A note on the equivalence of the graded response model and the sequential model