I've been trying to analyse a dataset of response times following Lo & Andrews' (2015) proposal here: https://doi.org/10.3389/fpsyg.2015.01171

They propose using a GLMM that assumes a Gamma or inverse Gaussian distribution (with an identity link function). For example:

glmer(rt.cut ~ (1|subject) + (1|targetnumber) + primetype * form * group + scale(order) + scale(rt.previous), exp, family=inverse.gaussian(link="identity"), control=glmerControl(optimizer="bobyqa", optCtrl=list(maxfun=20000))))

(I could only get this to converge by using bobyqa, increasing iterations, and simplifying the random effects structure)

I am concerned about the interpretation of coefficients when using treatment contrasts. Specifically, I get different t-values for interactions, depending on the reference level of one of the factors.

For example, these are the 3-way interactions when using one of the levels of "form" as the reference:

primeType2:formInf:groupL2  -39.094     15.266   -2.56  0.01044 *
primeType3:formInf:groupL2  -37.020     15.495   -2.39  0.01689 *

And these are the same interactions when using the other level of "form" as the reference:

primeType2:formFinite:groupL2  39.0939    17.5759   2.224   0.0261 *  
primeType3:formFinite:groupL2  37.0203    18.0101   2.056   0.0398 *

The estimates are exactly the same (as expected), but the SEs are larger in the second case.

Why does this arise and why should I choose one or the other treatment coding?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.