Sixty-three students in three undergraduate college classes were divided into three groups to investigate the effect of two delayed self-reporting techniques on attitudinal data toward a course and instructor. Nonparametric tests and intercorrelations of item treatment means, together with two-way analyses of variance on item clusters, indicated that the attitudes of students responding to a "then/now" format were significantly more negative than those of other students. It was concluded that delayed self-reporting (on the order of a two-week delay) would not significantly affect the data if investigators avoid use of the "then/now" question format. (Author)