This doctoral thesis examines whether inductive and deductive teaching methods are appropriate for different learners. Statistical interactions between the two instructional methods and various aptitude variables were examined among 275 sixth-grade pupils. Subjects were randomly assigned to rule-example (deductive) treatments or example-rule (inductive) treatments, which consisted of eight 40-minute periods of programmed instruction in selected skills and concepts of critical thinking. Student aptitudes were tested with five aptitude tests including Raven's Progressive Matrices test. The results indicate that the deductive and inductive teaching methods were appropriate for different learners, and that Progressive Matrices scores indicate which students should receive each kind of instruction. While both teaching strategies produced similar achievement scores, the deductive method produced better student attitudes. In this manner, the study provides empirical grounds for teaching different learners differently in order to maximize their attitudes and attainment and to personalize and humanize their instruction. The thesis presents the social and historical background of aptitude-treatment interaction research; explains the rationale, experimental design; and statistical analysis of the study reports and interprets the empirical results; and recommends several approaches to subsequent research. (Author/DE)