A study examining the effects of systematized feedback on mathematics performance was conducted in natural classroom settings to extend the external validity of previous feedback studies. A posttest-only, internal-external control group design was used with three sixth grade classes serving as subjects. All classes utilized a "step" approach to math instruction, where students advanced through a pre-determined sequence of math skills at individual rates. Two classes, each taught by the same teacher, were used to measure treatment effects. One class served as an internal control group, receiving only math instruction. The other class received math instruction and systematized feedback. The remaining class, taught by a different teacher, served as an additional external control class. It was included to control for possible teacher bias and received only math instruction. At the end of 12 weeks, all students were administered a 25-item test. A comparison of the scores of the three classes indicated that feedback is of significant value when applied in a systematic manner. The study confirmed the effectiveness and practicality of feedback based programs in "real world" settings. (Author/LLS)