Internet Document Delivery: An End User Survey

Walker, FL, Thoma, GR.
"Internet Document Delivery: An End User Survey,"
Proc. IOLS '97. Medford N.J: Information Today, 1997; 145-153.

Keywords: Internet, DocView, Survey, Document Delivery, TIFF, GEDI, MIME, Ariel, Raosoft

Abstract: The Internet is serving as an increasingly useful vehicle for electronic document delivery. Many libraries and information providers are now using the Internet for delivering electronically scanned versions of printed documents to patrons who have tools for receiving and using them. An example of such a tool is DocView, a Windows-based software application developed at an R&D division of the National Library of Medicine. DocView facilitates Internet document delivery by enabling a library patron to receive bitmapped images of documents sent by a library’s Ariel system. The Ariel system sends the documents rapidly and directly to the patron’s desktop personal computer running DocView. The DocView software also facilitates two other means of Internet document delivery, by serving as a TIFF viewer for scanned documents received through Internet client software such as World Wide Web browsers and MIME email. After a number of sites had beta tested DocView for several months, they received a beta version that included an electronic survey. The survey questions were automatically presented to the tester after 45 days of use, then the tester’s responses were electronically forwarded over the Internet to the developers at NLM. The survey solicited opinions from users, many of whom were receiving library documents electronically for the first time, instead of photocopies. The survey identified which features of DocView were useful, and revealed what users thought of Internet document delivery. This paper describes and analyzes the results of the survey.

1. Introduction

Document delivery by libraries and information service providers has evolved over the past two decades. Interlibrary loan has traditionally meant photocopies of journal articles being mailed to other libraries. While photocopies are still mailed to requesters, document delivery has since adopted facsimile transmission, and more recently, Internet document delivery. As was true for decades, and is still commonplace now, library patrons often have a long wait for photocopies of articles to arrive in the mail. Besides being slow, mail has a tendency for sometimes getting lost. Facsimile transmission of documents has only partially replaced mail. While fax is faster than mail, it tends to be more expensive, and is also prone to reliability and quality problems. Sometimes the analog nature of the transmission may garble fax transmissions. Also, the quality of faxed documents is limited to 200 dpi resolution. The 1990s has seen the arrival of Internet delivery of library documents, especially with the widespread use of the Ariel^TM system developed and distributed by Research Libraries Group [1], [2]. Ariel has enabled hundreds of libraries to do interlibrary loan electronically via the Internet. It is a technology that is faster than mail, more reliable than fax, and offers higher resolution images than currently possible through conventional fax. While libraries have been exchanging documents among themselves via fax and Ariel systems, document delivery to the patron’s desktop is still problematic. DocView, a software product under development at the National Library of Medicine, addresses this critical issue, viz., document delivery via the Internet to the patron’s desktop [3], [4], [5], [6].

DocView runs on all Windows^TM operating systems. It enables a library patron to receive documents sent by a library’s Ariel system. An Ariel system can scan a paper-based document, convert it to bitmapped images, then send it via FTP [7] to computers running DocView, as it does to other Ariel systems. DocView is capable of displaying monochrome bitmapped images in either the Group on Electronic Document Interchange [8] (GEDI) file format used by Ariel systems, or in the Tagged Image File Format [9] (TIFF). DocView gives the user many other features that facilitate use of library documents received via the Internet:

• It has a built-in document ordering function that allows a user to compose a text message and send it directly to an Ariel system or DocView running at a library. If received by an Ariel system, the document request may be automatically printed within seconds. If the library uses DocView, the request may be viewed or printed.
• Since DocView can serve as a TIFF viewer, it can display documents received by other Internet client software such as Web browsers or MIME [10] email.
• DocView’s image processing capabilities allow the user to view, zoom, shrink, scroll, pan, rotate and print received documents.
• DocView has an electronic bookmark feature that allows users to "mark" pages of interest to allow easy traversal through large (multipage) documents, and for printing selected pages rather than the entire document.
• It has a copy function that allows users to copy parts of a document to be pasted into documents created by other Windows applications (e.g., word processors).
• DocView's document management facility assists the user in organizing collections of bitmapped documents received through the Internet.
• DocView also contains a detailed help facility on any aspect of operating the user interface, in addition to a built-in tutorial. The user may obtain context sensitive help by pressing the help button associated with any function.

Beta testing of DocView began in August 1995 in the Washington, D.C. area and it spread to six other sites by late spring 1996. Initial comments received from beta testers were used to improve DocView’s functionality. These included the addition of the document management and messaging facilities. After the software underwent several design iterations, the DocView developers decided to expand the beta testing. During the summer of 1996, product testing was extended to twenty-six sites in six countries and three continents. At the same time the developers created an electronic user survey tool to aid in further product development. The survey software was based on Raosoft^TM Survey, a DOS software product that runs under all of the Windows operating systems. Raosoft Survey allows a questionnaire to be created, and includes tools for collecting and analyzing survey answers. DocView was modified to run the survey software forty-five days after initial installation by the user. At the end of forty-five days of use, DocView gives the user an optional opportunity to anonymously answer the survey questions. After the user completes the survey, DocView forwards it automatically over the Internet to a computer in the R&D labs at NLM. This computer, the "DocView Comments Server," runs continuously, collecting user comments and survey results sent electronically by DocView software over the Internet. This paper summarizes and analyzes responses from twenty-eight anonymous participants received at the DocView Comments Server.

The Survey

The survey contained a total of thirty-seven questions grouped into these seven areas: user background; computer capabilities; DocView usage; opinions on DocView; image legibility; ease of learning; and DocView features. While most of the questions were multiple choice, a few solicited written user comments. The survey was designed to require no more than fifteen minutes, on the average, for a user to answer all questions. A discussion of answers in each of the seven areas follows.

2.1 User Background

The survey’s first area of interest was user background. This section contained two questions. The first solicited the user’s opinion of his or her Windows experience level. The second question was on the user’s occupational (professional) area. These questions help to find out whether a user’s opinions on DocView could be influenced by software experience or occupation. About 10 percent of the users claimed "extensive" experience with Windows, 43 percent had "above average" experience, and 36 percent "average had average experience. Altogether, about 89 percent considered themselves at least average Windows users. The second question listed four occupational areas, and elicited a job description. Fifty-one percent said they worked in libraries; 26 percent were employed in basic research; 7 percent in clinical medicine; 7 percent in administration; one person was an educator.

2.2 The User’s Computer

The second survey area concentrated on the user’s computer. It elicited information on the user’s computer speed, amount of computer memory, and display resolution of the monitor. These questions were intended to understand the effect of the capability of the user’s computer on his/her perception of DocView. The computer speeds were fairly evenly distributed. Twenty-two percent of the computers were 33 MHz machines, 37 percent were 50-66 MHz, 15 percent were 75-90 MHz, 15 percent were 100-120 MHz, and 11 percent had speeds of 133 MHz or higher. By today’s standards, 60 percent of the computer would be considered slow (66 MHz or below), since the slowest PC commonly sold currently has a CPU running around 100 MHz.

The next two questions in this area asked about the computer memory and monitor. About 27 percent had the minimum-recommended 8 Mb memory in their computers, while the remainder had 16 Mb or more. A total of 42 percent had the minimum-recommended VGA resolution (640x480). Twenty-seven percent had super VGA (800x600), and about 31 percent had resolutions of 1024x768 or higher.

2.3 DocView Usage

The third area in the survey focused on the usage of DocView. Over 70 percent said they used it to receive documents from their library; and about 11 percent said they used it as a TIFF viewer. One user was a librarian using it to receive document requests from library patrons. Another question asked users how long they had been running DocView. Seventy-five percent said they had been using it for less than two months, while eleven percent had been using it for nine months or more. A third question asked users what they did with a document once they had received it. About 27 percent said they did not view the document; instead they printed it and deleted it from their computer. This is interesting, since these people were using only a fraction of the total capabilities of the software. Another 23 percent said they view, print and delete the documents they received. This amounts to 50 percent of respondents who do not store the documents on disk while the remaining 50 percent do at least for a while.

The next three questions attempted to find out whether having DocView increased the number of documents received from the patron’s library. These questions were as follows:

During the past month the approximate number of documents I received using DocView was…

During the past month the approximate total number of documents I received from all sources was…

Prior to using DocView, the approximate number of documents I received monthly was…

For the first question, the average number of documents received using DocView was 10, while the median was 3. The average was considerably higher than the median because two users had received a lot of documents (80 or more). For the second question, the average was 14, while the median was 3. For the third question, the average was 11 and the median remained at 3. The responses to these three questions are interesting, because having DocView meant that the average number of documents received in a month increased by 3, a growth of about 30%. Also, since DocView was supplying an average of 10 documents of the 14 received, this meant that it largely displaced the means of document delivery previously used.

2.4 Opinions of DocView

The fourth area of interest in the survey asked users for their subjective opinions on DocView. This consisted of four statements requiring subjective answers, plus a written comments section. The first statement was:

DocView improved the way I obtained documents.

Here, a total of 69 percent felt that DocView had indeed improved the way they receive documents. About 27 percent were undecided, and one person felt that document delivery had worsened with DocView. The next statement was:

My sessions with DocView were straightforward.

A total of 77 percent of users agreed with this statement. The third statement:

I want to continue using DocView.

An impressive 84 percent of users agreed with this statement. The fourth statement:

I would recommend DocView to colleagues.

Again, a high 77 percent of respondents agreed with this statement. The final portion of this section allowed users to enter comments. The vast majority were favorable, as in the following examples:

"As a librarian who is promoting DocView to our patrons, I love it! It is very easy to use. Our patrons who use it, love that they are able to receive their ILL requests electronically."

"It's a great program!"

"Excellent."

"Certainly is a pleasant advancement in document delivery."

2.5 Image Legibility

The survey’s fifth area had two questions dealing with image legibility, plus a written comments section. Eighty five percent of respondents felt that DocView’s images were easy to read. A few thought there was still room for improvement. As mentioned in section 2.3, 27 percent of respondents said they did not even view the document on the screen. Although most users felt that image legibility was good, there were some who obviously felt that viewing document images was not necessary, and that having a paper copy of the document was more comfortable or convenient.

2.6 Ease of Learning

The survey’s sixth area had questions dealing with the ease of learning DocView, as well as the usefulness of the help facility, the built-in tutorial for new users, and the online User Manual. A total of 88 percent felt that DocView was easy to learn, and about 84 percent felt that it was as easy to learn as other Windows applications. To a question on DocView’s built-in help facility, fifteen percent of the respondents said they never used it, and only 58 percent felt that it was useful. These low numbers suggest either that the help facility was not particularly useful, or that users really did not need help with DocView. One question asked about DocView’s User Manual, which is an online manual obtainable through the World Wide Web. A total of 23 percent of the users did not try to access the User Manual. Another 57 percent thought it was useful; the rest were undecided, except for one person who felt it was not useful. Another question asked about the tutorial built into the online help facility for new users. Forty-two percent of users responded that they did not use it.

The help facility, tutorial and user manual for any software application are intended to aid the new user in learning how to use the product. When a person starts using the product immediately without opening an instruction manual, this may be taken as a sign that the product is easy to use. The low numbers obtained concerning usage of the User Manual and help facility appear to indicate that most users felt that learning DocView was an easy experience.

2.7 DocView Features

The final area of the survey asked the user’s opinions on specific features of DocView. These are briefly described as follows:

DocView response time – 80 percent of users felt that response time (e.g., time to display images) was fast enough.

Bookmarks – 44 percent of users used the bookmark feature. This feature allows a user to electronically mark a page. Bookmarks make it easy to traverse a large document by marking just the pages of interest and to print only selected sections of a document. It is possible that not many people used bookmarks because the documents received were small. Most journal articles sent by interlibrary loan services between medical libraries average 10 pages. For a document of this size, it probably is not necessary to use bookmarks.

Zoom/Shrink – 77 percent of users felt that the ability to zoom and shrink an image was useful. This is something that is necessary on smaller monitors.

Rotate – 58 percent used the rotate feature. This feature is useful for rotating images that were scanned upside down, or appeared in the original paper document in landscape mode.

Copy to Clipboard – 47 percent used this feature; only 30 percent felt that it was useful. This feature is useful for creating new documents from the ones received over the Internet.

Print – 84 percent felt that printing was useful. Some people commented about slow print times, which is to be expected, since printing is 100 percent graphics-oriented. Printing speed depends very much on the speed of data transmission through the parallel printer port for a locally attached printer. Printing speed decreases appreciably when the printer is located on a network server.

Message Function – 70 percent used this feature and 53 percent felt it was useful. The message function allows a user to compose a text message and send to an Ariel system (for document orders) or to another computer running DocView. It will be interesting to see if usage of this function increases with the new functionality in DocView that was included after the survey was taken (see 4.0 Future Development).

Document Management Facility – 65 percent used this feature and only 34 percent felt it was useful. These low numbers reflect the fact that a lot of users did not store documents on the computer.

Reliability – 84 percent felt that DocView rarely failed.

3.0 Survey Analysis

An analysis of the answers to the survey questions revealed some correlations among the answers. In one case, users who said they would recommend DocView to colleagues also viewed the software in a positive light. For instance, they were the users who said that DocView improved the way they received documents, and that their sessions with DocView were straightforward. These were also the users who felt that DocView was easy to learn, who felt that DocView had a fast response time, and who felt that document images were easy to read on the monitor. In a second case, those users who stated that they wanted to continue using DocView were also the users who said that DocView improved the way they received documents, had straightforward sessions with DocView, and who felt that DocView was easy to learn. In a third correlated case, there was a direct relationship between users who felt that DocView images were easy to read with those who stated that DocView had improved the way they receive documents.

User experience level with Windows was analyzed to see if it had any effect on perception of DocView. There were only two respondents who stated they had very low initial experience with Windows. Both of these people said that DocView was about as easy to learn as other Windows applications, and that they would recommend it to colleagues.

An analysis of job function reveals that all the testers who worked in basic research and clinical medicine felt that DocView improved the way they receive library documents. It is interesting that testers with jobs in libraries split on this question. Sixty-four percent with library occupations felt that DocView had improved the way they receive documents. Twenty-nine percent were undecided, and seven percent were negative. It is possible that the testers in basic research and clinical medicine had previously received library documents primarily through mail, and that Internet delivery was so much of an improvement that they were positive about DocView. The difference of opinion might also reflect the fact that library employees, unlike researchers and clinicians, are probably not end users but facilitators of interlibrary loan, and therefore have access to other systems (fax, Ariel), and are less concerned about receiving documents on their desktops. The statistics for testers in library occupations improve when they responded to whether they wanted to continue using DocView. A full 85 percent said that they wished to continue using it.

Further analysis shows that 80 percent of the users with very slow computers (33 MHz) felt that DocView’s response time was adequate. Also, 83 percent of the users with computers having only 8 Mb of memory were satisfied with DocView’s response time.

This indicates that although DocView runs slower on older, slow computers, DocView’s design specifications calling for a minimum speed of 33 MHz and a minimum memory of 8 Mb are justified. Another hardware factor worth examining was the display resolution. It was found that 90 percent of the users having the minimum resolution monitor (640x480) felt that DocView images were easy to read. Although this is the minimum common resolution found on nearly all personal computers, it appears to be quite adequate for running DocView.

4. Future Development

After the user survey was conducted, an important new feature was added to DocView in late 1996: the ability to send documents using either Ariel FTP protocols or MIME email. This new feature gives both libraries and their patrons a powerful additional capability in using library materials. For the library delivering documents, this feature lets it receive documents sent by another library’s Ariel system, then forward it to a patron having DocView. It can also forward the received Ariel document via an external MIME email client to any patron having a UNIX or Macintosh platform and a suitable TIFF viewer, thereby allowing a degree of platform independence. For the library patron, this feature facilitates research collaboration, since comments can be added to a document received from a library, and then both the comments and document may be forwarded to colleagues.

DocView development continues to rely on user feedback. For example, a 32 bit version is being developed to fully take advantage of unique features in the Windows 95 and NT operating systems. Areas being researched and developed for the 32 bit version include image processing to enhance defective (e.g., skewed) images, and document delivery via the World Wide Web. It is likely that another user survey will be sent out with DocView later in 1997 to see how well libraries and their patrons take advantage of these future capabilities.

5. Summary

New technology is changing the way patrons of libraries and document delivery services get documents. With DocView, a library patron has a useful way of receiving and using library documents sent electronically over the Internet to the desktop. A survey sent to DocView beta testers indicates enthusiastic acceptance of Internet document delivery. Although not all beta testers used every feature in DocView, the vast majority felt that DocView was highly useful, that it improved the way they receive documents, and that they would recommend it to colleagues.

6.0 References

1. Berger, MA. Ariel Document Delivery and the Small Academic Library. College & Undergraduate Libraries, Vol. 3(2). The Haworth Press, 1996; 49-56.
2. The World Wide Web address for Research Libraries Group is located at this URL http://www.rlg.org/.
3. Walker FL, Thoma GR. Access to document images over the Internet. Proceedings IOLS'94. Medford NJ: Learned Information, 1994; 185-97.
4. Walker FL, Thoma GR. DocView: Providing Access to Printed Literature through the Internet. Proceedings IOLS’95. Medford NJ: Learned Information, 1995; 165-173.
5. Walker FL, Thoma GR. Internet Document Access and Delivery. Proceedings IOLS’96. Medford NJ: Learned Information, 1996; 107-116.
6. The DocView home page on the Internet is located through the Internet at URL http://docmorph.nlm.nih.gov/docview/
7. Postel, J. and Reynolds, J. File Transfer Protocol, Request for Comments #959, October 1985, available through the Internet at URL http://ds.internic.net/.
8. Agreements of the Group on Electronic Document Interchange are available through the Internet at URL: http://www.gu.edu.au/alib/iii/docdel/gediv21.htm.
9. TIFF Revision 6.0, Aldus Corporation, June 3, 1992.
10. Borenstein, N. and Freed, N. MIME (Multipurpose Internet Mail Extensions), Request for Comments #1341, June 1992, available through the Internet at URL http://ds.internic.net/.

CEB Home | CEB Projects | Related Work | Publications | Repositories | NHANES | Site Index

U.S. National Library of Medicine, 8600 Rockville Pike, Bethesda, MD 20894
National Institutes of Health | U.S. Dept. of Health and Human Services
Copyright information | Privacy policy | NLM Accessibility
USA.gov | Need a plug-in? | RSS

URL: http://archive.nlm.nih.gov/pubs/docview/iols97/iols97.php
Last updated December 08, 2003

Send questions or comments about this site to