Can Graphs Mislead Decision Makers When Formulated in Violation of Tufte's Lie Factor?
Abstract
This thesis investigated if decision makers could be mislead by computer generated positive and negative trend graphs formulated in violation to Tufte's lie factor. It also investigated if the misleading influence of graphs increases or decreases as the lie factor increases or decreases. A literature review revealed the need to test the lie factor in a controlled environment. By creating graphs with lie factors, stronger impressions may be produced on the viewer which may lead to misinterpretations of the underlying data. Experiments using a fully automated computer program were conducted on 68 decision makers. The experiment contained a control group and an experimental group. The control group received a program containing graphs constructed in accordance to Tufte's lie factor. The experimental group received a program containing graphs derived from identical data as that of the control group, but constructed as to violate Tufte's lie factor. Results showed that stronger impressions are created by graphs which violated the factor. It could not be concluded that as the lie factor increases, the misleading influence also increases.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1991
- Accession Number
- ADA243944
Entities
People
- Charles W. Kern Iii
Organizations
- Air Force Institute of Technology