This study was designed to investigate the effects of modeling or training with and without rule provision on the employment of strategies in solving four-dimensional, discrimination-learning problems. Subjects were 144 second and sixth-grade children from the New York City Public Schools. The blank-trial hypothesis testing paradigm was used. The children were individually pre-trained to familiarize them with the stimuli and the blank-trial methodology and then exposed to one of three 10-minute videotapes. Four problems were presented and solved by a modeling subject on each tape. Three tapes were used: one modeling focused checking, one modeling dimension checking, and a control tape. Strategies used by children fell into five categories: focused testing, dimension checking, hypothesis checking, stereotypes and unsystematic. Charts show the percentages of pupils using the various strategies after exposure to the experimental and control conditions. A number of implications are drawn from the data. Of note is the large percentage of focusing elicited among sixth graders after exposure to the focusing and control tapes compared with very little focusing elicited from second graders under any experimental conditions. (MS)