O'Hara K P, Payne S J
University of Wales, Cardiff, United Kingdom.
Cogn Psychol. 1998 Feb;35(1):34-70. doi: 10.1006/cogp.1997.0676.
This paper explores the idea that problem solving search strategies are chosen so as to optimize performance within the constraints of a particular situation. Four experiments are reported that examine the hypothesis that the cost of performing an operation affects "planfulness"--the level of planning during problem solving. The first experiment investigated problem solving with the 8-puzzle and compared strategies adopted when there was a high versus a low cost of making a move, manipulating cost in terms of command length. The second experiment used protocol analysis to provide more direct evidence for increased planning. The third and fourth experiments looked at the effects of these different strategies on learning: the third examined how problem solving performance on a direct manipulation interface is affected by prior problem solving experience in the same domain with either a high cost or low cost command-driven interface, demonstrating improved performance as a result of training on an interface with high cost operations; the fourth experiment, like the third, examined the effects of prior problem solving experience with either a high cost or low cost interface on subsequent problem solving performance in a different domain using a direct manipulation interface showing no effect to training interface on subsequent performance.
本文探讨了这样一种观点，即问题解决搜索策略的选择是为了在特定情况的限制下优化性能。报告了四项实验，这些实验检验了这样一种假设，即执行一项操作的成本会影响“计划性”——问题解决过程中的计划水平。第一个实验研究了使用八数码难题进行问题解决，并比较了在移动成本高与低的情况下所采用的策略，通过指令长度来操纵成本。第二个实验使用协议分析来为增加的计划性提供更直接的证据。第三和第四个实验研究了这些不同策略对学习的影响：第三个实验考察了在同一领域中，使用高成本或低成本命令驱动界面的先前问题解决经验如何影响直接操作界面上的问题解决性能，结果表明在高成本操作界面上进行训练会提高性能；第四个实验与第三个实验类似，考察了使用高成本或低成本界面的先前问题解决经验对在不同领域中使用直接操作界面进行后续问题解决性能的影响，结果表明训练界面对后续性能没有影响。
