Internet Delay Effects: How Users Perceive Quality, Organization, and Ease of Use of Information

ABSTRACT

Keywords

© 1997 Copyright on this material is held by the authors.

INTRODUCTION

While the Internet is being used more frequently to provide or locate information, research has shown that users do not like to wait for information to arrive. One of the most common complaints about the Internet is that some documents take too long to retrieve. There are several causes of delays: Web servers, Web clients, network latency[1]. However, while perceived latency is a concern of the average user, the cause of the latency is generally considered irrelevant.

It is widely recognized that system response time (SRT), the time between the user's input and the computer's response, is one of the strongest stressors during human-computer interaction [2]. Assessments of the effects of SRT have been conducted for personal computer use in a variety of contexts. As early as 1982, researchers determined that SRT and SRT variability act in concert to increase the stress levels of some personal computer users [3]. With increased SRTs, users rate their general well-being as lower [4]. Also accompanying longer delays are self-reports of annoyance [5], frustration, and impatience [6].

Given the results of previous research on system response time delays and the fact that Internet users' continue to complain about response delays, we examined the effect of Internet delays on the perceived usefulness, organization, and quality of information acquired from the Internet.

METHOD

The delays were generated using a trace-driven simulation technique [7]. This technique is based on measuring real network delays, processing the resulting data, and using these results to drive an instrumented WWW server. Using this technique, documents are delayed different amounts of time based on a number of factors including document size and the WWW site being simulated. Although means and medians do not adequately describe the distribution of document delays, they do provide an estimate of the differences between the three conditions. The following table provides mean and median delays for individual subdocuments (HTML file, graphic, etc.) for the three delay conditions. The specific delays were modeled after three heavily used web sites insuring that the delays were representative of real-world conditions.

	Delay Condition
	Short	Medium	Long
Median (msec)	385	2210	3600
Mean (msec)	575	3500	6750

To compel the subjects to explore the web site in detail, they were presented with 20 tasks requiring retrieval of information at the site. Subjects were allowed an unlimited amount of time to peruse the site. Subsequent to site perusal, the subjects were required to respond to a series of questions. The questions were designed to gauge the effect of delay on the subjects' perceptions of ease of locating information, organization of the information, quality of the information, and navigation problems. The subjects responded on a Likert scale of 1 (strongly favorable) to 7 (strongly unfavorable).

RESULTS AND DISCUSSION

The results consistently demonstrate that subjects perceive text to become progressively less favorable as the delay becomes shorter. In contrast, documents incorporating text-and-graphics become more favorable the shorter the delay. The results are consistent with previous studies that showed user sensitivity to delays during personal computer use. However, this study is the first that establishes sensitivity to delays of different lengths in the context of Internet use. The results indicate that users may prefer multi-media web sites but they are unwilling to tolerate the substantial network delays that may be associated with retrieving the necessary graphics, video, animation, and audio.

REFERENCES

1. Padmanabhan, V. N. & Mogul, J. C. (1995) Improving HTTP Latency. Computer Networks and ISDN Systems, 28, 25-35.
2. Thum, M., Boucstein, W., & Kuhman, W. (1995) human-computer interaction. Ergonomics, 38, 1342-1351.
3. Weiss, S. M., Boggs, G., Lehto, M., Hodja, S., & Martin, D. J. (1982) Computer system response time and psychophysiological stress. Proceedings of the Human factors Society 26th Annual Meeting, 698-702.
4. Kuhman, W., Boucsein, W., Schaefer, F., & Alexander, J. (1987) Experimental investigation of pschophysiological stress-reactions induced by different system response times in human-computer interaction. Ergonomics, 30(6), 933-943. Planas, M. A., & Treurniet, W. C., (1988) The effects of feedback during delays in simulated teletext reception. Behaviour and Information Technology, 7(2), 183-191.
5. Schleifer, L. M., & Amick, B. C. (1989) System repines time and method of pay: Stress effects in computer-based tasks. International Journal of Human-Computer Interaction, 1(1), 23-39.
6. Sears, A., & Borella, M. S. (1997) WANDS: Tools for designing and testing distributed documents. Technical Report #97-01, School of Computer Science, DePaul University, Chicago, IL.
   

   ---

   
   CHI 97 Electronic Publications: Late-Breaking/Short Talks