An investigation of some of the effects of '"verbal load" in achievement tests