In order to determine the extent to which the LOGO programming language can be used as a reinforcer for general problem solving ability, this field study used a production theory approach to problem solving as a conceptual model, and then translated the model into a LOGO oriented framework. The methodology tested the use of an algorithm in a quasi-experimental fashion by presenting it to students, and then determining the extent to which it increased their ability to solve LOGO problems. Four subjects, one each from grades 4, 5, 6, and 8, were presented with standard LOGO problems. (All students had received at least 4 hours of instruction in using LOGO commands.) Students were then rated using a Likert scale. The results indicate that a general problem solving algorithm does not significantly change the problem solving ability of students as it relates to LOGO problems, which implies that a "stronger" algorithm approach should be developed for teaching problem solving for different types of problems. However, the results also suggest ways that LOGO problem solving techniques might generalize to more specific algorithms because of LOGO's highly visual nature, recursive features, and provision of immediate feedback. For the reinforcement of these aspects, LOGO appears to be a highly useful teaching tool. Five references are listed. (DJR)