Some remarks on nevzorov's record model