A Statistical Analysis of Library Use at the University of Chicago: A Case Study of Nerd Power!
Dr. Abbott is clearly a smart man but somehow he was suckered into doing an indepth statistical analysis of library use. Some of the results are not shocking. Most students never take out a book from the library while a small fraction of nerds check out countless books. Faculty use the library by day and undergraduates use it by night. Abbott refers to this as "temporal zoning".
He reports one interesting result that Internet use and time spent at the library are complements. This is similar to Ed Glaeser's work that has argued that information technology increases the demand for living in big cities. Once you make new contacts face to face, you can use IT to stay in touch with them.
http://magazine.uchicago.edu/0610/features/reg-print.shtml
Heavy connections
The more today’s Chicago students use electronic research materials, the more they do research the old-fashioned way.
By Andrew Abbott, AM’75, PhD’82
If you have been wondering for years whether it’s true that students make love in the stacks of Regenstein, I’m sad to say that I can’t tell you the answer. But I can tell you that more students take a nap in Regenstein than ask a librarian a question. I can tell you that almost a fifth of students eat food in Regenstein outside the café on most of their visits. And much more important, I can tell you that the electronic library revolution seems—paradoxically—to be raising rather than lowering usage of the library’s physical materials. All these are results of the 2005 survey of library usage conducted by the Provost’s Task Force on the University Library.
The task force, convened by Provost Richard Saller in April 2005, conducted the survey as part of its effort to understand the current state of the University’s libraries. Our aim in conducting a survey was not to plan the future by projecting current trends; our vision for the library’s future must rest on a theory of library research, not on rationalizations of current behavior. But Regenstein’s double function as both a major research library and a study hall for undergraduates and early-career graduate students means that any plans for the building must understand the behaviors of the nonresearch users as well as of the researchers.
Going into the survey, our thinking about student library use was guided by vague opinions and precise but hard-to-interpret utilization figures. Faculty opinion sensed a move toward informality. Many found Regenstein noisier, less welcoming. Some thought it a “student union.” Turnstile data revealed some striking shifts. Over the four-year period 2000–04, undergraduate entries to Regenstein were up 40 percent. To be sure, part of the rise reflected the College’s expansion and the building of a huge dormitory next door, but even controlled for demographic expansion, the undergraduate entry rate was up nearly a third. In contrast, undergraduate circulation rates had declined slightly. Graduate entry and circulation figures revealed no noticeable trends.
There is little detailed information on longer-term trends. We know that around 2,500 patrons per day used the library in the mid-1970s and that roughly the same number of materials were checked out then as now. Because there were half as many students, rates of use were obviously far higher in the mid-1970s. But by the mid-1990s—an era for which there is detailed information—use had fallen dramatically. One main culprit was the personal computer, whose immobility in the pre-laptop, pre-wireless era meant users tended to take research materials out (to the home or office where the machines were) rather than use them in the library. The other culprit was the coursepack (and later, electronic reserve), which reduced the reserve use that provided almost half of library circulation in the 1970s.
By 2005 it was apparent that usage of all kinds had picked up again. To analyze that usage, in spring 2005 the task force commissioned a Web-based survey, which was developed and administered by the University Survey Lab. The sample of 14,000 included all registered students across the entire University. Overall response was 42 percent—and well over 50 percent in the groups which turnstile evidence tells us are the library’s major users: students in the College and graduate students in the Humanities and Social Sciences. But there were plenty of nonusers in the data; nearly a quarter of respondents, for example, did not take a single book out of the library last year. Thus, the data seem to provide a reasonable representation of student library use and opinions.
Of the 5,700 respondents, about two-thirds named Regenstein as their most used library. Another 10 percent favored Crerar, 10 percent claimed no favorite, and the rest were scattered around the system. Of the 90 percent who named favorites, 15 to 20 percent said they had spent less than ten hours in their “most used” library during the entire spring quarter. In other words, about a quarter of all respondents had no real connection with a University library as a physical entity, although they may have used library-licensed electronic sources remotely. Indeed, only about 1,200 respondents said that their primary study space was Regenstein, as opposed to, say, another library or a coffee shop or their residence. Extrapolation suggests that the “primary study space” clientele for Regenstein is between 1,500 and 2,000 students—or considerably less than half of the University’s 4,500 undergraduates.
Under the admittedly lax ten-hour criterion, about 2,600 respondents called Regenstein their most used library, and nearly all the findings below are based on their answers. We asked dozens of questions about their usage of the library, as well as about their desiderata, their demographics, and so on. The main usage questions included 41 different types of activities. Participants were asked if they did these things never, sometimes, on about half their visits, on most of their visits, or on all visits.
________________________________________
To the extent that Regenstein is a student union, it is so at a time that doesn’t interfere with its research mission.
________________________________________
After much preliminary statistical analysis we grouped this information into eight major scales. Five were direct measures of library usage. A traditional research scale combined activities like checking out a book, browsing the shelves, using printed library material without checking it out, or using the online catalog. A second scale charted the use of online databases, while a third pulled together use of online resources like the Research Libraries Group catalog, Worldcat, bibliographic search engines, subject guides, and online reference works. A fourth scale tapped use of specialized resources: archives, special collections, microforms, and CD-ROM databases. The fifth measure looked at circulation, taking data directly from library records.
Because research is not the only use of the library, these scales were complemented by three others charting more mundane matters. One was a purely social scale: items about meeting new people and hanging out on A-level, the undergraduate study space that is a major campus social center after 10 p.m. Another scale comprised items about use of library computers for assignments. Finally, there was a scale originally imagined as an indicator of the library’s status as a student union. Students were asked if they used the library to take a study break, make a cell phone call, answer a cell phone call, arrange to meet a friend, bump into a friend and chat, use e-mail, eat food brought into the library, eat food bought in the library coffee shop, surf the Web, or shop on the Web. It was, in fact, a laundry list of things that many faculty considered antithetical to serious use of the library.
This “student-union” scale produced the first surprise. The task force had expected an inverse relationship between the student-union scale and the traditional research scale. But there was no relation whatever, at the individual level, between being high or low on the traditional research scale and being high or low on the student union scale. The scales measure different things. The first measures traditional research practices; the other measures what for Chicago’s current students—both graduate and undergraduate—is a particular way of doing everyday life. Choosing to arrange your everyday life in this technology-based, somewhat sociable way is unrelated to whether or not you are a strong or weak user of traditional research practices.
It followed that our model of “student union versus research facility” was also wrong. While there is a negative relation between the purely social scale—the A-level scale—and the traditional research scale, the library’s A-level, social side is essentially an evening affair, when, as the turnstile data demonstrate, faculty and even graduate students are simply not around. To the extent that Regenstein is a student union, it is so at a time that doesn’t interfere with its research mission. A kind of “temporal zoning” keeps the two uses out of each other’s way.
There was a distinct difference between research and study use. Using library computers for assignments was much more characteristic of undergraduates than of graduate students. Undergraduates were also much more likely to bring materials of their own (books, notes, etc.) into the library. By contrast, graduate students were much higher on the traditional research scales. These three facts made a single consistent point about research versus study.
One hesitates to underscore this contrast, for fear of showing once again that sociology is the science of demonstrating the obvious. But we should recall the old joke about the mathematician who spends two days proving that he was correct in his offhand statement that some step in a proof was intuitively obvious: it is better to be safe than wrong. So it is important that we can say on the basis of real evidence, not hearsay, that undergraduates predominantly—almost overwhelmingly—use Regenstein as a study hall, bringing material there and doing assignments there. Graduate students use it predominantly as a research site. This is not an absolute difference by any means, but it is a strong theme throughout the data.
The pattern is driven to a great extent by the College curriculum. Both the Core and much upper-level undergraduate teaching emphasize careful analysis of a small number of selected texts, usually purchased. Indeed, when looked at by College year, all of the research scales—traditional research, both electronic scales, circulation, and even special resources—jump up slightly from first- to second-year (as students move out of the full-time Core) and take a somewhat more substantial jump up from third- to fourth-year as the honors students write BA papers and finally have to dig into the library seriously.
Although early-stage graduate students echo the undergraduate pattern of bringing their own materials and using Regenstein as a study hall, most graduate student use is research: graduate students were a full point higher on the traditional research scale. This means they averaged a point higher on each of the scale’s eight items, where “a point higher” means the difference between “sometimes” and “about half the time” or between “usually” and “always.” It’s a big difference—one that holds up for circulation levels and for use of specialized resources. More important, the graduate students’ edge also held up for both online-database use and electronic-research use. Even though the undergraduates were much higher on electronic-database use than on any of the other electronic measures, the graduate students were higher still.
This means that there was no evidence that younger people were somehow “more electronic.” Graduate student respondents, after all, averaged eight to ten years older than College respondents. Yet they made more use of electronic resources. And, as we saw earlier, they were also equally likely to conduct everyday life electronically.
The research-use data provide the surprising conclusion to this syllogism. The more an individual uses books, the more he or she uses electronic-research resources, and vice versa. The finding holds not only at the group level—graduate students are higher on both scales—but also at the individual level. This seems like very strong evidence in favor of a synergy hypothesis—that use of one of type of material is likely to reinforce the use of the other. At the very least, the survey data provides no evidence that traditional research practices are being replaced by electronic ones. Thus, the replacement hypothesis—a standard idea in library literature—must be rejected.
Hits on the library’s licensed electronic databases do show the one grain of truth in the mountain of the replacement falsehood. The main electronic workhorses—receiving the majority of total hits—are JSTOR and Elsevier Science Direct, which are merely delivery systems for journals. What’s being replaced is one type of material (journals) and that only in the final delivery stage: now people find journal articles online and print them directly. We know that users make hard copies, by the way, because more than half of all JSTOR hits result in a printing. At least in the humanities and social sciences, neither faculty nor students seem to read articles on screen.
THE DISCUSSION SO FAR CONCERNS LARGE CATEGORIES OF USERS—undergraduates, graduate students in this or that division. But there is much variation within these categories. It is clear from the survey—as from circulation data—that research use of the library is heavily skewed. Asked by the provost to facilitate the research enterprise, the task force had to ask in turn, Who, really, are the constituents of that enterprise, the heavy users of the library?
Gross circulation figures (from 2003–04) give some sense of who these heavy users are. Out of some 33,000 cardholders, about 13,600 “users” actually took out at least one book. About ten percent of those users provided 50 percent of all circulation. At the other end of the scale, the lowest 50 percent of users provided only 10 percent of the circulation. (Remember, this definition ignores cardholders who didn’t take out a book—a group that includes one quarter of Chicago undergraduates in a given year.) Basically, the library’s heavy users in circulation terms are the approximately 1,000 people (7 percent of the total users, 3 percent of all cardholders) who take out more than 100 books a year, providing close to 40 percent of all circulation. This group included about 80 faculty, about 500 PhD-level graduate students, about 100 MA-level graduate students, and about 140 undergraduates.
But the survey offered a more detailed way to characterize the students among these heavy users. We took our “never” to “always” scales, lumped together the top two categories—“usually” and “always”—and called that heavy use. We lumped the bottom two categories—“never” and “sometimes”—and called that light use. We left the middle as medium use. For circulation, we called 100-plus heavy use and 50-plus medium use. This created low, medium, and high ratings for each of our five research scales—traditional research, circulation, other resources, electronic databases, and general electronic sources. People with high levels on more than one scale were our student “heavy users.”
Believe it or not, 13 student respondents ranked high on all five scales. Another 56 rated high on four out of five; 163 on three of five, and 330 on two of five. This makes a total of 562 “heavy users” on two or more scales. In rough terms, Divinity School graduate students are 7 percent of these, social-sciences graduate students about 30 percent, humanities graduate students just under 50 percent, and undergraduates about 12 percent. Thus, although their rate of heavy-research use is quite low, the sheer numbers of undergraduates mean that they still make a significant contribution to the library’s heavy-user population. As a result, the library staff has to serve three very different types of heavy user—faculty, graduate students, and undergraduates.
As we saw earlier, these heavy users pursue their work in a library that is mostly filled with students using Regenstein as a study hall. But observational studies sponsored by the task force make it clear that study and research use live together in surprising harmony. Even at its heaviest use—just before exams when a third or more of the seats in Regenstein are occupied—the building is remarkably quiet and orderly, especially as one gets further away from the first-floor entry space. And fortunately, many of the things that can be done to improve Regenstein’s utility for its heavy users can be done without impairing its utility as a study hall for the perhaps 1,500 undergraduate and early graduate nonheavy users for whom it is their principal place to do assignments.
The task force went into its study knowing about the differences in graduate and undergraduate use and the largely graduate nature of the heavy-user community. But the noncorrelation of a student’s electronic everyday life with his or her research practices surprised most of us, and the powerful positive correlation between electronic- and traditional-research practices was quite unexpected. Such findings underscore the need to plan the library’s future not by extrapolating trends or imagining a speculative techno-utopian facility, but by thinking long and hard about the library-research process, the library-research community, and the specific ways in which both can be helped by a transformation of the library that has served the University so well for so long.
Andrew Abbott, AM’75, PhD’82, the Gustavus F. and Ann M. Swift distinguished service professor in sociology and the College, chaired the Provost’s Task Force on the University Library. His mother was a librarian.