 100 Estblished

101 Med Practice: Access Med Practice: CP Efforts 100 Estblished

102 Med Practice: Care Provider Med Practice: CP Concern 100 New

102 Med Practice: Care Provider Med Practice: CP Explanation 100 New

102 Med Practice: Care Provider Med Practice: CP Discuss treatments 100 New

102 Med Practice: Care Provider Med Practice: CP Efforts 100 New

102 Med Practice: Access Med Practice: Ease of contacting 100 New

102 Med Practice: Access Med Practice: Ease of scheduling 100 New

102 Med Practice: Access Med Practice: CP Efforts 100 New

Re: PROC CORR?

MarkusWeick — Mon, 14 Oct 2024 05:25:07 GMT

Hi @robertrao, would "are the average scores of the new IDs significantly different from the average scores of the established IDs?" be an alternative to "what is the correlation between average scores of the new IDs and their appointment type?" for you?

Best Regards

Markus

Re: PROC CORR?

robertrao — Mon, 14 Oct 2024 06:30:11 GMT

Thank you, Marcus

Did you mean i have to perform a ttest to look at the difference in Mean scores?

Would you be able to show th sample SAS code for my example to

Proc ttest;

class Appointment_Type;

var score;

run;

Re: PROC CORR?

MarkusWeick — Mon, 14 Oct 2024 07:22:33 GMT

Starting with yout table new_EST I would calculate the average score per ID and then start the ttest:

proc sql;
create table TTest_2 as
select
ID, APPOINTMENT_TYPE, avg(SCORE) as AVERAGE_SCORE
from new_EST
group by ID;
quit;

proc ttest data=TTest_2;
   class APPOINTMENT_TYPE;
   var AVERAGE_SCORE;
run;

Re: PROC CORR?

Ksharp — Mon, 14 Oct 2024 07:59:52 GMT

Check " intraclass correlation measures":
https://support.sas.com/kb/25/031.html

Re: PROC CORR?

robertrao — Mon, 14 Oct 2024 15:55:50 GMT

Hello MArkus,

It looks like they already know that the mean scores are different . This analysis provided information that the means are statistically significant.

The established had given a higher rating than the new patients.

Additionally, the important question that we need answer to is:

Between the two groups , is there a statistical difference between the questions they answered? !

Was Med Practice: Ease of scheduling higher in one group than the other???

Re: PROC CORR?

MarkusWeick — Mon, 14 Oct 2024 18:33:18 GMT

@robertrao , could you please provide the full text of your task.

Best

Markus

Re: PROC CORR?

robertrao — Mon, 14 Oct 2024 18:49:00 GMT

Hello MArkus,

I did not post the entire data since its huge and so i pasted a sample.

It was identified the Established category gave high rtings for the questions compared to NEW patients.
I changed the original sample to reflect those now -see below as i couldnt change the original post).

Our goal is to find if there is a statistical diffference if any for a question between Established and NEW!!!

ID DOMAIN QUESTION SCORE APPOINTMENT_TYPE

101 Med Practice: Care Provider Med Practice: CP Concern 100 Estblished

101 Med Practice: Care Provider Med Practice: CP Explanation 100 Estblished

101 Med Practice: Care Provider Med Practice: CP Discuss treatments 100 Estblished

101 Med Practice: Care Provider Med Practice: CP Efforts 100 Estblished

101 Med Practice: Access Med Practice: Ease of contacting 100 Estblished

101 Med Practice: Access Med Practice: Ease of scheduling 100 Estblished

101 Med Practice: Access Med Practice: CP Efforts 100 Estblished

102 Med Practice: Care Provider Med Practice: CP Concern 96 New

102 Med Practice: Care Provider Med Practice: CP Explanation 82 New

102 Med Practice: Care Provider Med Practice: CP Discuss treatments 75 New

102 Med Practice: Care Provider Med Practice: CP Efforts 70 New

102 Med Practice: Access Med Practice: Ease of contacting 100 New

102 Med Practice: Access Med Practice: Ease of scheduling 95 New

102 Med Practice: Access Med Practice: CP Efforts 99 New