next up previous
Next: Design Up: Related Work Previous: History of Windowing Systems

Psychological Studies

A number of psychological studies examine aspects of organization, access and manipulation of information. Studies typically model the processes behind these tasks and how they related to users' cognitive capabilities, interaction styles and system characteristics.

Malone observes that for users of different professions the cognitive demands of categorization determines the organization style [5]. According to his observations, people with more procedural jobs have more rigid organization of documents. People with more flexible jobs, however, typically postpone categorization, if not avoid totally, due to the difficulty of finding an appropriate category. Malone's observed that information typically falls into several overlapping and fuzzy categories and a rigid categorization reflects only certain aspects of the information. Lansdale argues that the problem of categorization stems from the fact that the human mind is sensitive to the meaning of information, but not to the details of how it is communicated [7]. Thus, lack of contextual cues in categorization contributes to the problem.

Studies on memory can provide an insight into how users find information in an organization. Memory is typically divided into three components: short- and long-term, and working memory. Short-term memory deals with perception and has limited capacity. According to Miller [63], people can rapidly recognize about seven chunks of information and can hold them in the short-term memory for 15 to 30 seconds, where the size of chunks depends on the familiarity with the information. Working memory is used in conjunction with the short-term memory for information processing and problem solving. Long-term memory, according to Norman [64], consists of organized knowledge units that structure knowledge and also procedural information necessary to control actions.

Recall from memory for large amounts of information is typically a reconstruction from the long-term memory rather than rote memorization from the short-term memory. Studies on stories by Barlett [65] show that recollection of a story is a reconstruction process of what the story was about rather than a memory of what was said. Recall for spatial information is similar. Typically it is very hard to give directions from memory, but most people can find their ways by actually taking the ride, if they have been there before. Even thinking about it most of the time reconstructs the directions.

Lansdale confirms these findings and adds that the ability to recall depends upon what people are thinking about at the time [7]. He argues that people also remember things like when the information is received, where it is put, what it looks like, and many other things. A number of studies are done to compare the effectiveness of the textual and spatial organizations for finding information. According to Dumais and Jones [66], spatial organizations are of limited use even when combined with textual attributes. This result is in particular due to the limited spatial space used in the experiment. Besides, their study examined only small number of information items. Lansdale did a similar study comparing words and icons in retrieval tasks of 18 information items and also found no significant difference. In another study, Lansdale revealed that finding with the aid of temporal attributes considerably increases the success rate. As a conclusion, Lansdale argues that people typically can not remember the exact information. Thus, recall by partial attributes is likely to perform better.

Lansdale argues that finding information is a two-step process, recall-directed search followed by recognition-based scanning. Recall-directed search is repetitive contraction of the information search space to get closer to the requested piece of information. Recognition-based scanning is done when the search space is small enough to look for each piece of information and recognize attributes of the requested information. Effectiveness of different cues is most likely to be different in these two steps.

Users typically engage in multiple activities simultaneously in their normal course of work. Miyata and Norman argues that limitations on human processing power, such as conscious resources and working memory capacity, affect planning of interaction with each activity [64]. Interruptions also suspend current activity unexpectedly and resumption may require time to concentrate. Thus, they classify activities as foregrounded, backgrounded, and suspended. They conclude that retrieval aids will help people to remind them of their tasks and their contexts. Norman et al. [67] proposed a theory of the cognitive layout of information presented in multiple windows. They categorized the modes of representing the interface between the user and the display as surface, machine, and cognitive layouts and claim that surface layouts created by designers must suggest a cognitive layout that increases visual understanding of the user.

next up previous
Next: Design Up: Related Work Previous: History of Windowing Systems

Eser Kandogan
Sun Sep 13 18:34:46 EDT 1998

Web Accessibility