Association of Research Libraries (ARL®)

5. UNDERSTANDING FACTORS INFLUENCING USE

One of our principal goals is to understand what distinguishes users from non-users, early adopters from late adopters, and so on. We have tentatively identified several kinds of variables that might influence users with regard to adoption of electronic books. We can broadly divide these into: resource factors (discussed here), attitude factors (discussed in section 6), and behavior factors (discussed in section 7).

5.1 Resource Factors

We believe the key resource factor to be possession of, or easy access to, adequate computer equipment and connections, so that online books are a reasonable alternative to paper versions. As noted earlier in a discussion of early research results, we have two primary methods of studying this factor:

1. Asking the question about access to computers with network connections that was discussed earlier; and

2. Developing a detailed profile of the computer resources available to representative samples of members of the Columbia community. As described earlier, we used the campus computing survey instrument in three sweeps through the community, in what we regard as pilot implementations in Fall 1995 and Winter 1996. The questionnaire asked respondents for detailed information about their computers. However, many respondents did not supply the requested data on computer power, size of hard drive, modem speed, and the like. Furthermore, observation of developments in the personal computer marketplace makes it clear that changes in personal computers will make the top of the original scales on these questionnaires the bottom of the scales in as little as two years.

Section 7 contains further analyses of the first of these factors.

5.2 Attitude Factors - In-Class Survey

We cannot probe attitude factors easily in a simple survey, whether in paper or online. We have designed some questions aimed at assessing whether the respondent thinks that members of his peer group use and/or prefer electronic access to books and other resources. On the one hand, this perception of others' preferences might precede and shape a user's own preferences and behavior. Alternatively, if use of computer modalities is, as some psychological research suggests, a very private activity, awareness of the behavior of others may, in fact, lag the move to using online books.

5.2.1 The In-Class Survey

Examples of such results include the data from Fall 1996 and Spring 1997 in-class surveys of students who were assigned a reading that was available online for the session during which the survey was administered. Students who had done the reading were asked to answer questions about what forms of the book they used, how long they used that form, where they did that studying, format preference, reasons for it and impacts, expectations for format most used by classmates, and the two benchmark questions about computer access and time online. (See Exhibit 4.) Those who had not done the assignment were asked to respond only to the last three questions.

In Fall 1996, most of this surveying (67 percent of a total of 439 responses) was done in sections of Columbia College's Contemporary Civilization course for which some of the readings are available in the Past Masters set of humanities texts. However, those students are expected to use the assigned editions of the readings and to bring a copy to class for use during the discussion. This may well have biased students' choices in methods of reading the assignments. About 16 percent of the cases in the sample came from graduate Social Work classes and 17 percent from a large upper level undergraduate political theory class.

In Spring 1997, the Contemporary Civilization course was the source of 106 (44 percent) of the 239 respondents. Two political science courses, one undergraduate and one graduate level, were the source of another 17 respondents (7 percent). Four graduate Social Work courses were the source of the remaining 116 respondents (49 percent).

5.2.2 Methods of Studying A Class Reading

Some students had not done the reading for the course session in which the surveying was done. Some others reported using more than one method.

This table shows that, in Fall 1996, 70 percent of the responses reported using one's own copy of the text. The next most common method was to use a friend's copy (14%). The shares for those two modes are insignificantly different in Spring 1997.

The questionnaire gives four alternative means of Using CWeb Online Text:

Reading it directly from CWeb

JAKE printout of text

Printout using non-JAKE printer

Download of online text to disk & reading away from CWeb

In Fall 1996, there were 19 reports (5 percent) of printing out or downloading from the CWeb books, but none of reading directly from CWeb. In 89 percent of those cases, the respondent was not in a Contemporary Civilization class. In Spring 1997, about 11 percent of responses reported using some form of the online text, but again none reported reading on screen.

5.2.3 Preferences for Studying Class Reading

There were far fewer responses (119 in Fall 1996 and 88 in Spring 1997) as to the preferred mode of studying. In both semesters about two thirds of respondents reported that reading their own copy was preferred.

5.2.4 Reasons for Preference

As the following table shows, in both semesters, the three strongest reasons for preference were always available, easy to annotate, and easy to read, with the last two reasons switching position between the semesters.

At present, these attributes are possessed only by a personal copy or photocopies from print copies or printouts from electronic copies. (Online books are always available, but one assumes that ready physical access to a computer does not meet the criterion always as students interpreted it here.)

The cross-tabulation of preferred method of use and reasons for that preference produces logically consistent results. For example, all of the respondents who gave Printout using non-JAKE printer or Download of online text to disk and reading away from CWeb as their preferred method gave less costly as one of their reasons while few of the those preferring their own copy gave that reason.

So we have a consistent picture of what makes a mode good, preferred, and used. In other words, our student respondents are behaving rationally.

5.2.5 Impact of Preferred Method

As the following table shows, when asked what the impact of the various possible modes was, a majority of the students selected more likely to do the assignment. Learned better and doing assignment faster ranked second and third, being cited by about one third of the students.

These are again entirely rational bases for preferring some particular mode, when we note that a student's role is to get assignments done and to learn.

5.2.6 Comparison of Personal Behavior and Perception of Others' Behavior

Students were asked what method they thought most of their classmates used in order to learn whether they perceived a shift to using the online materials. In Fall 1996, 81 percent of respondents chose own copy, and 14 percent chose library copy. In Spring 1997, these values were 80 percent and eight percent, respectively. This contrasts with the reality that 70 percent (73 percent in Spring 1997) used their own copies, 14 percent (ten percent) used a friend's copy and eight percent (nine percent) used a library copy.

Interestingly, while none of the respondents had used the CWeb book directly to do this assignment, in both survey periods almost two percent (five or six students) gave that response to this question. In the fall, another nine students (two percent) and, in the spring, another 13 students (six percent) thought that their classmates had used some form of print copy or downloaded file from CWeb. Thus, students are over-estimating their colleagues' propensity to read directly from CWeb and under-estimating their propensity to read printed copy and downloaded files from CWeb.

At the present time, there is not sufficient penetration of the market by the online modes for us to draw any meaningful conclusions about leading and lagging impacts. Results of this survey in Fall 1997, particularly in Social Work classes, should give some indication of trends as students will have had more time in which to gain awareness of the availability and attributes of the online format.

[Table of Contents] [Next Page]