Abstract
The research described represents an ongoing effort devoted to the development of systematic methodologies for the construction of expert models that can be used to assist and instruct the inexperienced computer user. Two methodologies are described: one based on sorting and clustering procedures aimed at describing the subjective characteristics of the task or target materials, and one based on performance protocols for assessing the command selection strategies of users. The methodology for assessing subjective characteristics of the target materials yielded a consistent set of 10 target clusters from experienced and novice subjects that could be mapped directly into the database of the file selection and search task. Based on preliminary data, these more homogeneous search categories were shown to be useful for constructing frequency, sequence, and plan-based models of expert command selection that will be used to aid inexperienced computer users.
Get full access to this article
View all access options for this article.
