How to create a null model for conditional logistic regression in R?

r goodness-of-fit
tumultous_rooster picture tumultous_rooster · Jun 15, 2014 · Viewed 9.1k times · Source

I would like to see if including some covariates will give me an AIC which is smaller than the null model, while working with conditional logistic regression in R (discrete choice).

I realize I can build a model, such as 

mymodel <- clogit(choice ~ dark+soft+nuts + strata(ID), data=candy)

and get the AIC by using 

extractAIC(mymodel)[2]

But I am a little unsure as to how I can create a null model to compare this against. Any advice would be greatly appreciated.

Answer

MrFlick picture MrFlick · Jun 15, 2014

Just fit a model without covariates

nullmodel <- clogit(choice ~ 1 + strata(ID), data=candy)

That's assuming that your null is that dark+soft+nuts have no effect on choice. Otherwise be clear about what your null and alternative hypothesis are.

Related questions

Chi-squared goodness of fit test in R

I have a vector of observed values and also a vector of values calculated with model: actual <- c(1411,439,214,100,62,38,29,64) expected <- c(1425.3,399.5,201.6,116.9,72.2,46.3,30.4,64.8) Now I'm using the Chi-squared goodness of fit test to see how well my model performs. …

Read More
Remove rows with all or some NAs (missing values) in data.frame

I'd like to remove the lines in this data frame that: a) contain NAs across all columns. Below is my example data frame. gene hsap mmul mmus rnor cfam 1 ENSG00000208234 0 NA NA NA NA 2 ENSG00000199674 0 2 2 2 2 3 ENSG00000221622 0 NA NA NA NA 4 …

Read More
Drop data frame columns by name

I have a number of columns that I would like to remove from a data frame. I know that we can delete them individually using something like: df$x <- NULL But I was hoping to do this with …

Read More