Login

Publications  •  Project Statistics

Glossary  •  Schools  •  Disciplines
People Search: 
   
Title/Abstract Search: 

Dissertation Information for Marilyn Tremaine

NAME:
- Marilyn Tremaine
- (Alias) Marilyn M. Mantei

DEGREE:
- Ph.D.

DISCIPLINE:
- Communication

SCHOOL:
- University of Southern California (USA) (1982)

ADVISORS:
- None

COMMITTEE MEMBERS:
- None

MPACT Status: Incomplete - Not_Inspected

Title: DISORIENTATION BEHAVIOR IN PERSON - COMPUTER INTERACTION

Abstract: Anecdotal evidence in the use of menu-driven information retrieval systems has indicated that users of these systems often feel lost. This disorientation is hypothesized to be a failure of the individual to adequately acquire and organize information about the environment for effective navigation, and the individual's recognition of this failure. Research was undertaken on a system developed at Carnegie-Mellon University's Computer Science Department (ZOG) to investigate this hypothesis. Two information structures were built in ZOG, both with different implicit organization information. An experiment was conducted asking subjects to search for given information in these structures. During the search, at random presentation times, subjects responded to a question assessing their sensation of lostness.

An analysis of the experimental results showed significant positive correlations between the subjects' decision behavior and degree of lostness and between decision times and lost measures. Recall tests indicated that subjects learned more about their local search environment when they felt lost, but had no global organization information.

Significant differences in "lostness" existed between the two information structures. To understand these underlying differences production system models of the subject's decision behavior were built and tuned to subject data. When rules were able to incorporate goal information from the local environment, lower lost scores occurred and subject decisions were less random. Subjects on the high-lost-score information structure used hill climbing decision rules and subjects on the low-lost-score information structure used means-end analysis.

Overall, the experimental results indicated that a computerized information structure could be built without maps, or where-have-I-been displays, but that exceedingly simple, easily learned implicit structures would be necessary if the users were not to become lost.

MPACT Scores for Marilyn Tremaine

A = 0
C = 4
A+C = 4
T = 0
G = 0
W = 0
TD = 0
TA = 0
calculated 2010-09-26 00:33:07

Advisors and Advisees Graph