APSR is back to publishing garbage can regression results