Locating tasks in psychology and education?

Abstract
The issue is raised: How does performance on one occasion relate to performance on another occasion? This is a problem both for practitioners and researchers working in classrooms and for psychologists concerned with the relation between results obtained on laboratory tasks and everyday performances. Naturally occurring events lack the constraints of the laboratory, making it difficult to compare one event to the next or to evaluate performance across events on a standard scale. In the study reported, attempts were made to instantiate the same task in a series of classroom lessons and activities. Two “process” coding schemes were tried out as ways of evaluating performance. The strengths and weaknesses of the Stallings Five Minute Observation system and Blank's system for coding cognitive demand are discussed with respect to specifying cognitive performance across situations.