Bandwidth Constraints to Using Video and Other Rich Media in Behavior Change Websites

Brian G Danaher,Stephen A Jazdzewski,Clinton R Hudson,H Garth Mckay

doi:10.2196/jmir.7.4.e49

Brian G Danaher, Stephen A Jazdzewski + Show 2 more

Open Access

https://doi.org/10.2196/jmir.7.4.e49

Copy DOI

Abstract

Background Web-based behavior change interventions often include rich media (eg, video, audio, and large graphics). The rationale for using rich media includes the need to reach users who are not inclined or able to use text-based website content, encouragement of program engagement, and following the precedent set by news and sports websites.Objectives We describe the development of a bandwidth usage index, which seeks to provide a practical method to gauge the extent to which websites can successfully be used within different Internet access scenarios (eg, dial-up and broadband).Methods We conducted three studies to measure bandwidth consumption. In Study 1, we measured the bandwidth usage index for three video-rich websites (for smoking cessation, for caregivers, and for improving eldercare by family members). We then estimated the number of concurrent users that could be accommodated by each website under various Internet access scenarios. In Study 2, we sought to validate our estimated threshold number of concurrent users by testing the video-rich smoking cessation website with different numbers of concurrent users. In Study 3, we calculated the bandwidth usage index and threshold number of concurrent users for three versions of the smoking cessation website: the video-rich version (tested in Study 1), an audio-rich version, and a Web-enabled CD-ROM version in which all media-rich content was placed on a CD-ROM on the client computer.Results In Study 1, we found that the bandwidth usage index of the video-rich websites ranged from 144 Kbps to 93 Kbps. These results indicated that dial-up modem users would not achieve a “good user experience” with any of the three rich media websites. Results for Study 2 confirmed that usability was compromised when the estimated threshold number of concurrent users was exceeded. Results for Study 3 indicated that changing a website from video- to audio-rich content reduced the bandwidth requirement by almost 50%, but it remained too large to allow satisfactory use in dial-up modem scenarios. The Web-enabled CD-ROM reduced bandwidth requirements such that even a dial-up modem user could have a good user experience with the rich media content.Conclusions We conclude that the bandwidth usage index represents a practical tool that can help developers and researchers to measure the bandwidth requirements of their websites as well as to evaluate the feasibility of certain website designs in terms of specific use cases. These findings are discussed in terms of reaching different groups of users as well accommodating the intended number of concurrent users. We also discuss the promising option of using Web-enabled CD-ROMs to deliver rich media content to users with dial-up Internet access. We introduce a number of researchable themes for improving our ability to develop Web-based behavior change interventions that can better deliver what they promise.

Highlights

The bandwidth usage index (BUI) is based upon a protocol that measures the number of bits of data received by the end user when accessing a Web-based program, in the following manner: (1) accesses a representative mixture of program content including text, graphics, and rich media; (2) produces a “good experience” as measured by the user being able to access intended content within a satisfactory response time; and (3) occurs over a designated test period
We used a HP ProCurve Switch 4104gl Ethernet switch with software counters that could be set to measure the number of bits received from the hosting computer by the test computer
This noise, or overhead, is important to acknowledge since it serves as a reminder that, in real world scenarios, BUI would compete with other network applications and activity for a finite amount of bandwidth

Summary

Introduction

The bandwidth usage index (BUI) is based upon a protocol that measures the number of bits of data received by the end user when accessing a Web-based program, in the following manner: (1) accesses a representative mixture of program content including text, graphics, and rich media; (2) produces a “good experience” as measured by the user being able to access intended content within a satisfactory response time; and (3) occurs over a designated test period. We used a HP ProCurve Switch 4104gl Ethernet switch with software counters that could be set (and reset) to measure the number of bits received from the hosting computer by the test computer. We used these counters to measure the noise or protocol overhead that resulted in measurable bit traffic being received by the test computer that was not attributable to the hosting of the Web-based program.

Results

Conclusion