![[Board accepted]](http://www.istl.org/graphics/b2.gif){kind=small}
Comparing Safari Tech Books Online and Books24x7 E-book Collections: A Case Study from the University of British Columbia Library
Science and Engineering Librarian
eugene.barsky@ubc.ca
Lisa Schattman
MLIS Candidate
School of Library, Archival and Information Studies
Aleteia Greenwood
Head, Science and Engineering Library
University of British Columbia
Vancouver, British Columbia
Abstract
Most academic libraries are seeking to provide electronic access to the very dynamic and changing field of technology related material. Safari Tech Books Online and Books24x7 are the major e-book collections in this area. We compared the Safari Tech Books Online and Books24x7 e-book packages as to their usefulness for the University of British Columbia Library, second largest academic library in Canada. In our sample, we found that Books24x7 had more titles to offer (25% more); the overlap between the collections was relatively small (13-16%) and publishers varied considerably; and although there were no major differences in the "usefulness" measures of the titles in the two packages (Amazon.com ratings, WorldCat holdings and Reviews.com reviews were quite similar for both packages), O'Reilly titles, available only through Safari Tech Books Online, were held by a slightly higher number of libraries worldwide. We conclude that in order to have comprehensive coverage of this constantly changing area of knowledge, a large research academic library needs to subscribe to both collections. If subscribing to both collections is not an option, we recommend selecting a package based on the pricing that the library can negotiate with ProQuest, the vendor of both products, since the collections are complimentary in their nature.
Introduction
With current financial markets in disorder and possible budget cuts to our libraries continent-wide, librarians may be forced to make tough decisions regarding their collections in the near future. In the area of Computer Science and other Science and Engineering disciplines, many academic libraries may feel a strong demand for fully searchable, full-text, online computer manuals and other technical reference books. At present, two major vendors offer comparable full-text, technical reference content: Safari Tech Books Online and Books24x7. While a number of articles discuss these e-book packages individually, there seems to be no research comparing the two (Dalzell 2007; Fernandes 2007; Golderman & Connolly 2004; Hawkins 2007; Kong 2006; LaGuardia 2004; Marien 2007; Stephens, Melgoza, & Gang Wan 2008; Turner 2003; Wallace 2006; Whang 2006). In an era of budget cuts and hiring freezes, we decided to evaluate these two packages of electronic technical reference books to see how they compare with each other and what their advantages and/or disadvantages might be.
Safari Tech Books Online, a joint venture between O'Reilly Media, Inc. and the Pearson Technology Group, is a collection of fully searchable, full-text computer manuals. The full Safari catalog contains 6,474 publications (as of December 19, 2008) from publishers including O'Reilly and Pearson imprints, Addison-Wesley Professional, Adobe Press, Cisco Press, New Riders, Peachpit Press, Prentice Hall PTR, Sun Microsystems Press, Que, and Sams. O'Reilly is the publisher of the popular "In a Nutshell," "Essentials," and "Definitive Guides" book series, identifiable by the animals on their covers. ProQuest is the distributor of Safari Tech Books Online to academic, public, and K-12 libraries.
A product of SkillSoft, Books24x7 makes available web-based technical and business reference content; it contains 7,501 digitized reference books (as of December 19, 2008). Like Safari Tech Books Online, the content is cover to cover, including full text, tables, sidebars, figures, drawings and formulas. Included titles come from McGraw-Hill, JohnWiley & Sons, Cambridge University Press, Sybex, Microsoft Press, Oxford University Press, MIT Press, IBM Redbooks, and Wrox Press. ProQuest is also the distributor of Books24x7 to libraries.
In comparing these two e-book packages, our goals were to determine for each collection:
- The number of titles covering our sample topics;
- The currency of those titles (based on publication years);
- The major publishers of those titles; and
- The "usefulness" of those titles, as measured by:
- The average number of libraries holding those titles (as listed in WorldCat),
- The number of positive reviews of those titles on Reviews.com (a computer science book reviews site from the Association for Computing Machinery), and
- The average rating for those titles on Amazon.com.
Methods
Unable to exhaustively compare Safari and Books24x7 title by title, we selected six topics that would best represent the interests of our user groups. We made our selection based on the reference questions received by the Computer Science liaison librarian (one of the authors) during the past year.
The topics selected were: Python and C# (programming languages), Dreamweaver and Maya (web and graphical design), cryptography and podcasting.
For each topic, we searched both Books24x7 and Safari for books about that topic, limiting the search to book titles only (i.e., book titles containing the word "python") and collected the following data:
- Number of titles,
- Distribution of publication years, and
- Publisher distribution.
We also wanted to evaluate the "usefulness" of those titles to general readers. For each title, we measured usefulness by looking at the number of libraries worldwide that owned that title (as reported in WorldCat), its average star ranking on Amazon.com, and whether it had been positively reviewed on Reviews.com, ACM's review site for computer science books. Our assumption was that if a title is held by more libraries, has a better ranking on Amazon.com, and is reviewed positively on Reviews.com, it will be more useful to our readers. Unable to spend the large amount of time required to compare the usefulness of all retrieved titles in our sample, we evaluated the titles for only two topics -- Python and Maya.
Results and Discussion
After compiling our results, we analyzed the 395 titles retrieved in our sample.
Number of titles: Our sample contained 220 books from Books24x7 and 175 books from Safari (see Appendix 1). Books24x7 had 25% more titles covering our chosen areas.
Years of publication: In our sample, Books24x7 offered more titles published in every year except 2003 and 2007; Safari had more books published in those two years for our selected topics (see Appendix 1).
Publishers: Although we had learned from the existing literature that the publishers for the two packages would vary, the results still surprised us. As anticipated, the major publisher in Safari was O'Reilly with 38 books (22% of the Safari titles in our sample). Other significant publishers were Sams with 26 books (15%), Addison Wesley and Microsoft with 17 books each (10%), and Peachpit and Prentice Hall with 16 books each (9%). In Books24x7, the major publishers were Apress with 52 books (24% of the Books24x7 titles in our sample), Wrox Press with 28 books (13%), Microsoft with 25 books (11%) and John Wiley & Sons with 21 books (9%). Moreover, publishers that were well represented in one package tended to be poorly represented in the other: there were 52 Apress titles in Books24x7 but only 7 in Safari; there were 21 John Wiley & Sons books in Books24x7 but only four in Safari; there were 38 O'Reilly books in Safari but none in Books 24x7; there were 26 Sams books in Safari and only two in Books 24x7. Finally, we found that Books24x7 represented a much larger variety of publishers (e.g., Cambridge University, Charles River Media, Friends of Ed, McGraw-Hill, Paraglyph, Sybex, Visibooks, etc.; see Appendix 2).
Overlap: Also of interest was the number of titles appearing in both Books24x7 and Safari. By comparing books on each of the chosen topics, we found the following overlap:
- Python: two titles in common, representing 9% of the Safari Python collection and 12% of the Books24x7 Python collection.
- C#: 22 titles in common, representing 26% of the Safari C# collection and 17% of the Books24x7 C# collection.
- Dreamweaver: three titles in common, representing 7% of the Safari Dreamweaver collection and also 7% of the Books24x7 Dreamweaver collection.
- Cryptography: one title in common, representing 25% of the Safari Cryptography collection and 6% of the Books24x7 Cryptography collection.
The total overlap was 28 titles in common, representing only 16% of the Safari collection and 13% of the Books24x7 collection. This was much less duplication than anticipated. Microsoft Press titles accounted for most of the overlap; if we eliminate Microsoft titles, the overlap would represent a mere 7% of the Safari collection and 5.5% of the Books24x7 collection. This was one of our most important findings: since the content overlap between the two collections is insignificant, the e-book packages are complementary, rather than duplicating one another.
Usefulness - WorldCat holdings: As one indicator of usefulness, we examined the number of libraries worldwide that owned each title in our sample, as reported in WorldCat. On average, each Safari title was owned by 203 libraries worldwide and each Books24x7 title by 194 libraries worldwide. This is a relatively minor difference (see Appendix 3).
Usefulness - Reviews.com ratings: As another indicator of usefulness, we investigated the number of titles in our sample that were positively reviewed on the ACM's computer science book reviews site. Six titles from each package were reviewed positively on Reviews.com.
Usefulness -- Amazon.com ratings: As a third indicator of usefulness, we examined user reviews on Amazon.com -- the major online bookseller -- and compared the rankings of each title in our sample. On average, Safari titles achieved a rating of 3.8 (out of 5) versus 3.45 for Books24x7 titles, not a sweeping difference.
Usefulness -- O'Reilly titles: Our user groups report that many people prefer O'Reilly books to those from other publishers. Using the three measures listed above, we compared the usefulness of the O'Reilly titles in our sample to the overall sample. We found that the O'Reilly titles in our sample were, on average, held by more libraries than the titles from Books24x7 or even the Safari titles as a whole -- an average of 280 libraries for O'Reilly titles, compared to 203 for Safari titles and 194 for Books24x7 titles (see appendix 4). The average Amazon.com rating for O'Reilly titles is similar to that for the Safari titles overall – 3.8, which is just a bit higher than the Books24x7 average rating of 3.45 (see Appendix 4).
Other Considerations: Both e-book collections provide MARC records for their content. The technical staff here at the UBC Library regards the Books24x7 MARC records very highly, reporting that Books24x7 is one of the most user-friendly and automated online book packages. Our library system has no experience with Safari MARC records.
Both e-book packages offer HTML viewing, displaying one page at a time with a table of contents link on each page. HTML pages can be printed and the URLs can be e-mailed or bookmarked. It is unfortunate that neither collection offers PDF versions of the books' text, the option that most of our users seem to prefer for viewing and printing.
Another important usability consideration is how users access the database. Safari authenticates users based on their IP addresses, which allows seamless access for patrons both on our campus and connected to the network through Virtual Private Network (VPN). Books24x7 requires users to create individual accounts and log in to view an e-book. Although UBC patrons can set up free accounts relatively quickly and easily, this extra step dissuades some users from taking advantage of this resource. It also impedes easy access should users discover Books24x7 content through any means other than the database itself.
Conclusions
Most academic libraries are seeking to provide electronic access to the very dynamic and changing field of technology related material. Safari and Books24x7 are the major e-content providers in this area, together with NetLibrary. Stephens et al. (2008) found that the most recent editions of electronic e-books are available in Safari (we found no comparable research for the Books24x7 collection), which helps to keep this rapidly changing field up to date. Our study found that the overlap between the collections is relatively small (13-16%); publishers vary considerably; and there are no major differences in the usefulness of those two packages: Amazon.com ratings, WorldCat holdings and Reviews.com book reviews were quite similar for both. Therefore, we conclude that in order to have comprehensive coverage of this constantly changing area of knowledge, a large academic research library needs to subscribe to both collections. If subscribing to both collections is not feasible due to budget constraints, as is the case at the University of British Columbia Library, we recommend selecting a package based on the pricing that the library can negotiate with ProQuest, the vendor of both products, since the collections are complementary in their nature.
Acknowledgments:
The authors would like to acknowledge Dean Giustini for reading and commenting on the preliminary version of the article.
References
Dalzell, C. M. 2007. Safari Tech Books Online. Choice: Current Reviews for Academic Libraries. 44(8): 1301-1301.
Fernandes, D. 2007. The Safari e-book route through the ICT jungle: experiences at Hillingdon libraries. Program: Electronic Library & Information Systems. 41(3): 227-238.
Golderman, G., & Connolly, B. 2004. Safari Tech Books Online. Library Journal. 129: 26-27.
Hawkins, D. T. 2007. Enter Safari Books Online. Information Today. 24(8): 40.
Kong, L. 2006. Books24x7. Choice: Current Reviews for Academic Libraries. 43(6): 979-980.
LaGuardia, C. 2004. Book searching made easy. Library Journal. 129(18): 32.
Marien, S. A. 2007. Safari Business Books Online. Choice: Current Reviews for Academic Libraries. 45(3): 453-454.
Stephens, J., Melgoza, P., & Gang Wan. 2008. Safari Books Online: currency, usage and book release policies of an e-book database. Collection Building. 27(1): 14-17.
Turner, S. 2003. Safari Tech Books Online: a valuable resource. Mississippi Libraries. 67(2): 48-50.
Wallace, R. 2006. Safari Tech Books Online as supplementary reserve materials. Science & Technology Libraries. 26(3): 7-11.
Whang, L. 2006. Safari Tech Books Online. Issues in Science & Technology Librarianship. [Online]. Available: http://www.istl.org/06-summer/electronic1.html [Accessed 1/14/09].
Appendix 1
Years of Publication
| Year of publication | Safari | Books 24X7 |
|---|---|---|
2008 | 16 | 17 |
2007 | 32 | 14 |
2006 | 22 | 40 |
2005 | 16 | 21 |
2004 | 12 | 28 |
2003 | 34 | 29 |
2002 | 22 | 43 |
2001 | 12 | 17 |
2000 | 6 | 9 |
1999 | 1 | 2 |
1998 | 2 | 0 |
total | 175 | 220 |
Appendix 2
Publishers
| Publishers | Safari | Books 24X7 |
|---|---|---|
A K Peters | 0 | 1 |
Addison Wesley | 17 | 0 |
Adobe | 3 | 0 |
A-LIST | 0 | 1 |
Apress | 7 | 52 |
Artech | 0 | 2 |
ASTD | 0 | 1 |
Cambridge University | 0 | 4 |
Charles River Media | 0 | 2 |
Course Technology | 0 | 2 |
Digital Press | 0 | 1 |
Electronic & Database | 0 | 1 |
Exam Cram | 6 | 0 |
firstPress | 0 | 1 |
Friends of Ed | 0 | 4 |
IGI | 0 | 1 |
iUniverse | 0 | 2 |
John Wiley & Sons | 4 | 21 |
Jones & Bartlett | 0 | 3 |
Macromedia | 8 | 0 |
Manning | 1 | 0 |
McGraw-Hill | 0 | 7 |
Microsoft | 17 | 25 |
MIT | 0 | 1 |
Morgan Kaufman | 0 | 1 |
New Riders | 14 | 0 |
O'Reilly | 38 | 0 |
Oxford Press | 0 | 1 |
Packt | 0 | 3 |
Paraglyph | 0 | 3 |
Peachpit Press | 15 | 2 |
Pragmatic Programmers | 0 | 2 |
Premier Press | 2 | 9 |
Prentice Hall | 15 | 0 |
Que | 5 | 0 |
Redmond Technology | 0 | 2 |
Sams | 26 | 2 |
Sitepoint | 2 | 3 |
Springer | 0 | 1 |
Sybex | 0 | 15 |
Syngress | 1 | 3 |
Visibooks | 0 | 4 |
Wordware | 0 | 1 |
Wrox Press | 0 | 28 |
Appendix 3
Usefulness of books
| Usefulness (done for Python and Maya titles only) | Safari | Books 24X7 |
|---|---|---|
Number of libraries in WorldCat (average) | 203 | 194 |
Number of positive reviews in reviews.com | 6 | 6 |
Average review score on Amazon.com | 3.8 | 3.45 |
Appendix 4
Usefulness of O'Reilly books
| Usefulness (done for Python and Maya titles only) | Safari | Books 24X7 | O'Reilly titles |
|---|---|---|---|
Number of libraries in Worldcat (average) | 203 | 194 | 280 |
Number of positive reviews in reviews.com | 6 | 6 |
|
Average review score on Amazon.com | 3.8 | 3.45 | 3.8 |
