Compared were the effects of two experimental arithmetic treatments, called Laboratory and Textbook, upon achievement and attitude development of fourth grade children. Prior to the study, the experimenter employed task analysis procedures to develop hierarchies of skills for the four operations on whole numbers. During the instructional phase of 29 days, the 73 students in the Laboratory group performed teacher-designed instructional activities that involved extensive use of games and manipulation of physical objects by children. The 74 students in the Textbook group were assigned traditional work from an arithmetic textbook. Children with behavior problems and learning difficulties in arithmetic were identified. At the conclusion of the experimental phase, three measures of children's arithmetic achievement were obtained: computation skills, knowledge of concepts, and application of processes. Measures were also taken of children's attitudes toward arithmetic, school, reading, spelling, and science. As measured by a commercial achievement test, no significant differences between treatments were found for each of the three areas of arithmetic achievement investigated. On experimenter-designed achievement tests, significant differences between treatments were found for Laboratory treatment of concepts and Textbook treatment of computation skills at the .05 and .01 levels, respectively. (Author/CB)