Learning strategies and exploratory behavior of interactive computer users

The cognitive modeling and laboratory studies revealed that the strategies effective for task-oriented exploration included label-following supplemented by active search for appropriate labels, along with a dual-search strategy that delves slowly deeper into both the actual interface and the users' associative memory.

Abstract

Users may engage in to investigate the capabilities of a new interface. In this form of behavior, the user's short-term goals are not clearly defined, even if there is a general long-term goal to learn the software. Related exploratory behavior may occur when real tasks are involved that are not well specified, or within a system that is not entirely understood. It has been suggested that exploratory learning is effective and attractive to users, but there has been little investigation of its occurrence outside of laboratory and training situations. The research described in this dissertation begins by examining the task of exploration in an abstract sense. The difficulties revealed by this formal approach help to guide a field study into the behavior and attitudes of computer users in everyday working situations. To provide further detail on the behavior found in the field studies, several cognitive models are implemented that describe how the user's general task description and the computer display interact to form interface-specific goals. Empirical work in the laboratory investigates the predictions of the formal theory and the model. The formal theory shows that truly task-free exploration is exceptionally difficult. The majority of users studied in the field were found to avoid this kind of behavior, preferring task-oriented exploration supported by manuals and personal interactions. The cognitive modeling and laboratory studies revealed that the strategies effective for task-oriented exploration included label-following supplemented by active search for appropriate labels, along with a dual-search strategy that delves slowly deeper into both the actual interface and the users' associative memory.