…with a newly accepted paper at Journal of Mathematical Psychology by Zeigenfuse & Lee.  The paper deals with deals with a general approach to dealing with subjects who don’t do you the kindness of actually doing your experiment after you were nice enough to give them extra credit.  Intuitively, the approach works by introducing augmenting the observed data with a latent variable for each subject.  We can infer the value of this variable using posterior simulation.  Most of the paper deals with 4 case studies: a rate problem, the Wiener decision model, signal detection theory, and the Win-Stay, Lose-Shift model of bandit problem performance.

And Matt takes the award for the first non-”Welcome to the Madlab Blog” post.  I’ve also flooded the “Papers” page with my submissions.