Distribution of medical school course materials through the Web: addressing the issue of printing through an analysis of the costs and utilization

INTRODUCTION

The use of computers to distribute curricular materials has grown significantly with the advent of the Web [1–3]. Furthermore, as future clinicians, students will increasingly rely on computer medical systems for patient information and outcomes [4–6]. One consequence of network distribution of curricular materials has been the subsequent increased demand on printer resources, whether through libraries or university computer labs [7, 8], a phenomenon often referred to as the “printing problem.”

In 1993, the Web-based Loyola University Medical Education Network (LUMEN)‡ was developed to enhance the use of computer-aided instruction in the medical curriculum at the Stritch School of Medicine (SSOM) and to distribute these resources freely to the health sciences community [9]. All the basic science courses at SSOM have traditionally distributed photocopy packets of course materials at the beginning of each course. In 2000, two freshman courses, “Structure of the Human Body” (SHB) and “Introduction to the Practice of Medicine” (IPM) elected to distribute course materials through the LUMEN Web server. Advantages of providing these materials through networks have been articulated [10, 11]. These included (1) enhanced availability to a wider audience of students, (2) ability to provide rapid updates, (3) inclusion of interactive materials, and (4) ability to evaluate utilization of the materials by analyzing server logs.

The present study was designed to obtain an objective, quantitative analysis of the costs associated with the transition to Web-based distribution of course materials. The study also sought to measure changes in the level of student access of the course materials.

METHODS AND RESULTS

The study compared data collected over a two-month period (August and September) in 1999 (the “baseline” period) and August to September 2000 (the “test” period), during which times the SHB and IPM courses were taught. The costs of preparing course materials (binders, photocopying, etc.) are provided in Table 1, which shows a 55% decline in the test period, resulting in a cost savings of over $2,200 for the course. These costs, which were budgeted through the dean's office, were not entirely eliminated, because some course materials were handed out the first day of class in the test period.

Table 1 Summary of costs for distributing syllabi and printer data for the two-month period in 1999 and 2000

The level of printing in the computer lab was obtained from the laser printers' logs (EPROM), which showed a 27% increase in number of pages printed and a 16% increase in the costs of disposable supplies in the test year following the transition to Web-based distribution of course materials (Table 1).

The degree to which students in the SHB and IPM courses contributed to the increased printing was addressed through analysis of Web server logs and a questionnaire. Server statistics for requests of all files (“hits”) and hypertext markup language (HTML) files were tabulated by time and by directory as described previously [12]. These statistics, which included only requests originating from the SSOM domain, revealed that approximately 50% of the increase in requests involved the two courses studied (Table 2).

Table 2 Data on server statistics for the two-month period in both years*

Additional data substantiating the printing activities of the students in the test period were obtained through a questionnaire returned by all 130 students in these courses. The results showed that 80% printed 0 to 10 pages per day, 17% printed 11 to 20 pages per day, and 3% printed 21 to 30 pages per day in the computer lab. No student printed more than 30 pages per day. Based on the midpoints of each range, it was calculated that students in the two courses accounted for 56,570 printed pages in the test period. It was noteworthy that this number accounted for approximately 50% of the total pages printed during the two-month period of that year (Table 1), which agreed closely with the authors' conclusions based on server statistics. The survey also sought information on the printing behaviors of the students by providing data on the time of day students printed in the computer lab. The results showed that 23% printed mostly in the morning, 35% printed in the afternoon, and 42% printed mostly in the evening and night (the computer lab was accessible 24 hours a day).

The authors were further interested in the degree to which the level of printing in the computer lab correlated with the number of server requests. Regression analysis of either monthly total “hits” or HTML requests from the SSOM domain and the monthly level of printing in the computer lab showed highly significant correlations (P < 0.005), suggesting that increased printing in the computer lab was related to increased use of course materials by the students. Either statistic (hits or HTMLs) could be used for this analysis, because there was a highly significant (P < 0.001) direct correlation between the two variables.

DISCUSSION

The results of this study showed that it was cost effective for the courses studied to make the transition to Web delivery of course materials. One explanation is that students in the baseline period received photocopied packets whether they wanted them or not, whereas in the test period students self-selected what they needed. Comparisons with other institutions may be more difficult, especially if the transition involves separate cost centers (e.g., shifting costs from departmental or course budgets to libraries). As libraries and educational programs react to the increasing demands on printing, many have imposed quotas and printing charges [13–15]. These charges typically range from five to ten cents per page. In the present study, the actual cost of 1.5 cents per page for paper and toner cartridges was significantly less than the five cents per page reported elsewhere [16].

Our study demonstrated the importance of server logs for evaluating the use of technology in the curriculum. The significant correlation between server requests and printing activities provided an important index to glean information about the sources of printing. The conclusion that 50% of the printing was attributed to students in the two courses was substantiated by data collected from the survey. It was also encouraging to demonstrate a highly significant correlation between level of printing and server requests (either hits or HTMLs), suggesting that most of the printing in the computer lab was related to the curriculum.

An important consideration in addressing the “printing problem” was the degree to which printer resources in the computer lab were affected by the change to Web-based delivery of course materials. One obvious step is to determine if printing demand exceeds printer capacity, but this information would not take into account the behaviors of the users. For example, if all students decided to print at once, printer capacity could be exceeded, but only temporarily. Our survey, showing that printing occurred rather evenly throughout the day, with twenty-four hour a day access, led us to conclude that printer capacities were not a limiting factor (i.e., students could have printed more, if needed). The study focused on printing in the computer lab and did not address student printing elsewhere (e.g., home, library, research labs, etc.). However, a follow-up survey indicated that 80% of respondents did not print course materials from home (another 16% printed less than 5 pages per day). Because students were not charged for printing in the computer lab, these results confirmed our expectations that most students would not print course materials outside the computer lab in significant amounts.

CONCLUSION

Our study demonstrated that distribution of course materials via the Web resulted in a significant cost savings compared to the more traditional method of distributing photocopied packets. Therefore, cost savings could be added to the list of advantages of providing course materials through the Web. The study also showed how Web statistics could be used as a reliable tool along with student surveys to evaluate the level of use and printing of course materials.