What do faculty and students really think about e‐books?

Current use of e‐books

An obvious benchmark question to start with is the extent to which members of the UCL community had, at the time the survey was administered, experience of using e‐books (Figure 1), whether supplied by UCL or indeed through any other channel, such as a bookshop or the Web.

Users form a large minority (44 per cent) of the UCL community, a finding that is closely in line with other surveys (for example, a 2005 study of undergraduates at New Hall Cambridge found a 38 per cent penetration rate).

A number of interesting demographic patterns emerge when we look at the data for existing e‐book users more closely (diversity is a strong leitmotiv in this article, as indeed in most of CIBER's recent work). For example, existing users are:

• •

  more likely to be dependent upon searching other libraries' catalogues[1];

• •

  more likely to be dependent upon publishers' catalogues or web sites[2];

• •

  less reliant on following up references on reading lists[3];

• •

  less reliant on personal recommendations[4];

• •

  more likely to be dependent on Google or another popular web search engine[5];

• •

  more likely to report a worse experience of UCL Library Services[6];

• •

  likely to exhibit very much higher levels of awareness of UCL e‐book provision[7];

• •

  much more likely to be male[8]; and

• •

  more likely to be a graduate student, less likely to be a staff member[9].

As might well have been anticipated, age is a good predictor of e‐book use (Figure 2), and the extent of this effect is very striking.

The demographics of e‐book use are complex, since age, gender, and broad subject interest all interact with one another in various ways. As an example, Figure 3 presents the findings of a univariate General Linear Model (GLM) analysis where gender and broad subject were held constant in relation to e‐book use (for the sake of simplicity, not all subjects are shown here). For most disciplines, existing e‐book users are much more likely to be male than female (i.e. to appear lower on the y‐axis). However, in engineering and the social and historical sciences, the gender pattern reverses. Numbers are admittedly too small to make more general claims here, but this GLM analysis does point to the complexity of the demographic interactions and the need to avoid “one‐size‐fits‐all” prescriptions based on a superficial reading of highly aggregated data, as in so much library survey work.

Surprisingly, perhaps, neither status (full‐, part‐time or occasional), nor regularity of use of print library collections are associated with existing e‐book use.

Table IV shows the findings of another more detailed analysis, this time a binomial logistic regression. The intention here is to develop a simple predictive model that speaks to whether or not someone is likely to be an existing user of e‐books based on their responses to the other questions in the survey. The odds ratios in this model (Table IV, final column) show that current users of e‐books are very likely to be already aware of UCL's e‐book offerings (three times as likely), to be male (one and a half times more likely), and to be less than wholly satisfied with UCL's printed book collections (about 13 per cent more likely). These three factors enable a correct prediction to be made in 63.5 per cent of all cases. These findings suggest that many users' interest in e‐books may have been activated by their experience at UCL, a working hypothesis that will be pursued (the effect could of course equally run in the opposite direction: having already used e‐books they then seek out this form of provision more actively in a library context).

Sourcing e‐books

For those currently using e‐books, a key question for both publishers and librarians is how they sourced those materials in the first place? At this point in time, our survey finds that existing UCL e‐book users are relatively independent of library provision; with a majority (61 per cent) sourcing titles under their own steam (Figure 4).

Again, there are pronounced gender differences, with men exhibiting a greater degree of “library independence” than women (Figure 5). Doctoral students show the greatest degree of self‐sufficiency of any of the groups, with nearly two‐thirds, 65 per cent, currently obtaining their e‐books for themselves.

Reading format preferences

A crucial question for the widespread adoption of e‐books is the issue of reading format preference. E‐book users at UCL claim a strong preference for reading from screen rather than paper, and this seems to be relatively independent of their age (Figure 6).

The demographic profile for this question is particularly interesting:

• •

  men say they are more likely to read from a screen than women[10]; and

• •

  undergraduates appear to be the most likely group to read from screen[11].

The youngest group (undergraduates) shows a very marked preference for reading from screen, but the overall pattern appears not to be very age‐dependent, except for a marked fall off (on the basis of a small population) after the age of 65. These findings raise more questions than they answer: Are people really reporting their behaviour accurately? What do they mean when they say “it varies”? Why and under what circumstances?

Purpose of reading

One of the most potent guiding principles in librarianship is “To each book its reader”, a maxim coined by the famous Indian librarian (and UCL alumnus), S.R. Ranganathan. What kinds of needs do e‐books meet for academics at UCL, in their professional or personal capacities? The purpose of reading is shown in Figure 8.

At the time of the survey, users clearly associated e‐books mainly with work and study rather than leisure. The extent to which this finding is constrained by the availability of suitable titles, or whether it reflects deeper undercurrents is something that will be investigated further during the course of SuperBook.

There are no significant age or gender differences with respect to this issue. Part‐timers (staff and students) do tend to use e‐books more to support their leisure activities[12], another finding that deserves closer scrutiny. Perhaps unsurprisingly, there is a strong tendency for users to read leisure materials from a computer screen but print out the contents of work or study‐based materials[13].

The findings in Table V show that libraries are very much the preferred source for e‐book materials relating to work or study and that users generally tend to find other sources for their leisure reading. Finding other ways into this question will be important, again because the answers could be shaped as much by title availability as by any innate preference.

Types of e‐books used

A more detailed breakdown of the particular types of e‐books used is presented as Figure 9. (Note that respondents were invited to tick as many boxes as applied, thus the total responses are greater than the number of respondents).

Textbooks are clearly the most popular form for academic users, followed by reference works.

This question offers some valuable insights into the current academic market for e‐books and these are explored further in Tables VI‐VIII. Highlights include a strong preference for popular non‐fiction by men (Table VI), and for fiction and popular non‐fiction amongst undergraduates (Table VII). The findings in Table VIII suggest that there is strong competition from non‐library sources (i.e. people buy or obtain them on their own account) across all e‐book formats.

Advantages and disadvantages of e‐books

E‐books and conventional print titles offer users a very different kind of experience and this question explores user perceptions of the relative advantages and disadvantages of each format across a range of factors. The precise form of the question was that respondents were invited to tick one of the following three options for each factor:

1 = “e‐books are better”2 = “printed books are better”3 = “I don't know”

Again, only those who had actually used e‐books were surveyed.

The data in Figure 10 reveal users' perceptions of the advantages (positive values) and disadvantages (negative values) of e‐books when compared with conventional hard copy titles (the units are percentage point differences and “don't knows” are excluded for ease of assimilating the essential message).

E‐books clearly compare very unfavourably indeed with print titles for perceived ease of reading. The benefits of e‐books cluster around convenience: ease of making copies, perceived up‐to‐dateness, space‐saving, and around the clock availability. That hard copy is so decisively favoured in terms of ease of reading is a little surprising given the responses reported earlier regarding reading format preferences (Figure 6) and suggests that more work on actual, rather than self‐reported, reading behaviour is urgently needed.

Confidence in the relative up‐to‐dateness advantage of e‐books declines markedly with age[14], similarly, their around‐the‐clock availability[15]. There is a big difference between men and women in respect of features and functionality: men tend to rate these aspects much more highly[16].

Current use of library print collections

All respondents rejoined the survey at this point, where the attention shifts from actual experience of using e‐books to profiling current use and satisfaction with UCL Library Services' print collections and general book discovery behaviour.

The UCL community is segmented in Figure 11 by their current use of library print titles.

The two critical demographics relating to the intensity of print collection use are gender and age:

1. 1.

   women (42 per cent) are much more likely to classify themselves as “regular users” than men (35 per cent)[17]; and

2. 2.

   regular use peaks at 22‐25 years (44 per cent) then declines to a minimum of 28.7 per cent (46‐55)[18].

The extent of the age effect on intensity of print collection use is starkly portrayed in Figure 12. This is hardly unexpected, since the information needs of undergraduates and professors emeriti are hardly likely to be co‐extensive! The more interesting issue is the striking switchover in behaviour in middle age and whether this will persist as we move through time, or simply disappear as a generation of digital immigrants passes by?

Book discovery preferences

The question of precisely how readers identify the books they want to read or consult is seriously under‐researched in the literature, amazingly so given its significance for publishers and librarians (let alone readers). Here, we asked respondents to rate their perceived level of dependence on a range of formal and informal mechanisms for (print) book discovery. A four point scale was used (1 = “very dependent”; 2 = “quite dependent”; 3 = “not very dependent”; 4 = “not at all dependent”).

Book discovery behaviour is complex, and we can only scratch the surface in this paper.

At first glance, the findings presented in Figure 13 might suggest that individuals employ a wide range of strategies to identify the books they need for work or study. It is certainly the case that most of us do most of the things on this list at some point in our work or leisure time. However, a closer examination of the data suggests that there is real underlying structure to these preferences. Demographic factors, notably broad subject, age, and gender all play a significant role in determining which strategies tend to be emphasised.

Figure 14 is an automatic classification[19] that reveals something of how book discovery behaviour is structured at UCL. This hierarchical classification was achieved by means of some fairly heavy number crunching: briefly, the profile of answers to this multi‐part question for each and every respondent was compared with all other respondents to see if there is any discernible pattern.

Three clusters of factors are apparent. The first cluster brings together visiting other (i.e. non‐UCL) libraries and catalogues, publishers' catalogues (and web sites), and book reviews. These are all formal or semi‐formal systems of literature control that lie outside the ambit of UCL Library Services and are suggestive of specialist research requirements. The second cluster brings together searching Amazon or Google Book Search, searching Google or other popular web sites, and visiting bookshops. This grouping suggests a fundamental mode of behaviour that is highly independent of library systems and provision. The third cluster comprises activities (following up reading lists, personal recommendation from friends and colleagues, consulting UCL Library Services in person or consulting the catalogue) that are institutionally focused. While not suggesting that any of these strategies is “more important” than another, an interesting line of enquiry would be to see whether these archetypes have any explanatory power and if they do indeed map onto different kinds of information requirement.

That book discovery behaviour is a complex and highly structured set of activities, is further illustrated by Figures 15‐18. These all illustrate very significant demographic differences that clearly have implications for service delivery.

Space here does not permit a fuller exploration of the issues around book discovery: we will simply note that they are complex and beg further examination. The key message is that service planning and delivery might well benefit from a better understanding of how people find books and that publishers and librarians might do well to segment their offerings in a much more sophisticated way: “one‐size‐fits‐all” solutions would not seem to be a good idea at all looking at the evidence here.

Satisfaction with current provision of printed library books

At some point, every library user experiences difficulty in obtaining copies of conventional printed books and so the purpose of the next question is to determine the wider institutional context within which to understand the take‐up or neglect of e‐books (see Figure 19).

Overall, satisfaction levels among the UCL community regarding library provision are very high: 41 per cent of those expressing an opinion find provision to be “excellent” or “good”, only 8 per cent “poor” or “very poor”. But it is also clear that satisfaction and dissatisfaction are highly granular, with significant differences by gender[20] (men are generally less satisfied), age[21] (satisfaction increases with age), broad subject[22] (there is greater dissatisfaction in engineering and the sciences), and status[23] (part‐timers are less satisfied). The next two Figures (20‐21) illustrate just some of the complex interactions between these demographics again to make the point that we should not treat headline findings as necessarily being generally applicable across the board.

These figures highlight particular local hotspots of low satisfaction: for example, men in engineering and in the mathematical and physical sciences (Figure 20, upper left) and part‐time males (Figure 21, upper left) that we might otherwise overlook.

Awareness of e‐book library provision

The next question explored levels of awareness of the (then current) e‐book provision made by UCL Library Services (Figure 22).

Students, especially at Masters’ (41 per cent) or undergraduate (34 per cent) level are much more aware of e‐library book provision than are academic (24 per cent) and research staff (21 per cent). Amongst the faculties (faculty and students), engineering sciences (41 per cent) and social and historical sciences (38 per cent) are the most e‐book aware, life sciences (22 per cent) and mathematical and physical sciences (20 per cent) the least. Surprisingly perhaps there are no significant differences between full‐ and part‐time members of the UCL community.

Library e‐book awareness channels

For those faculty and students that were aware of library provision, a subsidiary question revolved around the channels by which they had found out about those e‐books in the first place (Figure 23).

The main channels forming initial awareness of UCL Library Service e‐books were the web site and library catalogue. This is especially so for men: staff briefings and course tutors were more effective awareness‐raising channels for women. Course tutors play a vital role: 68 per cent of undergraduates said that they found out about UCL provision this way.

Library e‐book marketing preferences

Regardless of whether they had any e‐book experience, or whether or not they knew about local UCL library provision, all respondents were asked at this point what they felt would be the single most effective way to get the message out about e‐library books (Figure 24).

Overall, the most effective marketing channels for e‐books in this context are likely to be information on the library web site and e‐mail user guides, but more precise targeting for different groups may well pay dividends. Within this broad picture, the survey findings suggest that staff would very much welcome a user guide posted on the library web site, that undergraduates would benefit enormously from making sure that e‐books are included and signposted on reading lists, and that graduate students are able to access a printed information guide in the library.