Bayesian workshop - STEP 2023
University of Alberta
We often log-transform our data (especially reaction times)
Why is this normally done?
Using lognormal distribution similar
Has some advantages
model_fit <- brm(..., family = lognormal(link = "identity", link_sigma = "identity"), ...)
exp(6)
[1] 403.4288
exp(7)
[1] 1096.633
exp(8)
[1] 2980.958
Open up script S4_E1_lognormal_priors.R
S4_E1_lognormal_priors.R