Download Cognitive Ergonomics. Understanding, Learning, and Designing by Pierre Falzon PDF

Show description

Figure 3. 6 Method graph of the experimental tasks. PROCEDURE Four groups of subjects (total η = 88) performed the four types of editing tasks mentioned above in four different task orders. Ten instances of each task type were provided, each involving the same number of operations in the same order. There was one session for each task type. Figure 4 shows the task order for each group with the associated number of new production rules to be learned in each session. A session was terminated after three subsequent error-free trials (learning criterion).

This is because the subject's response to any particular experiment is strategic, and will reflect a flexible and possibly ungeneralizable response to the particular task demands. On the other hand, when it is those strategies which are themselves of interest (for example, we might want to know the circumstances under which one filing strategy is preferred over another), then we can only observe them by creating the task in which they are observed. When we are talking about personal information management, this means everyday information management, over realistic periods of time, and over realistic quantities of information.

The action part of the rule is evaluated). A very simple but in some cases sufficient solution is to write such specific rules, that only one rule applies in a given context. Rules interact with W M by adding or deleting items. Those 'internal' actions are the only means for handling flow of control and information between different rules. In other words: rules communicate only via W M . W M contains descriptions of current goals, notes about actions that have been executed and assumptions about the current state of the 'external' world.