An approach to training based upon principled task decomposition

Abstract The primary purposes of the present research are (1) to specify a principled basis for analyzing the skill and knowledge components of expert performance within a domain. (2) to develop an instructional strategy based upon this decomposition, and (3) to study the knowledge and skills resulting from such instruction. The domain of application, the Space Fortress game (Mane et al. 1984; Mane 1985), is a complex task which involves the concurrent and coordinate use of perceptual and motor skills, conceptual and strategic knowledge, in the service of multiple goals. There are three primary aspects to our approach. The first is a decomposition of the task domain from the perspective of the inherent structure of the task, its human information processing demands, and the characteristics of expert performance. The decomposition identifies the top level goals of experts and the strategies, skills, and knowledge developed by them in pursuit of those goals. The second aspect of our approach is an analysis of hierarchical relations among the skills and knowledge components to be acquired. Hierarchical relations derive from conditional and functional relations among the components. This process attempts to determine a possible set of transformations that will turn novices into experts. As part of this process, skills may be identified that are not present in expert performance but which are necessary precursors to the acquisition of some skill or knowledge that is a component of expert performance. The third aspect of our approach is a construction of activities for training the individual components and their integration based upon a set of design principles, and an evaluation of the nature of the skills and knowledge acquired. In contrast to the training strategies pursued by other investigators within the Learning Strategies project, our training activities are not necessarily set in the context of the Space Fortress game. This allowed us to constrain task performance in such a way as to necessitate the development of particular concepts needed for expert performance. In addition, our training strategy focused, more than any of the other training strategies, on the cognitive aspects of expertise in the game. In particular, we were concerned with facilitating the development of the conceptual knowledge that our analyses indicated is necessary for expert ship control. Our experimental results indicate the subjects exposed to our series of training tasks not only performed better than control subjects on the Space Fortress game but also performed better on other dynamic control tasks as well as on a paper and pencil physics test that measured their understanding of how accelerations affect the motion of objects.

[1]  Walter Schneider,et al.  Degree of Consistent Training and the Development of Automatic Processing. , 1980 .

[2]  Barbara Y. White,et al.  Intelligent Tutoring Systems Based Upon Qualitative Model Evolutions , 1986, AAAI.

[3]  Emanuel Donchin,et al.  Unsupervised Practice: The Performance of the Control Group , 1989 .

[4]  Emanuel Donchin,et al.  Working memory in the acquisition of complex cognitive skills , 1989 .

[5]  W. Schneider,et al.  Automatic category search and its transfer. , 1984, Journal of experimental psychology. Learning, memory, and cognition.

[6]  E. Donchin,et al.  The space fortress game , 1989 .

[7]  K L Shapiro,et al.  Training of efficient oculomotor strategies enhances skill acquisition. , 1989, Acta psychologica.

[8]  D. Gopher,et al.  Practice under changing priorities: An approach to the training of complex skills☆ , 1989 .

[9]  John R. Frederiksen,et al.  A Componential Approach to Training Reading Skills: Part 1. Perceptual Units Training , 1985 .

[10]  John R. Frederiksen,et al.  A Componential Approach to Training Reading Skills , 1985 .

[11]  Barbara Y. White,et al.  Progressions of Qualitative Models as a Foundation for Intelligent Learning Environments. Report No. 6277. , 1986 .

[12]  Walter Schneider,et al.  Training High-Performance Skills: Fallacies and Guidelines , 1985 .

[13]  P. Rabbitt,et al.  Space fortress as an IQ test? Predictions of learning and of practised performance in a complex interactive video-game , 1989 .

[14]  D. Gentner.  Mental Models , 1983 .

[15]  A. D. Fisk,et al.  Degree of consistent training: Improvements in search performance and automatic process development , 1982, Perception & psychophysics.

[16]  J. Adams.  Historical review and appraisal of research on the learning, retention, and transfer of human motor skills. , 1987 .

[17]  G E BRIGGS,et al.  The relative efficiency of several training methods as a function of transfer task complexity. , 1962, Journal of experimental psychology.

[18]  Gavan Lintern,et al.  Part-Task Training for Tracking and Manual Control , 1985 .

[19]  Karl M. Newell,et al.  Whole-part training strategies for learning the response dynamics of microprocessor driven simulators☆ , 1989 .

[20]  Barbara Y. White.  Designing Computer Games to Facilitate Learning , 1981 .

[21]  Emanuel Donchin,et al.  The Design and Use of Subtasks in Part Training and Their Relationship to the Whole Task , 1984 .

[22]  Emanuel Donchin,et al.  Adaptive and part-whole training in the acquisition of a complex perceptual-motor skill , 1989 .

[23]  G E BRIGGS,et al.  Effects of task complexity and task organization on the relative efficiency of part and whole training methods. , 1963, Journal of experimental psychology.