Introduction
In 1996, the State University of New York (SUNY) Office of Educational Technology (OET) awarded the libraries at SUNY Buffalo, Binghamton, and Plattsburgh an $85,000 grant to conduct a year-long study of the feasibility of delivering library and other resources over the Internet to students in distance education programs. One of the goals of the project, addressed in this article, was to determine the technical requirements for developing and operating a Web-based distance learning support service. A variety of project data was collected to determine such matters as the optimal scanning equipment for this type of work and the optimal file formats, in terms of both efficiency and accessibility. This article summarizes the data that was collected over the course of two semesters at the University at Buffalo (UB). Another goal of the project was to determine if the Internet was the most effective means of delivering instructional materials to students for the purpose of distance learning. A second article, currently in preparation, will address that question.
Background
With OET funding, the UB Libraries operated an experimental distance learning service during the fall and spring semesters of the 1996/97 academic year. The service made use of the hypertext linking and organizational capabilities of the World Wide Web to provide integrated access to a variety of course-related materials and electronic library resources and services (for an overview of the OET project, see http://ublib.buffalo.edu/libraries/course/dlproj.htm). At UB, the project built on a prototype Web service originally developed in the fall of 1995 to support distance education courses in engineering (for more information on this phase of the project, see http://wings.buffalo.edu/publications/mcjrnl/v4n1/schiller.html) With OET funding, that service was expanded in the fall of 1996 and spring of 1997 to include distance education programs in nursing, education, and library and information science.
For the OET project, the UB Libraries designed and maintained a World Wide Web site from which students could access course-related materials. These included materials developed by the course instructors, such as lecture notes, handouts, and homework assignments, as well as course reserve readings, primarily journal articles. Access to these materials was integrated with access to a variety of electronic library resources and services, including online library catalogs, bibliographic and full-text databases, electronic journals, and online interlibrary loan and reference services. In addition, links to course-related Internet sites--suggested either by subject librarians or selected by instructors--were incorporated into the Web site, providing students with a wealth of supporting research materials.
A project log was kept for statistical purposes. The log included the following information:
The log also provided space for comments, which typically noted whether any unusual processing problems were encountered.
As much as possible, materials were formatted as HTML files. However, early in the project it became apparent that this was not going to be possible for material that was handwritten or contained charts, graphs, or equations. Instead of HTML files, these materials were scanned and put up on the Web site as GIF or JPEG files. Once the service began handling journal articles and other highly formatted documents, the Adobe Acrobat Portable Document File (PDF) format also was adopted for the project. A variety of software and hardware tools were used to process materials. Software packages included HTML Assistant Pro3 (for HTML coding); OmniPage OCR (to scan text); DeskScan II (to scan images); Adobe Acrobat Capture 1.0 (to convert files to PDF); and Adobe Photoshop 3.0 (for clean-up). Project hardware included two PC workstations (a Dell OptiPlex GXMT 5133 PC workstation with Windows 95 and a 17-inch monitor and a Dell 5133 GXMT PC with Windows 3.1 and a 15-inch monitor). The project scanner was an HP ScanJet 4c.
Day-to-day processing of materials for the Web site was performed by both salaried and hourly library staff. A reserve clerk devoted five hours a week to the effort during the fall semester. In addition, a UB School of Information and Library Studies (SILS) student was hired with grant money at the start of the fall semester to work 15 hours a week on the project, primarily to process materials for the Web site. She was joined in the spring by a second SILS student who worked on the project for course credit, spending an average of 10 hours a week over a 12-week period. Part of her time was spent assisting with the processing of materials. The students' work was supervised by a full-time librarian whose project responsibilities (representing an average of approximately five hours a week) included providing training and technical support and assisting with day-to-day processing activities, as necessary.
The project supported courses offered by a number of distance education programs at UB. These included:
Fall 1996 Statistical Data Summary
In the fall of 1996, course materials, including copyrighted reserve readings, were placed on the project Web site for nine courses. The first materials were received on July 31, 1996, and linked on August 1; the last materials submitted were received on December 12, 1996, and linked that same day. A total of 176 items, or "jobs," totaling 490 pages, was processed for the Web site during this 4-1/2-month period. The average page length per item was three pages (the longest item was 36 pages; the shortest, a single page). Total processing time was approximately 110 hours, with an average processing time per item of 37 minutes. The average time spent per page was 13 minutes. The average number of items per course was 19.5. The average time spent per course was 12 hours.
For comparison purposes, materials received and processed for the Web site were grouped into three major categories--static, dynamic, and reserve. Dynamic material consisted of homework assignments, laboratory assignments, handouts, class notes, and, to a lesser extent, corrections and hypertext links. These types of materials were categorized as dynamic because they were received on an ongoing basis throughout the semester. Static materials were so characterized because they represented unique material that was typically received once, at the start of the semester. This type of material consisted of course syllabi and course schedules. Like course syllabi and schedules, reserve readings also were typically received at the start of the semester. Reserve readings were considered a separate category of material, however, because most of this type of material consisted of copyrighted published documents that required special handling (conversion to PDF, passwording, citation verification, and posting of the verified citation, as well as a copyright warning statement, on the Web site).
Fourteen percent of the material received in the fall, representing 8% of the total volume of material processed, consisted of static material, which represented 11% of the total processing time. Eighteen percent of the material processed in the fall consisted of reserve readings. Processing of this type of material (which numbered 32 items, totaling 221 pages) took up 39% (42.8 hours) of the total processing time, representing the most time- and labor-intensive type of material the project dealt with. The bulk of the material received, 68% of all material processed in the fall semester, fell in the dynamic category.
Besides representing the largest category of material submitted in the fall (120 items, totaling 230 pages) and accounting for the greatest amount of processing time (50% of the total processing time), dynamic materials were also often "time-sensitive," thus placing further demands on the service. Homework and laboratory assignments, in particular, have a short "shelf life" and required quick turn-around times. The chart below gives information on processing and linking times for these materials:
Homework and Lab Assignments - Processing and Linking Times
| Number of Items: | 45 items (26% of total) (7 lab & 38 homework assignments) | |
| Number of Pages: | 111 pages (23% of total) | |
| Processing Time: | 24.8 hours (23% of total) | |
| Linking Times: | ||
| 1-2 days | all 7 lab assignments | |
| 1-3 days | 15 homework assignments | |
| 4-10 days | 16 homework assignments | |
| > 10 days | 6 homework assignments |
In terms of the overall timeliness of the service,
In general, the amount of time spent on each item was determined by the length of the item (number of pages) and the format in which the item was received. The exception to this appears to be materials which took more than 10 days to process for the Web site. Analysis of the 33 items that took more than 10 days to link indicates that the size of the item (i.e., number of pages) had no bearing on the extended processing time. Much of this material contained difficult-to-scan typefaces or fonts; for example, a large portion consisted of practice questions in a nursing textbook that were typeset in very small fonts. Comments made on the project log sheets indicate that at least two of these extended time items had to be redone because of the poor quality of the photocopies received.
Of the total (176) items submitted to the project in the fall, slightly more than half (91, or 51%) were received in paper format. Although the guidelines for faculty participating in the project stated that electronic submission of materials was preferred, it was decided that paper copy submissions would also be accepted. These consisted of typed as well as handwritten originals, though the majority of the materials consisted of photocopies of originals that included published documents such as journal articles, book chapters, and reports, all of which had to be scanned. All of the reserve readings were received in paper format; 43% of all dynamic materials were received on paper; and 25% of all static materials were received on paper.
Paper submissions represented 78% of the total processing time spent on the fall 1996 project; electronic submissions represented 22% of the total processing time. Of the 91 items received on paper:
Of the 85 items received in electronic format (this included materials received on disk and via e-mail, as well as links to course-related Internet sites and resources):
Thus, 59% of all materials received in electronic format was linked within 2 days, compared to only 29% of the materials received on paper.
In terms of file formats, 40 items, or 23% of all items, were processed as PDF files. This material consisted of reserve materials (31 items) and dynamic course materials (nine items, consisting of one lab assignment and eight class handouts), for a total of 241 pages, or 49% of the total volume for the fall project. All but one of these items was received in paper. The materials processed as PDF files took 50 hours to process, representing 46% of the total processing time for the fall project. The average processing time per page for PDF files was 12.5 minutes per page. In comparison, 132 items, or 75% of all items, were processed as HTML files. These materials totaled 245 pages, or 50% of the total project volume for the fall. Fifty of these items were received in paper; 82 were received in electronic format. Processing time for these materials totaled 57 hours, representing 52% of the total time spent processing materials in the fall. The average processing time per page for HTML files was more than the average time per page for PDF files--14 minutes per page for HTML versus 12.5 minutes per page for PDF. Other file formats were used to a limited extent in the fall--there was one GIF file and three JPEG files. Together these four items represented just 2% of the total number of items and only 2% (or just under two hours) of the total processing time spent on the fall project
Spring 1997 Statistical Data Summary
In the spring of 1997, materials were placed on the project Web site for three courses. The first of these were received on November 11, 1996, and linked that same day (these consisted of the schedule and syllabus for a graduate course in public school finance, and they were intentionally placed on the Web site well in advance of the spring semester to help market the course); the last materials submitted were received on February 27, 1997, and linked by March 4.
The nature of the spring project was very different from the fall project. There were fewer courses to support and a shorter project time frame (all materials had been received and linked within the first two months of the start of the spring semester). In addition, far fewer items were received although there were more pages, and one of the courses was taught entirely on the Web. This course represented a cooperative effort between the project team and the course instructor, whose teaching assistant was given the responsibility of operating a Web course page for materials such as the syllabus, handouts, and assignments. The OET project team supported these efforts by processing the course reserve readings as PDF files and placing them on the project Web site. The teaching assistant then linked to the reserve readings from the course page he maintained.
A total of 19 items was processed in the spring (compared to 176 processed in the fall) for a total of 665 pages (compared to the 490 pages processed in the fall). The average number of pages per item was 35 (compared to an average of three pages per item in the fall). The longest item received in the spring was 206 pages; the shortest was a single page. Total processing time in the spring was 35 hours--less than one-third the total processing time in the fall, although the total volume of the material received in the spring project was 1.4 times the total volume of materials received in the fall (665 pages versus 490 pages). The average processing time per item was 1.9 hours. The fact that it took longer on average to process an item in the spring than in the fall can be attributed to the greater average size (number of pages) of the items received in the spring. The average processing time spent in the spring per page was 3.2 minutes, compared to 13 minutes per page for the fall. While the average number of items per course was lower in the spring (6.3 compared to 19.5 for the fall), the average time spent per course was about equal--11.7 hours per course for the spring project versus 12 hours per course in the fall.
Twenty-six percent of the items received in the spring were static materials consisting of course syllabi and schedules. All static materials received in the spring were in electronic format. These materials represented 3% of the total processing time and 6% of the total number of pages processed in the spring. Fifty-eight percent of the items were dynamic materials, consisting mainly of handouts and quizzes. All of the dynamic materials received in the spring were in paper format. They represented only 5% of the total number of pages processed but took up 17% of the total processing time. Sixteen percent of all items received in the spring were reserve readings. Reserve materials, however, took up 80% of the total processing time and represented 89% of the total number of pages processed. All reserve materials received in the spring were received in paper format, as in the fall. In comparing the figures for the spring project and the fall, the large percentage of work time and work volume spent on reserve materials is notable. The amount of time spent on reserve materials constituted virtually all the time spent on the project in the spring.
As mentioned above, the majority of materials received in the spring were in paper format (74%, representing 97% of the total processing time). Only 26% of the material received for the spring project was in electronic format, representing 3% of the total processing time. In terms of file formats used, seven items, or 37% of all items, were processed as PDF files in the spring. This material totaled 599 pages, or 90% of the total volume for the spring project. Most of it consisted of reserve materials (594 pages). The materials processed as PDF files in the spring took 28.6 hours to process, or 82% of the total processing time for the spring. The average processing time per page for PDF files was 2.9 minutes, significantly less than the 12.5 minutes per page it took to process PDF files in the fall. In comparison, 12 items, 63% of all items submitted in the spring, were processed as HTML files. These materials totaled 66 pages, or only 10% of the total project volume. Eight of these items were received on paper; the remaining four were received in electronic format. Processing time for these materials totaled 6.4 hours, or 18% of the total time spent processing materials in the spring. As in the fall, the average processing time per page for HTML files was greater than the average time per page for PDF files--5.8 minutes per page for HTML versus 2.9 minutes per page for PDF.
In terms of the overall timeliness of the spring project:
These figures suggest that by the spring the project team had become faster at processing materials. The data also indicate that converting materials into PDF was faster and more efficient than converting materials into HTML.
Summary Figures for Both Semesters
Twelve courses, in nursing, education, engineering, and information and library science, were supported in 1996/97 by the OET-funded distance learning project at UB. A total of 195 items, totaling 1155 pages, was processed for the project Web site in the fall and spring semesters. Total processing time for the project was 144.5 hours. The average length of an item was six pages. The average time to process a page was 7.5 minutes.
Concluding Remarks
The aim of the experimental service was to create an information-rich online learning environment using the World Wide Web to deliver course-related materials and resources to students in distance education situations. A major objective of the OET grant was to find out what this entailed, with a particular focus on the technology involved.
File Formats. The project data indicated that converting materials into PDF was faster and more efficient than converting materials into HTML. In addition, PDF preserves the original page layout and design of a published document--of particular importance when the material being scanned is formatted in columns, the standard format for most journal articles, or when the material contains mathematical notation, such as equations. Moreover, unlike a GIF or JPEG file, a PDF file is not a static, or "frozen," image but a dynamic, searchable document that can be enhanced with hypertext links. A disadvantage to PDF is that a special software program (the Adobe Acrobat Reader) is required for viewing it. The Acrobat Reader, however, is available for free on the Internet and is increasingly being bundled with operating system software and loaded on campus servers along with other "helper applications." All reserve materials for the Web site were processed as PDF files. In addition, all copyrighted reserve materials were password-protected and access to them was limited to students enrolled in the course for which they were reserved. The Adobe Acrobat Capture software, used to convert files into PDF, proved well-suited to these purposes, since the software includes the ability to assign a password to a PDF file.
Scanning Equipment. One of the most critical pieces of equipment in this type of work is the scanner. All of the statistics reported in this article are based on work performed using an HP ScanJet 4c. Toward the end of the project, the UB Libraries acquired a Fujitsu M3097G High Performance Scanner, which was not used in the fall of 1996 or spring of 1997, but has been used beginning in the fall of 1997 to continue the service. Although to date no statistics are available to compare the performance of the Fujitsu to the HP ScanJet, anecdotally the difference is significant in terms of both speed of scanning and accuracy. A significant amount of time, particularly during the first semester of the project, was spent proofreading and correcting scanned materials. The need to proof and correct scanned documents has since been eliminated with the Fujitsu. The built-in option on the Fujitsu to scan directly into PDF and the automatic document feeder that was purchased with the scanner (the HP ScanJet was used without one) have also helped to speed up and streamline the process. Although we no longer use the Adobe Acrobat Capture software to convert materials to PDF, readers should note that later releases of the Capture software have improved on the version used for the OET project.
Submission Formats and Deadlines. Faculty were encouraged but not required to submit their materials in electronic format for the OET project. As the project data indicated, the processing of materials received in electronic format was much more efficient than processing of materials received in paper format. Since most of the course materials developed by faculty--both static materials, such as syllabi and schedules, and dynamic materials, such as handouts and lecture notes--are routinely prepared by them (or their assistants) in an electronic format, electronic submission of these types of materials could be made a requirement of the service. Similarly, no deadlines were established for the submission of materials, although it would not be unreasonable to set deadlines for certain categories of materials, such as reserve readings or course syllabi and schedules. However, requirements that specify electronic submission or early deadlines are not necessarily meaningful for other kinds of materials. Reserve readings, for example, unless they are supplied directly from the publisher in an electronic format or are available through a full-text database, will be received in paper format and will therefore require scanning and conversion, most likely to a special file format such as PDF. For dynamic materials, deadlines set at the beginning of the semester are not applicable. Very few faculty have all of their materials prepared in advance of the first day of class, nor would it be desirable in every case to put up all of the materials for a course at one time, particularly homework and lecture notes.
Types of Materials. Whether the aim is to develop and operate a service similar to the one described in this article in support of distance learning or an electronic reserve system (which also, of course, could serve the needs of distance learners), decisions must be made about the types of materials that will be handled. The service developed by the UB Libraries handles materials developed by the course instructors and course reserve readings. Other services may choose to handle just course materials or just reserve readings. A better understanding of the demands these different types of materials are likely to place on a service will aid in establishing workable guidelines, such as those described above under submission formats and deadlines, and will facilitate making decisions about resource allocations within the library. The project at UB indicated that reserve materials are by far the most time- and labor-intensive of all material types. Also, dynamic materials, particularly assignments, place special demands on a service of this type. Whenever possible, dated material of this kind was given priority processing, but it was difficult to predict in advance how much of this type of material might be received in any given week or how long it would take to process it for the Web site. An analysis of the time it took to link homework and lab assignments indicates that the project did not always process these time-sensitive materials in a timely manner. Sufficient and consistent staffing is needed, as well as alerting mechanisms, to ensure timely processing of these materials.
At UB, we continue to support distance education courses with course Web pages integrated with electronic library resources and services, with a particular focus on providing online access to course reserve readings. A new position has been created within the UB Libraries, that of Library Information Technology Specialist. Forty percent of that person's time is devoted to supporting distance learning initiatives within the UB Libraries. This past fall, that effort included supporting a graduate-level geology course offered through a recently established distance learning consortium made up of the geology programs at SUNY Buffalo, Albany, Binghamton, and Stony Brook. In this same period, a trend has been developing on campus toward increased faculty development of Web course pages that provide access to many of the types of course-related materials that the library has been making available to students in distance education courses. While we continue to process and place on the UB Libraries Web site all types of course-related materials for distance education courses, our supporting role may in time shift more to building and maintaining a comprehensive electronic reserve system to serve the needs of all UB students--a process we feel more confident embarking on with the experience and expertise gained through the OET project.
Nancy Schiller
University at Buffalo
E-mail: schiller@acsu.buffalo.edu
Nancy Schiller is Engineering Librarian in the Science and Engineering Library at the University at Buffalo. She served as the Project Director for the SUNY OET grant reported on in this article.
Nancy A. Cunningham
Buffalo Museum of Science
E-mail: nac@acsu.buffalo.edu
Nancy A. Cunningham is Assistant Librarian in the Research Library at the Buffalo Museum of Science and Research Assistant Professor of Microbiology at the University at Buffalo. As one of the two SILS students who worked on the OET project at UB, she was responsible for inputting the project data and generating data reports.
| Copyright © 1998 - All Rights Reserved. All commercial use requires permission of the author and the editors of this journal. |
|
The Journal of Library Services for Distance Education <http://www.westga.edu
/library/jlsde/>
State University of West Georgia - Carrollton, Georgia Vol. I, No. 2 - June 1998 - ISSN: 1096-2123 |