Describes a study focusing on different presentation methods (pairs of students; entire class; and noncomputer list game-type setting) of a computer-based science simulation to sixth-grade students to determine the most beneficial method for student achievement. Results indicate large group instruction is slightly more beneficial. (MBR)

Criticizes the statistical methods used in a previously published article that compared a commercially produced computer program with traditional teaching methods. Measures of instructional effectiveness are described, the assignment of student pairs is examined, and the use of blocking variables with analysis of variance is discussed. (LRW)