I use `svyglm`

to conduct the logistic regression of a complex survey. I am wondering if this function includes the check of linearity assumption. Is the check of linearity assumption necessary for logistic regression?

`regTermTest`

does this.

Provides Wald test and working likelihood ratio (Rao-Scott) test of the hypothesis that all coefficients associated with a particular regression term are zero (or have some other specified values).

“Linearity” in this context means that coefficients of the regression are not of quadratic or higher order. \beta_0 +\beta_1x rather than \beta_0 +\beta_1x^2.

Thanks for your advice! I checked this function, but I am wondering how it works to check the linearity assumption as mentioned. It seems to be designed for model comparison and feature selection.

Linearity is assessed on a model. Were you expecting it to be assessed on data?

Yes. I found that one of the logistic regression assumptions is " The explanatory variables and the Logit of response variable have a linear relationship between them." I am wondering how to check this in survey package. I didn't find relevant information, so I don't know if this step is necessary.

`regTermTest`

does this.

Provides Wald test and working likelihood ratio (Rao-Scott) test of the hypothesis that all coefficients associated with a particular regression term are zero (or have some other specified values).

“Linearity” in this context means that coefficients of the regression are not of quadratic or higher order. β0+β1x rather than β0+β1x2 .

Can it access the data?

The `regTermTest`

accesses the model, which in turn accesses the data. Linearity is a relationship among coefficients of a model of data, not of data points themselves.

I see. Thanks for your advice. I'll try this.

This topic was automatically closed 7 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.