Vancouver, British Columbia
May 15-17, 1996
Carol Mandel, Deputy University Librarian
Columbia University
Thank you for welcoming me here today. My presentation this morning is about books, a rather familiar topic to this audience. But what we are finding in the Online Books Evaluation Project at Columbia University is that we still have a few new things to learn about books and their use.
We already know that books are used in a variety of ways. You may recall that, when asked about the usefulness of books, Mark Twain observed that a leather-bound volume was an excellent razor strap, a thin book was good for fixing a broken table leg and a thick heavy book was ideal for throwing at a noisy cat. We are amused by Twain’s remark because we think that a book is important not for its form, which he was emphasizing, but for its content. But the relationship of object to content is complex.
A print book is a known package with a title page, pages, and chapters. It has been scaled to human needs and perfected over centuries of use. The advantages of moving book content out of this likable package and onto a computer screen are not immediately evident. Thus, the Online Books Project is focused more on evaluation and product development than it is on production. And, so, at the heart of our evaluation project are the questions: What is an online book? Why would anyone want to use one?
How will we find that out? For the project we are repackaging books for online delivery from typesetting tapes. It is an interesting experience. We don’t want to recreate the electronic text and redo copy-editing. So we must address the problem of legacy systems. We are definitely working with publishers who have legacy systems and we’re not trying to change that at this point.
We are making these electronic texts available to the Columbia University community, and we will then study how they are used. We will analyze the costs of development, delivery, and use, and we will relate that use to the kind of environment in which our users are working, their level of computer sophistication, to intellectual property features, and to system features.
The books come from a number of publisher partners. We began our work with Columbia University Press. We have been working with them for a while, putting reference books online. With the Columbia Press we became very interested in what other kinds of book formats would work online. We are starting to think that different kinds of books will need different kinds of products.
We are also working with Oxford University Press. The Oxford University Press has a very specific set of issues. They are interested in trying to rescue the scholarly monograph, and finding out whether there are aspects of online delivery that can make the scholarly monograph economically feasible. So that is another area we are studying.
I’m pleased to note that Garland Publishers joined our project a couple of weeks ago. They have a different perspective, as well. They publish specialized encyclopedias as well as anthologies and specialized monographs. They wonder if, by linking these products and creating new hyperlinked databases from them, they can create a new product.
We have some titles from Simon & Schuster in higher education, essentially from Prentice-Hall, that we are also trying to work with, selecting high-use titles. For comparison, we have on our network, as many of you do, a set of humanities texts: texts that are used in our core curriculum. Notice I use the word “texts.” We tend to think of these things more as texts than books because they are “classic” content. But they are, after all, books and we want to study their use in comparison to other kinds of books.
I think it is important to note that we’re putting up these books in the context of our overall program of digital library services, which includes a broader digital university concept of services. That the scholarly content piece is a part of that is particularly important because many of these books will initially be used more in instructional roles than just in typical take-the-book-out-and-read-it kinds of roles. We’re working with faculty because we want to see how these books are used in instruction, giving us an opportunity to compare the way the digital texts are used with the other kinds of instructional support tools that we have made available electronically.
Another important part of our environment is the wide availability of network access and that we make a point of creating public access to the Web and its services throughout the campus. What’s important here is that we are aiming for ubiquitous delivery of these online books. We aim to deliver them in a format that anyone can use anywhere, and what that essentially means, at the moment, is web browser technology. It may be something else later. But the point is, we are trying to deliver these in a way that everyone will have the interface to use it and specialized software won’t need to be developed.
Something that we do have on our campus that isn’t as universal is a Kerberos network authentication system that we have used for some time for verified access to our information resources. From that we have built, over recent years, a very full user database, so that we have demographics on our users and we can match users to uses. This will be very important in our evaluation. By this summer (it is currently being tested) we will have developed a way of linking that user database and authenticated sessions to our web delivery, so that we won’t just measure a “hit” for a web use, but we will know that the hit is a use of a particular book and which hits were linked to that person, etc.
Mr. Ekman asked us to talk about our testable propositions, so I will just list some of the key questions that we have. Of course, we want to know whether online books will complement print ones, will replace them, or will not be used at all. We want to know which formats and what kinds of books they’ll be used for and by what kinds of users. We also are trying to look at the impact of book use on users, and are asking users if they feel that online books help them be more productive or more effective. We want to look at who will be using online books and who will be the early and who the late adopters.
We have begun a study of the life cycle cost of online and paper books. We will also be looking at the implications of intellectual property regulations and commercial traditions. That word “traditions” is particularly important because publishers have a certain perspective on how things should be delivered. In fact, publishers have their own set of questions. They’re interested in all the questions we’re interested in, but they also have their own. They’re in this because they want to understand how to package and market online versions of books, and want to find out what, ultimately, the product will be, an answer that is much more straightforward, I think, for journals than for books.
Oxford University Press, as I mentioned, is particularly interested in whether there are ways in which online delivery can save the scholarly monograph. One thing that they are examining is whether online versions can promote print sales. We’ll be offering the ability to order a book online and then pay a discount price for the print version, looking at whether the online version might be a teaser or a promotional device for people who may want to pursue it further and buy it in print. Of course, publishers are also worried that online availability may just decrease print sales.
Oxford is also curious about what kinds of printing sales might result from electronic delivery. They would like to know, for instance, whether selling books by the chapter might become viable. Later I’ll talk a bit more about how technology is foiling us in that area. Finally, publishers want to find out if they can develop licensable products from this. And, of course, publishers are worried about whether authors will be willing to publish in online-only versions.
We have developed an evaluation plan with a matrix of evaluation methods. We will be studying eight classes of relevant variables.
We are interested in sharing our evaluation methodology with others. We have put it up on the Web and will continue to put up information because we are interested in having others replicate our studies. We are investing a lot in designing questionnaires and want to make them available for others to use. If you are interested in doing these kinds of studies, we would be happy to help.
Mr. Ekman is anxious that we share some of our preliminary findings with you. Our first finding is something you already knew: that start up takes longer than we are capable of imagining. I mostly mention this to alert you that what I’m talking about are really trends and emerging issues more than findings at this point.
I do, though, have one conclusive result, because, of course, we’re using print use as a benchmark and we don’t have to worry about start-up with print, it’s already out there. Our project coordinator, Mary Summerfield, has conducted a study of the use of print materials based on literature review and some studies of her own. One of the things she has found, and you probably know this intuitively, too, is that books are for use but not necessarily for reading. The statistics indicate just how infrequently people read.
So what are people doing with books? They’re searching, they’re browsing, and they’re trying to assimilate the information in books. This bodes very well for online books because we know that it’s not particularly comfortable to read on the screen. But it is fun and useful to do these other kinds of functions, and our work with potential and current users of online books is confirming that what they are seeking in an online book is functionality.
I believe that the future of online books is dependent upon the kind of system functionality that will be available. In fact, our users are telling us that. The future of online books is not in scanned bitmaps or in “silicon microfilm,” it’s in functionality. Again, this may be different from journals because for journals people are looking for a print server; if I can find my article and get my information and print it out, that’s fine. But that’s not what they want from books. So the kinds of features that we can deliver will be the deciding element in the user’s choice between print and online. The importance of functionality, then, has affected us in the way we’ve looked at web browsers. We’re not trying to put out specialized technology, we’re trying for ubiquitous delivery. For a while at least, the success of online books and whether they work or not will be linked to the technology that’s currently out there. Right now web browser functionality for books leaves something to be desired in certain aspects, particularly navigation and referencing. It’s hard to navigate books, it’s hard to know where pages are, and it’s hard to cite them. It’s a problem.
This leads us to the realization that the success of online books will really be determined by things that we can’t plan: unplanned system features and some unexpected human factors. Agility is important in planning, in terms of seeing what the best developments are and seizing on them and shaping them, because we won’t be able to control this aspect.
I want to give you just a couple of examples in terms of system features. As I said, Oxford University Press came into this project with a set idea of printing as an intrinsic part of what the product of an online book would be. Potentially they would like to charge for printing and would like to sell books by the chapter. When we started talking about this a couple of years ago it seemed possible to control printing in this way. We knew we couldn’t control screen prints, but those were screen prints, and we thought we could control printing. Well, in an HTML web environment there is a print function that we’re not about to disable on campus, from which some pretty nice printing can be done that can’t be monitored at all. You can’t tell whether people are reading or printing text. This is really a site-licensing mode, it’s not a print-packaging mode. We’re still struggling with this for Oxford. We may discover other ways of making printing available and see if we can make them attractive enough to interest users. But I think technology is telling us that this is not a direction in which we will be able to go.
In terms of unexpected human factors, user acceptance is really one of things we can’t predict. Here I want to compare online books to another online instructional product that we’ve created. One of the most popular services that we have put up this year on our digital library has been a collection of art reserve images. Under the Columbia core curriculum requirement, all first- and second-year students are taking many of the same courses. The Art Humanities course then has a set body of images that are studied in all the classes. We have put those images online— is, the ones we could get intellectual property rights for— it’s a very popular resource. In a typical week there are approximately 3,000 hits on that system. During exam week there were 29,000 hits on that database. So the students are definitely using those images to study from.
We worked with the instructors to put that up. The instructors love it. The students love it. And we thought, well, this is a good model. So we went to the literature humanities instructors and said, “Hey, we have these core texts online. Wouldn’t you like to use them, and how can we do that?” And they said, “We’d have to change all our assignments because the students have the books, you see. They don’t need these texts to study from.” To take advantage of the functionality of the texts they would have to change the way they taught and the kinds of assignments they gave, whereas the art instructors didn’t have to change anything they were doing, they could just point to the images. These literature instructors are mostly junior faculty and graduate students who can’t really invest in curriculum development at this time in their careers, so we’re going to go next to the senior faculty who are in charge of this curriculum and work with them. We’ve discovered in other fields, particularly the sciences, that these are the people who have the time to work with curriculum change. This is an example of human nature relating to the acceptance of these tools, something we just can’t predict.
My final observation is one that Ursula Bollini of Oxford Press recently made when we gave presentations at the CNI meeting. At the end of this project’s time frame online books will not yet be a widely available commercial product. I feel confident in this finding now. I guess that’s related to our first finding about start-up. Again, this is in contrast to online journals, which are already proliferating. The translation from paper to online will be more complex for books and it is unlikely to be a one-for-one replacement of digital for paper. Perhaps this makes our project an “unnatural” experiment, rather than the natural experiments that the Mellon Foundation had in mind. In any case, to end this talk with a book analogy, we’ll need to keep reading right up to the end to see how this plot turns out.
Thank you.
Copyright © 1998 by Carol Mandel