How to get pooled results for PROC PLM on imputed data

ejay0503 — Wed, 15 Nov 2017 17:46:43 GMT

Dear all,

I have two interaction terms in my model and would like to run a simple slope analysis using PROC PLM. I already got pooled results for the mail effect but am having trouble with the PROC PLM part. Is there a way to get pooled results for PROC PLM on imputed data?

Please see my code below. Please note that I used PROC SURVEYLOGISTIC because the data is weighted and the outcome is ordinal.

(proc mi...........truncated)

proc surveylogistic data=perm.19m2;
strata RAESTRAT; cluster RAEHSAMP; weight R11WTRESP;
by _imputation_;
model DV_r (order=data) = R11AGEY_B | stress
RAEDYRS h11itot self_health R11WGIWK R11JCTEN
computer pres_retire young_promo less_demand physical_eff
gender race2_b race2_his race2_o
occupation_1 occupation_2 occupation_3 reduce_hrs part_full
/link=clogit clodds expb rsq;

ods output parameterestimates = perm.19m2print;
store perm.test; run;

(proc mianalyze....truncated)

ods graphics on;
proc plm restore = perm.test;
estimate 'stress slope, R11AGEY_B = mean-sd' stress 1 stress*R11AGEY_B 2.71
RAEDYRS 1 h11itot 1 self_health 1 R11WGIWK 1 R11JCTEN 1
computer 1 pres_retire 1 young_promo 1 less_demand 1 physical_eff 1
gender 1 race2_b 1 race2_his 1 race2_o 1
occupation_1 1 occupation_2 1 occupation_3 1 reduce_hrs 1 part_full 1,

'stress slope, R11AGEY_B = mean' stress 1 stress*R11AGEY_B 2.73
RAEDYRS 1 h11itot 1 self_health 1 R11WGIWK 1 R11JCTEN 1
computer 1 pres_retire 1 young_promo 1 less_demand 1 physical_eff 1
gender 1 race2_b 1 race2_his 1 race2_o 1
occupation_1 1 occupation_2 1 occupation_3 1 reduce_hrs 1 part_full 1,

'stress slope, R11AGEY_B = mean+sd' RAEDYRS 1 stress 1 stress*R11AGEY_B 2.75
RAEDYRS 1 h11itot 1 self_health 1 R11WGIWK 1 R11JCTEN 1
computer 1 pres_retire 1 young_promo 1 less_demand 1 physical_eff 1
gender 1 race2_b 1 race2_his 1 race2_o 1
occupation_1 1 occupation_2 1 occupation_3 1 reduce_hrs 1 part_full 1 /e ;
run;
ods graphics off;

(and then how to pool these results?)

I would really appreciate your help!

Re: How to get pooled results for PROC PLM on imputed data

SAS_Rob — Thu, 16 Nov 2017 14:37:59 GMT

You could do something similar to the example below (it assumes the Proc MI step was already run).

/* Getting Started Example
Generate Data */

proc format;
value ResponseCode 1 = 'Very_Unsatisfied'
2 = 'Unsatisfied'
3 = 'Neutral'
4 = 'Satisfied'
5 = 'Very_Satisfied';
run;

proc format;
value UserCode 1 = 'New Customer'
0 = 'Renewal Customer';
run;

proc format;
value SchoolCode 1 = 'Middle School'
2 = 'High School';
run;

proc format;
value DeptCode 0 = 'Faculty'
1 = 'Admin/Guidance';
run;

data SIS_Survey;
format Response ResponseCode.;
format NewUser UserCode.;
format SchoolType SchoolCode.;
format Department DeptCode.;
do _imputation_=1 to 2;
drop j;
retain seed1 111;
retain seed2 222;
retain seed3 333;

State = 'GA';

NewUser = 1;
do School=1 to 71;