Introduction

Online Format

Or, How This Dissertation Got Its Spots

The version of the dissertation you are currently reading is only one of several ways to see this material. You might be reading a printed copy, viewing a version running off of a CD-ROM disk in a computer, or visiting the version accessible via the Internet. Because the focus of the report is an Internet-based-web site, my advisor and I decided that the best way to discuss the site would be to have "readers" see the site and any commentary about it online. However, this decision brought up several tricky questions (not necessarily in the order in which they were considered):

• What social barriers are there to an online dissertation?

• How should the information best be structured?

• What technical issues need to be addressed?

• How does an online dissertation affect the reading?
  What burdens does an online dissertation place on the reader?
  What advantages does an online dissertation offer the reader?

Social Issues

As far as I was able to determine, my request to submit an online dissertation was the first of its kind, and not surprisingly, was denied by the University of Washington Graduate School. The Graduate School requires that all dissertations be in printed form. Representatives of the Graduate School acknowledged that in part, the request was denied because it was the first time it had happened, and there weren't any policies in place to deal with the situation.

However, there were also more legitimate reasons for requiring a printed version, the most important being that dissertations are supposed to be accessible to as many people as possible, and the current way of insuring that is to store printed copies of the dissertation in the university library. Since my dissertation is concerned with community participation, and by extension, community access, I could hardly deny that reason.

On the other hand, at the University of Washington libraries, there are publicly accessible computers that can either read CD-ROM disks or connect to the Internet, so the argument of physical accessibility is less persuasive. Applying the accessibility argument to the skills required to "read" the dissertation paints a different picture. The printed version requires only reading skills to understand. The online versions also require the ability to understand and manipulate a computer, and possibly, familiarity with use of the Internet. While these computer literacy skills are becoming more widespread, they are still not prevalent in a large segment of the general population.

In addition, accepting an exclusively online dissertation would require the university to have some way to insure that the dissertation could be viewed in perpetuity. Printed documents still have a major archival advantage over electronic documents. The lifetime of the media used to store electronic documents is on the order of decades, rather than the centuries that archival quality paper can last. More importantly, the rapidly changing nature of technical standards and incompatible forms of storing electronic information make it increasingly difficult to read historic electronic documents. For those reasons, it is unlikely that the university could easily guarantee future access to an online dissertation.

Despite those concerns, we decided that an interactive, online version of the dissertation would provide additional benefits to those who were able to access them, without detracting from the material that was available in the printed version.

Structural Issues

This dissertation has ended up as a hybrid document—the Project Description and Analysis are best viewed as an online document, while all of the other sections are probably more easily read as a printed document. This split resulted from the decision to enhance the Project Description and Analysis by making it interactive, allowing the reader to see how the original Budget Builder web site worked, and immediately view a discussion of the development of the site.

Although the Project Description and Analysis is the heart of the dissertation, it is not the entire work. I also needed to decide how to handle the other sections, such as the Introduction, Theoretical Background, and Research Methods. Solely due to having written much of those sections earlier in the process, they were already in a form that favored linear, printed text for ease of reading. We decided it was best to leave them in their linear form (with slight modifications for the online version) primarily because of the additional time it would take to fully develop both printed and modular, hyperlinked, online dissertations. Furthermore, we decided that instead of writing a linear, printed version of the Project Description and Analysis, that I would reproduce the online version on paper.

At this point I was imagining two documents: the static, printed version required by the university, and the interactive, hyperlinked version available online. It bothered me that readers of the printed version would miss out on the interconnections of the online version. In addition, I worried that neither the university nor I could guarantee that the online version be available. Therefore, I decided to add a copy of the online version stored on a CD-ROM attached to the printed dissertation. If the readers then choose, they can supplement the printed copy by viewing the Project Description and Analysis online.

Although future readability of CD-ROM disks is not assured, it is the most promising current choice for longevity. CD-ROM has become the dominant form of removable storage for computers. CD-ROM drives have been standard equipment in virtually all computers produced since 1994. This ubiquity is fueled in part by CD-ROM sharing the same form and production as audio compact discs, leading to miniscule production costs. Furthermore, newer disk drives with advanced features are designed to read CD-ROM disks because of the large number of CD-ROM disks available. Although new disk formats are sure to arise in the future, the sheer volume of CD-ROM disks and drives will insure that they are accessible for many years. In terms of longevity, optical disks such as CD-ROM are considered to have a longer storage life (50-100 years) than magnetic disks (5-10 years) (Stepanek 1998).

Technical Issues

After deciding to move forward with the online dissertation, several technical decisions were required to effectively present the information. First, I decided that the most effective way to discuss the development of the Budget Builder Web site would be to have a parallel commentary Web site "behind" the original site. Specific features and principles of the original site are analyzed in detail in the second site.

Each decision along this path spawned additional questions and issues. First, because the project being described is still proceeding, it is not possible to directly annotate the working site without creating confusion for users of that site. Also, the features and content of the working site will certainly change. Therefore, I needed to create a copy of the Budget Builder web site at a particular point in time. This means that there are two versions of the site in existence: the current working site which will be continually upgraded and enhanced, and possibly moved to a different computer; and the Budget Builder site discussed in the dissertation, which must remain constant.

Second, deciding to create a second, parallel web site containing annotations required a number of organizational and practical challenges. The second web site needed to be designed with its own look, graphic consistency and navigation aids to distinguish it from the first site, and to enable the reader to read the information easily. In addition, the first site needed to show that there is an annotation "behind" it. Because web pages are rigidly structured, it is difficult to overlay an annotation mark (such as a footnote) on the original page without changing the entire layout of the page. My solution to the problem involved making a single, large, noticeable change to the original page:

0. The header on each page of the original Budget Builder web site looked something like this:
   
   Note the lined paper background.

1. The header for each page of the online dissertation looks like this:
   

2. Combining the two headers yields this:
   
   Note the folded-over corner to the right. This is what readers see at the top of each page of the annotated Budget Builder web site.

Third, the online version (and CD-ROM copy of the online dissertation) required me to consider how I might create and store an electronic copy of the dissertation that would be as long-lived and accessible as possible. Guidelines for the electronic version included:

• Storing the information in a form that was readable by as many varieties of computers as possible.

• Storing the information on a CD-ROM disk (see above for a discussion of CD-ROM as the appropriate disk format) in ISO-9660 format.

• Choosing and sticking with a defined standard of HyperText Markup Language (HTML) so that the document was likely to be interpreted and displayed correctly by older, current and future HTML-aware software.

• Removing references to Java and JavaScript programs, which supposedly work on many different types of computers, but in practice are often incompatible.

• Removing dynamic aspects of the original Budget Builder web site which depend on a specific type of computer or a specific piece of software, any of which might not be available in the short-term future.

Adding an online version of the dissertation creates many issues for both the readers and author. Two interrelated, although different, issues come to the forefront: the differences between electronic and paper documents, and the differences between linear and hypertextual information.

Electronic Documents
An electronic dissertation requires that the reader view the information on a computer screen. There is much anecdotal and experiential evidence that many readers find it uncomfortable to read on screen for extended periods of time. However research into reading speed, accuracy, fatigue and comprehension is inconclusive (Wright and Likorish 1983, Gould and Grischkowsky 1984, Gould et al. 1987). Additional studies seem to indicate that as the resolution of computer screens increases, differences between paper and electronic documents disappear (Wilkinson and Robinshaw 1987). On-screen reading also requires the reader to remain in one location. With a portable computer, that requirement is not true, but portable computers often trade smaller, dimmer screens for portability, making viewing text more cumbersome.

Electronic reading offers the advantages of being able to search quickly for any text, and, if accessing the online version of the dissertation, it could be done from anywhere in the world there is an Internet connection. The reader does not have to go to the library to get the printed dissertation, nor worry about returning the dissertation, or whether or not it is overdue. The online version does, however, require that the reader have access to a computer with either an Internet connection or a CD-ROM drive, and know how to use it, although both the computer and assistance are likely to be available in the library.

Hypertext Documents
Hypertext is defined simply as "nodes (or 'chunks') of information and links between them." (McKnight, Dillon, and Richardson 1991, 2) The term often implies electronic documents, although there is much disagreement about making that connection. For example Farkas (1992, 157) believes that "hypertext concepts are relevant to the study of paper as well as electronic documents," and can "reveal underlying similarities and differences among such diverse documents as Mark Bernstein's pioneering hypertext history The Election of 1812; multimedia kiosks; newspapers; encyclopedias; and the full gamut of conventionally designed non-fiction books, technical reports, and other print documents." The most obvious print example of hypertext is the footnote; the main text and the footnote are the nodes of information, with the footnote number being the link between them. Other common examples are sidebars in magazine articles that are referenced in the text, or "see also" references in encyclopedia entries, which direct the reader to additional information.

On the other hand, McKnight, et al claim that "what makes hypertext different, what sets it apart from the most conceptually inter-linked paper document, is that in hypertext the links are 'machine-supported.'" The computer makes it possible for "the 'movement' between the two nodes to take place automatically," reducing the work of the reader. The concept of an electronic hypertext was created by Vannevar Bush in 1945 when he was serving as President Roosevelt's chief science advisor. The term 'hypertext' was coined by Theodor Nelson (1972), who envisioned linking the all of the world's information in a giant 'docuverse.' Both of their visions are moving closer to reality with the recent introduction of the World Wide Web, the largest and best-known example of an electronic hypertext. Of course, the view of hypertext expounded on by Bush and Nelson assumes that only information that can be reduced to a computerized form is important. Other ways of knowing that cannot be represented by digital media are not part of their discussions.

Hypertext is also often thought to imply non-linearity. Although it is possible to construct linear hypertexts—and many hypertext authors do—the ability to link information in a variety of ways is what makes hypertext documents most useful. The arrangement of links, or topology, of the information can exist in many forms, from the simple linear form to trees, hub and spokes, webs, and many incredibly complicated structures.

Figure 1. Hypertext topologies

Hypertext is most successful when taking advantage of the flexibility in arranging complex information combined with computer support for keeping track of the connections (Horton 1990a, Parunak 1991). Because of this flexibility, "hypertext may permit authors to describe relationships that are difficult to express in sequential prose." (Bernstein 1990, 104)

Although the computer keeps track of the links and automatically brings linked information to the user when requested, the reader must determine how to navigate through a potentially confusing document structure. A disadvantage of electronic information is that it has no physical presence—it is usually difficult for a reader to gauge how much information is at hand when there is no physical object to examine for size and weight, or to flip through to see the number of pages or how many pictures there are. References to "feeling lost" are common among hypertext users and researchers (McKnight, Dillon, and Richardson 1991, 65). However, Bernstein (1990, 106) claims that "if readers cannot possibly become lost, they feel little sense of achievement in exploring complex relationships. If readers never stray far, they will not feel that the road they have travelled was either long or significant." Nevertheless, hypertext authors are typically encouraged to add navigation aids to provide some sense of the document's structure and assist readers in locating themselves in the document.

Another effect of not having a clear sense of the full extent of a hypertext document is the sense of wondering whether all of the information has been seen. Again, without physically seeing how many pages of a document have been read, it can be difficult for readers to get a sense of how much of a document has been covered and how much is still hidden.

Many of these effects of hypertext can be controlled to some degree by authoring techniques. The author has control over the structure and organization of the information, and the navigational and informational clues provided to the reader. This can make it more difficult to author hypertextual material. The tradeoff is that with well-designed and written hypertext, the reader has more flexibility in choosing the path through the information than might be available in a printed text.

The World Wide Web (WWW) is a specific instance of hypertext that does not have many of the features of other hypertext systems. The methods for connecting web pages were originally developed by physicists in Switzerland to communicate research results and scientific documents (Berners-Lee, Groff, and Pollermann 1992). Since then, the web and its underlying technology have grown to encompass the needs of an enormous variety of disciplines. Today, almost all people who use the World Wide Web interact with it using a web browser—a piece of software that graphically displays nodes of information (typically referred to as web pages) and the links between the nodes.

Unfortunately, there are several problems with the current incarnation of the web standards and the web browsers used to read information:

• All web links look the same
  There's no information built into the links to give the reader clues about the nature of the information on the other side of the link. Is the link to a definition or short explanation of a concept, or is it a link to another web site with in-depth background material?

• Links never die
  There are no automatic mechanisms to update links if the information on the other end changes, moves or is deleted.

• Information is difficult to find
  The overwhelming amount of information available on the Internet is simultaneously one of its strengths and weaknesses. Although many companies are competing to provide indices to the web and many others offer to add web sites to the indices for a fee, it is estimated that only about 25% of the information on the web is indexed. (*** newspaper?, MacWeek? not New Media)

• Web browser support for user navigation is awful
  Most of the navigation improvements in the last four years "are aimed at serving media publishers: very few new features actually help users get around and manage their information needs better." (Nielsen 1997)

Overview | Research Goals | Wider Context | Online Format

Contents | Introduction | Background | Methods | Description | Conclusion | References