Guidelines for design and measurement of interface consistency on the World Wide Web
Abstract
Previous studies have indicated that when interfaces are designed consistently with regards to structure and physical attributes, higher performance and lower error rates are achieved than when interfaces are designed inconsistently. The objective of the current study was to develop a methodology to measure all aspects of computer interface consistency and assess the impact of linguistic inconsistency of interface design on user performance. Based on the background literature, seven factors were identified as affecting overall consistency. Based on this identification, a structured questionnaire of 125 items was developed and a factor analysis was conducted which reduced the number of items in the questionnaire to 94 and identified the following nine factors which contribute to consistency: Text Structure, General Text Features, Text Content, Information Representation, Lexical Categories, Meaning, User Knowledge, Physical Attributes, and Communicational Attributes. A series of four experiments were conducted with 140 subjects. The internal reliability of the questionnaire was 0.81, and the interrater reliability was 0.75. The instrument effectively identified all of the inconsistencies in interface designs. The experimental results indicated the following: Both direct manipulation and text-based interfaces with linguistic consistency reduced performance time and error rates in comparison to inconsistent linguistics interfaces. Direct and perceptual representations resulted in higher linguistic consistency levels of interfaces than episodic and abstract representations. The instrument can be utilized both as an evaluation and design tool for Web-based interfaces. The 9-factor, 94-item questionnaire can be used to evaluate consistency of interfaces. Based on the data obtained from 20 tasks, 25 guidelines have been established which can be utilized to design consistent interfaces.
Degree
Ph.D.
Advisors
Salvendy, Purdue University.
Subject Area
Industrial engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.