The Internet Encyclopedia (Volume 3)

P1: JDW

Zimmerman WL040/Bidgolio-Vol I WL040-Sample.cls June 20, 2003 17:20 Char Count= 0

Usability Testing: An Evaluation Process

for Internet Communications

Usability Testing: An Evaluation Process

for Internet Communications

Donald E. Zimmerman,Colorado State University

Carol A. Akerelrea,Colorado State University

Introduction 512

What is Usability Testing? 512

Why Test the Usability of Internet Products? 512

Foundations of Usability Testing 513

Methodological Foundations 513

Theoretical Foundations 513

Usability Methodologies 514

Card Sorting 514

Contextual Inquiry 515

Heuristic Evaluations 515

Verbal Protocol Analysis 516

Ethical and Legal Considerations 516

Avoiding Usability Pitfalls 517

Integrating Usability Into the Design Process 517

Develop Project Management Plan 517

Develop Clear Objectives 518

Analyze Audiences 518

Conduct Task Analysis 518

Develop Scenarios for Usability Testing 518

Select and Develop Measurements 518

Recruit Participants 519

Collect Data 520

Analyze Data 520

Interpret the Data 520

Identify Strengths and Weaknesses 521

Make Recommendations 521

Reporting on Usability Testing Results 521

Prepare the Report 521

Preparing the Presentation 521

Producing a Videotape 522

Conclusion 522

Glossary 522

Cross References 522

References 522

Further Reading 524

INTRODUCTION

As Internet programs have proliferated over the last de-

cade, “usability testing” has emerged as a buzzword for

organizations claiming to have developed user-centered

products. In an atmosphere of customer-focused re-

sponses, a multitude of methods called usability testing

have surfaced, some of them similar processes and others

different. Terminology varies widely, with some authors

calling the activity “usability testing,” and others referring

to it as “usability engineering,” or simply “usability.”

This chapter presents an overview of usability testing

methods, suggests when to use the methods, and lists

references for further guidance. Usability testing can be

used to assess users interactions with Web sites, Inter-

net communications, software, printed publications, and

hardware. The following discussion


provides an overview of usability methodologies,


discusses theoretical foundations for usability testing,


reviews different usability methodologies,


identifies ethical and legal considerations,


notes usability’s major pitfalls and how to avoid them,


discusses verbal protocol analysis technique in detail,

and


suggests strategies for reporting usability testing results.

WHAT IS USABILITY TESTING?

Usability testing, an evaluation research methodology,

should be a systematic process of assessing the design,

organization, presentation, and content of proposed In-

ternet products. A diverse series of methodologies, as dis-

cussed later, are used today. For our discussion, individ-

uals conducting usability testing are considered usability

practitioners, and we focus on Web sites to illustrate the

evaluation of Internet communications. The methods dis-

cussed here can be applied to other Internet products.

Usability testing methods can be used for formative

and summative evaluations as well as basic research. For-

mative evaluation assesses a Web site through its develop-

ment and provides guidance to designers, engineers, and

programmers as they continue to develop the product.

Summative evaluation assesses a Web site once finished.

Although the following discussion focuses on usability

testing of Internet communications and Web sites, keep in

mind that usability testing can be used in diverse commu-

nications and products: testing software packages; spe-

cific kinds of software, such as accounting and inventory

software; instructions and warnings; print information,

such as books, booklets, brochures, and manuals; and

equipment, such as medical equipment, video recorders,

and cameras.

WHY TEST THE USABILITY

OF INTERNET PRODUCTS?

Hard-to-use Web sites may require lengthy learning time

and waste hundreds of hours of users’ time, but too often

the costs of bad products are hidden. Consider an e-mail

program in a company with 1,000 employees. The system

and e-mails reside on a server with an inbox and multiple

512