Crude and adjusted OR

[kyla123]

I am trying to determine the Odds ratio of the association between the number of hpv shots and poverty status adjusting for age, race, and sex. we have grouped the number of shots by 0-1 and>=2 and using the poverty status Above Poverty >$75k as our reference. This is the code ive been trying to use but it gives me an error message of "invalid reference value for incpov1" and a syntax error for my event statement 

proc logistic data = puf.nisteenpuf19 descending;

class incpov1 (ref="Above Poverty >$75k") / param=ref;

model hpvi_group(event= "0-1" ">=2") = incpov1 age sex raceethk;

run;

[PaigeMiller]

What are the actual values (formatted) in your data set puf.nisteenpuf19? Did you type the ref= properly? Does it match exactly?

[Reeza · Accepted solution]

I'm surprised Sex or Race are not also categorical variables in the CLASS statement.

Please show the actual log.

REF= value should be the formatted value and must match exactly. If you're not certain, run a PROC FREQ on the INCPOV1 variable and post that output.