I am quite curious about details of the model. For example, the single largest contribution to real world interview process variability is interviewer (for resume screening, who screened that resume, etc.). Wouldn't it be possible to code interviewer as categorical variable and separate resume-intrinsic? effect and interviewer effect? They must have tried this, haven't they?
I am quite curious about details of the model. For example, the single largest contribution to real world interview process variability is interviewer (for resume screening, who screened that resume, etc.). Wouldn't it be possible to code interviewer as categorical variable and separate resume-intrinsic? effect and interviewer effect? They must have tried this, haven't they?