Difficulty re-leveling variables in multivariable GLM

General
1

I'm doing a multivariable analysis using Poisson regression. Below is my code for the code.

HighBPGLM <- glm(formula = dta$HBP ~ dta$SmokingLVL + dta$alcohol + dta$hiv_status + dta$diabetes, family = poisson(link = log), data = dta, na.action = na.omit)

I re-leveled the smoking LVL to be "never" "current" "past"

Thus, the reference for the analysis should be never. However, when I run the regression, my output is (the exponentiated GLM summary):

FYI "...." represents the estimate

(Intercept) .....

dta$alcohol .....

dta$hiv_status .....

dta$diabetes .....

As you can see, I am missing the estimate and the levels for smokingLVL.

What did I do wrong? and how can I fix it to be

(Intercept) .....

dta$smokingLVL1 ....

dta$smokingLVL2 ....

dta$alcohol .....

dta$hiv_status .....

dta$diabetes .....

2

In your glm call, you set the data to come from plos.dta but your formula uses data from dta. Did you relevel in plos.dta? Try

HighBPGLM <- glm(formula = HBP ~ SmokingLVL + alcohol + hiv_status + diabetes, 
                  family = poisson(link = log), data = plos.dta, na.action = na.omit)
3

I apologize. That was an error from creating dummy variables. It's edited now to dta. Thanks.

4

This topic was automatically closed 42 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.