Underlying dimensions of human problem solving and learning: implications for personnel selection, training, task design and expert system

Abstract

This study presents experimental evidence on the dimensions underlying the interaction of the learning process with human cognitive representation of problem solving. An experiment was conducted in three sessions. Each session consisted of performing 12 problem solving tasks and obtaining their similarity rating. The observed four-way data made up of 12 tasks × 12 tasks × 6 subjects × 3 sessions were analyzed using a multidimensional scaling model. The analysis resulted in three major dimensions across the learning process. Two of the dimensions are described as distinctions of bottom-up vs. top-down and conscious vs. subconscious reasoning. The third one is identified as a dimension inherent to the task characteristics. It is shown that the former two dimensions are affected significantly by learning and hence become more dominant in the process of understanding. The implication of the findings for personnel selection, job rotation, task design, and expert system design is discussed.