Metrics For Designing Web Pages

Citation:

Surendra N. Singh, Nikunj Dalal, and Sanjay Mishra (2002) ,"Metrics For Designing Web Pages", in AP - Asia Pacific Advances in Consumer Research Volume 5, eds. Ramizwick and Tu Ping, Valdosta, GA : Association for Consumer Research, Pages: 313-314.

Asia Pacific Advances in Consumer Research Volume 5, 2002      Pages 313-314

METRICS FOR DESIGNING WEB PAGES

Surendra N. Singh, University of Kansas, U.S.A.

Nikunj Dalal, Oklahoma State University, U.S.A.

Sanjay Mishra, University of Kansas, U.S.A.

EXTENDED ABSTRACT -

Motivation

In the current age of rapidly proliferating business-to-consumer electronic commerce, companies are rushing to build "sticky" Web sites to retain visitors who could be at their competition’s doorstep in a mouse-click. Their rushed efforts have led to complex, user-hostile Web sites, increasing site development costs. To a large extent, the efforts of Web site designers have been stymied by the lack of theories explaining how users perceive Web pages in their ordinary, everyday Web browsing. Traditional methods used by designers and usability engineers to incorporate user input in the Web page design and testing process tend to be ad hoc, atheoretical, and focused on molecular page attributes rather than the totality of the user experience. Moreover, there is little input form marketers in the design and testing process.

Conceptualization

Contemporary methods of seeking user input in Web page design and testing are firmly rooted in usability testing and focused on utility, usability, acceptability, and costs. Furthermore, user interface design guidelines help usability inspection and design by professionals; they are not meant to be the primary means of incorporating user input in the design. When such guidelines are used in Web page design and testing they are excessively focused on efficiency and functionality issues (e.g., Do the animations load quickly? Are the fonts appropriate? Are icons intuitive?).

Based on studies in social psychology and marketing (e.g., Abelson, Kinder, Peters and Fiske 1982; Batra and Ray 1986; Edell and Burke 1987; Holbrook and Batra 1987), we propose a general model of how users might experience Web pages. The model posits that exposure to a Web page generates feelings (affective responses) as well as evaluative judgments that directly influence audiencesCattitude toward the Web page (A_wp). In addition, feelings and evaluations have an indirect effect by serving as causal antecedents to A_wp, which mediates their effect on behavioral. We develop measures for web-surfers reactions to Web pages. In our study, we demonstrate the validity of this model and measures based on it. The proposed model provides an empirically validated, theory-based approach for incorporating user input in Web page design as well as for assessing the effectiveness of Web pages on a pre- and post-test basis.

Method

We tested our model in a computer lab where student subjects were exposed to four home pages on individual computers. After a brief exposure to each home page, subjects filled out a questionnaire, providing their reactions these home pages. Seventy-seven undergraduate students participated in the main study. The home pages were extracted from their actual Web sites and loaded on individual machines from which they could be accessed locally. This eliminated variations due to differences in server response times while retaining the Web environment. We arranged Web pages in four sequences by using a randomized Latin square design to neutralize any sequence effects. Upon arrival, subjects read the instructionsC that the subjects would be shown four home pages, each appearing on the screen for a short timeCon their computer screens. The subjects were asked to view each home page, as they would normally do if they were casually browsing the Web.

Measures

Forty six semantic differential items were generated from three primary sources: (1) in-depth interviews with six subjects, who were asked to visit several home pages; (2) literature in information systems, advertising, and social cognition, and (3) nineteen pretest subjects who viewed 10 home pages and provided a listing of their thoughts and feelings about each page. The 11 affective items were: amuse, playful, irritated, joyous, pleased, annoyed, cheerful, stimulated, warmhearted, drowsy, and soothed. The 35 evaluative items were: organized, meaningful, dynamic, not well-designed, unconvincing, pleasant, fun, lively, cluttered, orderly, useless, likeable, clear, unique, interesting, enticing, bad, intriguing, cool, captivating, passive, tasteful, appealing, ugly, attractive, relevant, important, simple, attention-grabbing, unusual, surprising, curiosity-arousing, credible, unrealistic, and exaggerated.

After viewing a home page, subjects were asked to remember as best as they could how they were feeling while looking at the home page and to indicate their feelings in response to the prompt, "Did this home page make you feel ____?" by marking the appropriate responses for the 11 affect items on 7-point semantic differential scales with endpoints labeled "very much so" and "not at all@Ca scale format adapted from Abelson et al (1982) and Madden, Allen, and Twible (1988). After providing their affective reactions, subjects were instructed to evaluate the home pages on the 35 semantic differential items on a 7-point scale. We measured A_wp (the global attitude toward the Web page) on a single, 7-point semantic differential item with anchors "favorable" ad "unfavorable." On a 7-point semantic differential scale with endpoints "probable" and "improbable," subjects indicated the probability that they would click on a hyperlinked word/image/icon on the home page if they encountered the page in their casual Web browsing.

Major Findings

With only 26 items measuring feelings and evaluations, our scales are parsimonious, easy to use, and relatively inexpensive. This is a major advantage, for complex testing methods tend to be more expensive and intimidating, which limits their wide usage. The results also verified the bi-dimensional nature of the feelings: the same Web page produced both positive as well as negative feelings. Unlike extant subjective, idiosyncratic methods of soliciting user input, which require users to consider micro aspects of Web pages that may be unique to one or perhaps a few Web pages/sites, the psychometrically valid scales developed in this study enable gathering user input on theoretically meaningful (molar) dimensions. In essence, the theoretical model we have developed using the marketing and allied literatures and the scales accompanying the model provide a good copy testing tool for Web pages, a tool that could serves the evaluative and diagnostic functions simultaneously.

REFERENCES

Abelson, R.P., Donald Kinder, Mark D. Peters, and Susan T. Fiske (1982), "Affective and Semantic Components in Political Person Perception," Journal of Personality and Social Psychology, 42 (April), 619-30.

Batra, Rajeev, and Michael L. Ray (1986), "Affective Responses Mediating Acceptance of Advertising," Journal of Consumer Research, 13 (September), 234-49.

Edell, Julie A. and Marian Chapman Burke (1987), "The Power of Feelings in Understanding Advertising Effects," Journal of Consumer Research, 14 (December), 421-33.

Holbrook, Morris B. and Rajeev Batra (1987), "Assessing the Role of Emotions as Mediators of Consumer Responses to Advertising," Journal of Consumer Research, 14 (December), 404-20.

Madden, Thomas J., Chris T. Allen, and Jacquelyn L. Twible (1988), "Attitude Toward the Ad: An Assessment of Diverse Measurement Indices Under Different Processing 'Sets’," Journal of Marketing Research, 25 (August), 242-52.

----------------------------------------