topic Re: relative risk for each level of an ordinal variable in Statistical Procedures

relative risk for each level of an ordinal variable

Melk — Wed, 15 Nov 2017 00:14:16 GMT

I am interested in determining the relative risk of mortality for each level of an independent variable (8 levels), with the lowest being the reference level. What model would be appropriate to calculate this?

Re: relative risk for each level of an ordinal variable

Reeza — Wed, 15 Nov 2017 00:16:54 GMT

PROC LOGISTIC or a PROC FREQ should give you what you need.

FREQ has examples in the documentation for relative risk.

@Melk wrote:

I am interested in determining the relative risk of mortality for each level of an independent variable (8 levels), with the lowest being the reference level. What model would be appropriate to calculate this?

Re: relative risk for each level of an ordinal variable

Melk — Wed, 15 Nov 2017 00:18:05 GMT

I think relrisk does not work for rxc tables. I am not sure what the denominator component of the relative risk calculation should be for a 2x8 table.

Re: relative risk for each level of an ordinal variable

StatDave — Wed, 15 Nov 2017 15:42:22 GMT

See this note.