Reaction times analysis with Mixed Models


I want to use Mixed Models for interpreting reaction times, instead of ANOVA repeated measures with the mean for each subject/condition.

I have several subjects, condition (2 levels) and time delay (6 levels). Each subject obtained 20 reactions times (RTs) for each time delay in the two condition.

Could someone explain to me how to analize the data using Mixed Model, thus using each reaction times for the analysis, and not only the mean per condition as we usually do with classical ANOVA?

I both SPSS or Jamovi, not programming software like R beacuase I have no experience with it.


Less is more. Stay pure. Stay poor.
It is usually pretty straight forward. Things to consider that deviate the approach is time vary confounding or cross-level interactions, otherwise you select your outcome and covariates like normal regression and list your clustering (grouping variable, e.g., persons ID). Next you also list whether any of the terms (covariates) are at the individual grouping level. I haven't used SPSS for this, but I bet they have some resources. Multilevel models are definitely preferred over repeated measures ANOVA.
I would guess that reaction times would be a skewed distribution. So maybe a gamma distribution fits. Or maybe a lognormal distribution. Maybe they don't have that on the standard software. Then try to take the logarithm (so that data transforms from the lognormal to the normal distribution).


Less is more. Stay pure. Stay poor.
Good point @GretaGarbo about still needing to meet model assumptions. Side note, when you recommended the gamma distribution to me last month, I searched around on the topic it seemed like quite a lot of people said they preferred lognormal distribution over it. I believe it was since they both get you to about the same place, but lognormal is more interpretable - I think that is a general comment I saw. I am sure this is one of those if I could recall how all of the distributions are related, I could better articulate and provide insight.