Accuracy and Completeness of Problem Solutions with Example-Solutions

摘要

This research investigated the ability of computer users, both programmers and non-programmers, to specify problem solutions in the form of example-solutions. This ability was evaluated as a function of the complexity of the processor, i.e. the degree of generalization of the user inputs, the complexity of the problem, and the complexity of the feedback-aids. The experimental task employed in this study required the specification of a logic for the formation of a naval task-force. The performance both of programmers and non-programmers decreased with increasing levels of problem-complexity and with reduced processor support. For both the groups, errors-of-commission were relatively infrequent compared to errors-of-omission. It was found that the degree of processor-complexity was much more influential than problem-complexity in predicting performance scores. When little computer generalization of user-input was provided, performance was significantly lower than during all other experimental conditions. Results also showed that participant-strategy in the generation of problem solutions was a significant factor in performance, though years-of-experience and years-of-education were not found to be good predictors of performance. The feedback-aids were shown to be most effective when they included the logic implied by the example-solutions. These experiments demonstrate the effectiveness of the on-line use of computer software to create and modify software routines. (Author)