Academic libraries provide access to scholarly articles, books and other materials to support the work of academic staff and students, yet academics today also have many alternative ways to access scholarly information. Measuring the use, outcomes and value of the library's role in scholarly reading helps librarians make decisions about the future direction of their collections and services, and helps demonstrate the value library collections bring to scholarship.
The Scholarly Reading and the Value of Library Resources project1 measured the value and outcomes to academic staff members from access to scholarly publications in six universities in the United Kingdom. It sought to answer questions such as: What is the value and outcome of scholarly reading for academic staff? How do academic library collections support research and teaching activities of academic staff? How do reading patterns of articles, books and other materials differ? What is the role of the academic library collections in teaching and learning? Six higher education institutions – Cranfield University, Durham University, Imperial College London, University of Dundee, University of East Anglia and the University of Manchester – participated in the 2011 survey of scholarly reading by their academic staff members. (See Appendix for survey questions.) JISC Collections funded this study led by the Center for Information and Communication Studies at the University of Tennessee.
This article summarizes the results of the combined findings, with a particular emphasis on those findings that focus on the value of the reading that comes from the library collections.
The study builds on reading surveys conducted by Carol Tenopir and Donald W King in the United States since 19772 and in Australia and Finland in 2005 and 2006.3 Together, these studies track scholarly reading patterns and the continued contribution of academic library collections to the academic enterprise.
Tenopir and King4 and King and Tenopir5 summarize reading patterns of academic staff members through the 1990s. These two sources provide extensive literature reviews and serve as background for the data presented in this report. Multi-university studies by others focus on how academic staff use electronic journals, online resources and libraries.6 Further studies show that staff members in the sciences prefer and read more electronic journal articles than in humanities or social science disciplines.7 Access and convenience, especially electronic access, are important to all academic staff.8 A 2011 study by the Research Information Network (RIN) found a link between the library and the institution's research performance.9 For the first time, the 2011 survey includes sections on reading from books and other publications and questions on the use of social media.
Earlier surveys examined just the reading of scholarly articles. This survey includes those questions, but expands the survey to look at reading books and book chapters and other scholarly materials, including conference proceedings, government documents, and other material from websites used for work-related reading. The questions include both reader-related (demographic) and reading-related questions. Reader-related questions include age, gender, percentage of work time spent on various activities, number of personal subscriptions and two measures of recent academic success: publication record and record of recent awards.
The reading-related questions are mostly based on the ‘critical incident technique’ first developed by Flanagan,10 which has since been applied to many contexts, including libraries and readings.11,12 We focus on the last scholarly reading as the ‘critical’ incident of reading.13 This allows us to ask questions about a specific most recent reading, so the respondent will have a better memory of that reading, rather than having to reflect back on multiple readings over a longer period of time. The questions cover many details of that reading, including time spent on the reading, source of reading, purpose of reading, value of the reading to the purpose and outcomes of the reading.
Starting in March 2011, the head librarian or library representative at the six UK universities sent an e-mail message with an embedded link to the survey instrument housed on the University of Tennessee's server. We received 2,117 responses by the official closing date of 9 June 2011. Approximately 12,600 invitations were distributed in total, providing an overall estimated response rate of 16.8%. Of the 1,102 respondents who chose to give their academic discipline, 34% are in sciences, 24% are in social sciences, 16% are in engineering/technology, 13% are in medical/health, 11% are in humanities and 2% are ‘other’. The ages of respondents varied from under 30 to over 60, with 18% under 30 and 7% over 60.
Since respondents were allowed to leave the survey at any time, skip questions, or were timed out automatically if they began the questionnaire and did not complete it, most of the questions have a lower number of responses than the total of 2,117 who answered at least one question. All respondents for a particular question equal 100% for that question.
Since surveys rely on self-reporting, the main limitation of this method is that we assume that respondents are replying accurately and the exact numbers should be viewed as approximations. Relatively low response rates may mean that those academics who do not read as much did not reply. The academics surveyed here, and in earlier surveys, are all affiliated with universities that have robust libraries with electronic and print collections available to all of their affiliated academic staff.
Results and discussion
Total amount of reading per academic staff member
Academic staff report reading from a variety of materials, but they read the most articles. Respondents read from an average of 39 scholarly readings per month, comprising 22 articles, seven books, and ten other publications. On this basis an annual total could be estimated at 468 readings per year (i.e. 39 X 12). Academics who spend more time per book reading also spend more time per article and other publication reading. A significant relationship between the number of article readings and the number of other publication readings also exists. If a respondent reads a lot, they do not focus on one type, but read from of all types of material. For example, academic staff members who read more articles than the average also read more books and other publications.
“Successful academics read more on average.”
A significant relationship exists between successful academics, defined as those who publish more and earned an award in the past two years, and the number of article, book and other publication readings. Successful academics read more on average. While we cannot draw a cause and effect relationship, these relationships show that scholarly reading is an important part of the work activities of successful academics.
When the average time spent per reading is multiplied by the average number of articles, books and other publications read each month, the average academic staff member spends 18 hours reading articles, 12 hours reading books and seven hours reading other publications (outliers excluded). The estimated annual total commitment to scholarly reading is at least 448 hours per academic, or the equivalent of 56 eight-hour days each year, illustrating the huge time commitment that the average academic staff member invests in scholarly reading each year.
Respondents who spend more time on per book reading also spend more time per other publication reading and article reading. Readers from every discipline report more article readings than from other types of scholarly material. Humanities respondents report the most readings – on average 58 scholarly readings. Scientists read many articles but overall report the fewest total readings, an average of 38 readings (Table 1).
|Article readings||Book readings||Other publication readings||Total readings|
How academic staff become aware and obtain scholarly material
Academics find more articles through searching, while books are often found through word-of-mouth. Citations are a useful tool for discovering articles and books (22% and 11% respectively). Overall, UK academic staff members use a variety of methods to become aware of articles and books (Table 2).
Excluding outliers, respondents spend on average 16 minutes browsing or searching for articles, 11 minutes for books, and nine minutes for other publications. Articles, on average, may require more time to find because there are so many published monthly and large databases include millions of articles. While academics appreciate the ease and accessibility of online access to articles, the sheer amount of information may be requiring more time to become aware of relevant material. One respondent states, “There is far too much information out there and very little time to screen through and read articles”. Respondents are also more likely to previously own or find out about books through another person, which may decrease the amount of time they spend.
|Found while browsing||134||11.2||61||6.5|
|Found while searching||392||32.9||218||23.4|
|Found through citation in another publication||211||17.7||106||11.4|
|Another person told me about it||205||17.2||262||28.1|
|Promotional e-mail or web advert||43||3.6||34||3.6|
|Don't know or don't remember||34||2.9||54||5.8|
|Already owned or knew||–||N/A||73||7.8|
|For review or was a contributor/co-author||–||N/A||38||4.1|
Academics use different sources to obtain the different materials; articles are predominately found through the library, while books are often purchased and other publications are frequently obtained from either a free copy from a publisher or website (Table 3). While respondents are using the library to obtain the article, they are usually interacting with an electronic version (93.5%) and are rarely reading in the physical location (1.7%).
|Article %||Book %||Other Publication %|
|I bought it for myself/personal subscription||4.7||39.1||14.9|
|The library or archives collection||65.2||25.7||13.8|
|Inter-library loan or document delivery service||2.1||1.9||1.4|
|School or department collection||4.5||2.6||2.7|
|A colleague, author or other person provided it to me||5.6||10.5||13.0|
|A free, advance, or purchased copy from the publisher||–||14.0||17.8|
|Website/free web journal||13.4||N/A||26.6|
Academics spend more time on receiving, printing and other tasks involved in obtaining books than they do for articles or other publications. On average, they spend approximately six minutes to obtain an article or a publication, and 14 minutes to obtain a book. While approximately 5% of respondents spend over 30 minutes to obtain a book (not including printing, photocopying or other task), less than 1% spend over 30 minutes to obtain an article or other publication.
“… articles are more often accessed through the library's electronic services, while books and other publications typically are obtained from other sources.”
One explanation of why books and other publications, on average, take longer to obtain than articles once the reader is aware of the item is because articles are more often accessed through the library's electronic services, while books and other publications typically are obtained from other sources. The UK academic libraries' e-journal collections are saving the time of the reader in obtaining articles, and as one respondent comments, “without wide electronic access through library subscriptions many aspects of my work would either take substantially longer or be done to a lower standard”, and another respondent says, “subscriptions to electronic journals improves my research and makes it faster and easier for me to perform my job effectively”.
We did not explicitly ask if the book reading was from an electronic source, but from the responses to how they found out about and obtained the book reading, we assume the majority of book readings are done from physical copies. As a result, respondents must spend time getting to the book's physical location, whether a bookstore, library, or a colleague's office, requiring more time than downloading an article or government document from a website or electronic journal while sitting in an office or at home.
Overall, respondents spend more time reading from books than articles or other publications. Respondents, however, are reading more articles and as a result spend more time on total article readings than book or other publication readings (Table 4). Earlier we said academics spend 448 hours per year on scholarly reading, but their total commitment goes beyond the time spent reading to also include the time spent finding and obtaining the material. This illustrates an even greater commitment of time invested in scholarly reading. The figures in Table 4 are one measure of the ‘exchange value’ placed on scholarly reading. The average academic staff member spends over 300 hours each year on journal article readings, 180 hours on book readings, and around 120 hours on other publication readings, for a total commitment of 76 eight-hour days each year. If scholarly reading was not a valuable and useful aspect of their work, they would not spend as much time using articles, books and other publications.
|Total time(in minutes) per material||71||131||57|
|No. read per month||22||7||10|
|Total time reading per month||26 hours||15 hours||9 hours and 30 minutes|
|Total time reading per year||39 eight-hour work days||22 eight-hour work days||15 eight-hour work days|
Principal purpose of scholarly reading
Research and writing is the most frequent purpose for reading for all types of scholarly materials (Table 5). Teaching is the second most frequent purpose for articles and books. Other publications are more frequently read for current awareness/keeping up than are books and articles.
|Article % (n=1161)||Book % (n=921)||Other % (n=731)|
|Research and writing||74.3||57.6||45.2|
|Current awareness/keeping up||5.9||2.5||28.3|
|Internal or external presentations||1.6||1.1||1.1|
|Continuing education for self||2.5||5.2||5.2|
|Engagement activities (to wider community)||0.3||0.8||1.2|
|Knowledge transfer or enterprise activities||N/A||0.4||0.5|
Value of scholarly reading
Book readings were rated as ‘very important’ to ‘absolutely essential’ (49.6%), while other publications have the lowest importance ranking (Table 6). Approximately 45% of other publication readings are ‘somewhat’ or ‘not at all’ important. Only a few (approximately 1%) of article and book readings are considered not at all important.
|Article (n=1160)||Book (n=918)||Other (n=729)|
|Not at all important||1.1||0.9||5.3|
We see a wide range of outcomes from scholarly readings, and less than 1% of article and book readings and 3% of other publication readings are considered a waste of time (Table 7). Inspiring new thinking and improving the result are the two most frequent outcomes.
|Article% (n=934)||Book% (n=935)||Other% (n=749)|
|Inspired new thinking||53.7||44.7||39.1|
|Improved the result||37.8||47.7||33.8|
|Narrowed/broadened/changed the focus||28.1||26.5||21.6|
|Saved time or resources||9.6||14.5||10.9|
|Resolved technical problems||9.9||21.8||10.1|
|Resulted in faster completion||4.6||13.5||7.1|
|Resulted in collaboration/joint research||4.4||3.1||4.0|
The role of library collections
How academic staff members obtain scholarly reading material can be assigned to three basic categories: library-provided, personal source (e.g. personal subscription or purchase) and others. Most scholarly article readings are obtained from the library (67.3%), a finding that is consistent with previous studies.14 In the UK, less than 5% of article readings come from personal subscriptions (Table 8), which is similar to the 2005 US findings. Unlike article readings, the majority of book readings are obtained through personal copies (39.1%), and only 27.6% are obtained from a library. Other publications are more likely to be obtained from another source, including other online sources (27%) or another person (31%).
A 2011 study by the Research Information Network found a relationship between the institution's library and its research performance.15 The RIN study concludes that easy access to high-quality content is a key foundation for good research, and when the library works in partnership with researchers it enables better library services and creates top researchers. We found a similar association between the library's resources and its support of research. Approximately three-quarters (76%) of the articles obtained through the library are principally for research and writing (Table 9). Just over half (57%) of article readings from a personal subscription are for research and writing (29 of 54). Seventy-two percent of article readings obtained from other sources are read for research and writing. Approximately 73% of book readings and 68% of other publication readings obtained from the library are primarily for research and writing.
|Personal subscription||Library provided||Others||Row Total|
|Principal Purpose||Research & writing||31||600||230||861|
Article readings from personal subscriptions are more likely to be of recent publications. Eighty-three percent of articles obtained through a personal subscription are published in the past year (2010–2011). Over half of the library-provided and other source articles are 18 months old or older (55.4% and 53.3% respectively). No articles obtained through personal subscription are over ten years old (Table 10). Regardless of the age of the publication, library-provided article readings are primarily from an electronic subscription. Eighty-six percent of pre-1996 library-provided articles are obtained from an electronic library subscription (76 of 88), and 95% of the library-provided articles in their first year of publication are from an electronic library subscription (315 of 331).
|Personal subscription||Library provided||Others||Row Total|
One way to represent the value of the library for scholarly work and research is to formulate how many hours per year each academic staff member dedicates to library-provided reading. Based on past methodology that creates a formula to measure academic staff output based on library input,16 we measured the library's value by the time spent using library reading material, assuming that scholarly readings are important for quality research, teaching and other work activities. By using a simple formula of time spent reading each material multiplied by the number of each material read per month multiplied by 12 to calculate an annual total, we can illustrate the total amount of reading by each academic staff member. We then multiply the total amount by the percent obtained from the library to determine the number of hours per year each staff member devotes to library-based work (Table 11). Most time is spent on library-provided article readings, approximately 144 hours each year. Approximately 40 hours is spent on library-provided book readings and 13 hours on library- provided other publication readings each year. Annually, academic staff spend 197 hours of their work time with library-provided material, or the equivalent of 25 eight-hour days.
|Minutes per reading||Number read per month||Multiplied by 12 months||Percent from library||Total|
|Other publication||42||10||12||0.15||13 hours|
Academic staff members use multiple sources of information every day from scholarly articles, books and other publications for their teaching and research. As a result, time has become an increasingly important deciding factor for where to obtain desired material. Academic staff members spend a large portion of their work time on scholarly reading and reading has many positive outcomes on their research and teaching. The amount of time they spend on reading from the library's collections is evidence of the importance of library-provided scholarly materials to academic work.
“Annually, academic staff spend 197 hours of their work time with library-provided material, or the equivalent of 25 eight-hour days.”
A vast majority of article readings come from the library's e-journal collection, while personal copies are popular for print books and the open web for other materials. Convenience and ease of accessibility is likely the most important factor in making choices of where to get material to read. In order for the library to maintain its function as a central source of information at the university, it must strive to keep its collections as accessible and convenient as possible. Providing access through the library to e-resources is crucial to the research and teaching missions of the university, and collections which include back-files in addition to current subscriptions are a key investment. Scholarly library-provided e-books have not yet reached their potential and may drive changes in reading behavior if they are convenient and relevant to the academic staff.
Library usage data from COUNTER reports and SCONUL confirms that usage is high and the number of articles downloaded has steadily increased over time. The library may be providing more readings than people are aware of because often it is not possible to distinguish library-provided resources from free ones. There is often a perception that more content is ‘free on the web’ than is actually the case. Especially as the library's role in providing access to content becomes less visible, it is more difficult for users, funders and librarians to judge the library's value using only usage measures. Branding of the library's role in e-collections will help academics more fully realize the benefits of their academic library. The value academic reading has on the outcomes of the work of the university is apparent, and the university library, especially for article readings, is essential to the quality of the academic enterprise.