Background: Digital health information is available on a wide variety of platforms including PC-access of the Internet, Wireless Application Protocol phones, CD-ROMs, and touch screen public kiosks. All these platforms record details of user sessions in transaction log files, and there is a growing body of research into the evaluation of this data. However, there is very little research that has examined the problems of comparing the transaction log files of kiosks and the Internet. Objectives: To provide a first step towards examining the problems of comparing the transaction log files of kiosks and the Internet. Methods: We studied two platforms: touch screen kiosks and a comparable Web site. For both of these platforms, we examined the menu structure (which affects transaction log file data), the log-file structure, and the metrics derived from log-file records. Results: We found substantial differences between the generated metrics. Conclusions: None of the metrics discussed can be regarded as an effective way of comparing the use of kiosks and Web sites. Two metrics stand out as potentially comparable and valuable: the number of user sessions per hour and user penetration of pages.
