My Library is a personalization tool that allows the library user to consolidate frequently used library resources and services. Since 1998, My Library has been available as a web page to users of the Virginia Commonwealth University Libraries. An evaluation of My Library use at Virginia Commonwealth University (VCU) was completed and published in 2000. This article reports on a follow up study that took place between March 2000 and December 2001. The article also discusses the value of My Library as a teaching tool at VCU.
My Library is a personalization tool that allows the library user to consolidate frequently used library resources and services. At the time of its inception at the Virginia Commonwealth University (VCU) Libraries in the fall of 1998, My Library was one of the first such services to be offered through a library web site [1, 2]. In the fall of 1999 a log file was added to measure My Library usage. Between October 1999 and March 2000, the log files were analyzed in order to evaluate the service. The VCU Libraries reported a general pattern of use whereby the lion's share of accesses was accounted for by a minority of users [Ghaphery & Ream, 2000]. In addition, the use of class My Library pages as part of a library instruction session was also seen as driving a strong percentage of the overall use of My Library. This article seeks to follow up on the earlier results with particular attention to My Library as a teaching tool.
My Library at VCU
Judging from programs and roundtables held at various conferences and the general traffic of e-mails I receive on the topic, interest in personalized services like My Library remains high. At the same time, there have not been many follow up studies on the measured use of My Library. A possible reason for this might be the fragmented nature of library personalization. In the case of VCU Libraries, the initial concept for the My Library service was ahead of technological infrastructure to support it. For example, in the fall of 1998 the VCU legacy Integrated Library System (ILS) did not support online borrowing transactions and Interlibrary Loan was mediated through a simple web form. While the spring of 2002 finds these types of web services commonplace, a unified package of services remains to be fielded. At VCU Libraries we have settled, for now, on combining all our personalization services into a single web page, even though the various services available through that page require separate logins. We do, however, have high hopes for being able to offer a unified and personalized login to library services in the near future, as we begin to migrate to a new ILS.
Growth in the number of My Library accounts at VCU
At VCU, the total number of My Library accounts at the end of 2001 was 2,586. This represents a substantial increase in the number of accounts over the preceding year. In March 2000, after inactive accounts had been purged there were a total of 840 accounts. Thus, from March 2000-December 2001, 1746 new accounts had been created. Before the purge of inactive accounts in March 2000, there were a total of 2,336 accounts.
Of the total accounts at the end of 2001, 824 (32%) had not accessed My Library at all in past year. Only 13 of these inactive accounts can be clearly tied through an email domain to guests from other educational institutions. In 2000 there was a greater percentage of guests, and since that time the general curiosity from the outside community has cooled and a clearer pathway to a guest login has been established. Forty percent (327) of the inactive accounts contain e-mail addresses with a vcu.edu domain. The remaining 59 percent (484) of inactive accounts are from .com addresses (i.e., hotmail.com, aol.com, etc.), which may or may not be affiliated with VCU. In comparison with previously reported data, the number and percentage of inactive accounts has declined (see Table 1).
An analysis of My Library access activity
There were 25,842 My Library accesses for the calendar year 2001. Further analysis supports the trend from previous years where the majority of these accesses are tied to a minority of the My Library accounts. Likewise, a large percentage of My Library accounts are mostly inactive. For example, 780 accounts accessed My Library either once or twice in the past year for a total of 1,069 accesses. If we add in the 815 accounts that did not access My Library at all, 1,608 of the 2,585 My Library accounts (62%) were seldom seen at about 4% of the total My Library traffic. Looking at the other end of the spectrum, 103 accounts (4% of all accounts) had accessed My Library more than 50 times in the past year. This accounted for 15,665 accesses and 61% of the total My Library activity. Thus, 4% of the My Library accounts drove more than 60% of the activity. Conversely, in a weird parallel 62% of the My Library accounts dented the log files for only 4% of the total activity.
The high end of the accesses is skewed by the presence of My Library accounts that had been set up by librarians for a specific library instruction session. More than 50% (54 of 104) of the "power" accounts that had accessed My Library more than 50 times were from such class accounts. Nevertheless, even after removing the My Library class accounts from the picture, we still see the general pattern of the majority of use by a small minority of users. The 49 power users that were not part of My Library classes show up 4,895 in the log files (19% of all My Library activity).
In comparing these patterns with the previously reported use for October 1999 through March 2000, we see a similar portrait of a small number of users accounting for the greatest demand as well as a sizable percentage of rather inactive accounts. In comparison, both sets of data reveal that 9% of the active users accessed their accounts more than 20 times. The greatest difference in the two sets of data is in the percentage of total accesses for these frequent users. Here the number jumps from 44% to 57%. One explanation could be that, as the data set from 2001 covers a longer period of time, the increase may reflect frequent users over time becoming more persistent in logging into their accounts. It is interesting that even given the longer time period, the percentage of disinterested users remains around 10% for both sets. That is, the extra six months did not provide an increase in inactive user visits to My Library (Table 2).
In terms of measuring overall growth or decline of the service, it is helpful to compare similar time periods. In looking at a snapshot of the service over three years for the month of October, there is a significant increase in both the number of active users and the volume of logins. It is curious to note that the number of accesses has grown at a larger rate than the user base. This pattern is both true for individual and class accounts and serves to temper some of the global patterns of inactivity by a majority of users (see Table 3).
Customization of My Library accounts
The VCU implementation of My Library includes a very simple interface and does not include many options for advanced levels of customization. Two advanced options are the use of a cookie to remember the login information and the ability to create a custom list of electronic journals. The custom list of e-journals is created by entering a second level edit screen, which requires more time and commitment from the user. Not surprisingly, those users who had selected either of the advanced options were more likely (by a factor of 3) to use their pages (see Table 4). This certainly raises the question of whether more advanced features or custom features could increase user loyalty.
My Library as a teaching tool
Perhaps the most unique aspect of VCU's My Library is its integration into the bibliographic instruction program at James Branch Cabell Library. This is probably due to the fact that the original developers of My Library at VCU were faculty from the education and outreach services department there. The continued increase in the use of My Library class pages points toward an adoption of this teaching strategy by other librarians. From the perspective of the teaching librarian, the My Library toolkit provides a quick and efficient way to create a customized page for a specific class.
In comparing class use for the past three years, we see that class accesses for the month of October jumped 66% from October 1999 to October 2000, and an additional 108% from October 2000 to October 2001. For calendar year 2001, 68 My Library class accounts were responsible for 11,155 accesses or 43% of the total My Library activity. In comparison, for the previously reported data from October 1999 through March 2000, class accounts were found to be 28% of the total My Library activity. At that time it was also noted that the total number of accesses was in part due to the fact that the entire class was using a single account. Further, each class account registers a significant number of accesses from the use during the library instruction session.
In order to take a closer a look at the use of these class accounts, it is useful to filter on the IP address of the library classroom. Of the 11,155 class accesses, 2,037 (18%) were logged from the library classroom. In terms of the global use of My Library, these 2,037 classroom accesses represent almost 8% of the total My Library activity for the 2001 calendar year. It is heartening to see that 82% of the total My Library class use was outside of the classroom. In terms of a pedagogical strategy, the My Library class account is not merely meant as a way to demonstrate a customized collection of library resources; it also serves as a tool to transfer learning, whereby the student can revisit the page after the class session.
The IP address portrait of the class accounts also offers up some sobering data. In order to access library databases from off campus through My Library, the user must configure his or her browser to go through the library proxy server. The My Library log files recorded 1,763 off campus accesses that were not validated through the proxy server. Thus, almost 16% of the My Library class accesses could not take advantage of the majority of the resources on the My Library page. Conversely 2,283 (20.5%) of the class accesses were validated through the library proxy server. The remaining 5,072 (45.5%) class accesses were from elsewhere on campus, outside of the library classroom (see Table 5).
There are several competing directions for future development of My Library at VCU. On one hand, the greatest loyalty to and success with the My Library tool involve individual accounts that have taken advantage of some of the advanced features. This seems to bode well for increased and more centralized customization features such as access to patron data, interlibrary loan, new acquisitions, etc. On the other hand, development of My Library in this direction could exclude generic class accounts by requiring individual authenticated logins. The use and popularity of My Library as a teaching tool points toward the need for librarians and other educators to be able to easily customize and present digital content.
Ideally, further development in both of these directions could be accommodated once the larger management systems mature. That is, the advanced personalization and authentication features might be handled out of an Integrated Library System while the digital content for educational purposes would dovetail with the institution's Course Management System .
The real mark of success in this vein would be a transparency of the My Library service whereby patrons have unfettered access to customized library resources and services depending on their information need. In the context of a class led by a professor or librarian, selected resources would sit freely and prominently alongside online course syllabi. For more general library use, a secure and single login would offer individual users a robust portal to digital content and services.
 Other early starters in this field were the University of Washington, California Polytechnic State University, and North Carolina State University. For further discussion of these initial efforts, see the ITAL Special Issue: User-Customizable Library Portals [Morgan, 2000].
 In a similar vein as early as April of 2000, Cornell University Library published on their efforts at fielding personalized electronic services [Cohen et al. 2000].
 Incorporating library content and services into course management systems is one of the top issues for the Academic Library Advisory Committee of the Council on Library and Information Resources [Cohen, 2002].
Cohen, David. "Course-Management Software: Where's the Library?". EDUCAUSE Review. May/June 2002. Available at <http://www.educause.edu/ir/library/pdf/erm0239.pdf>.
Cohen, Suzanne with John Fereira, Angela Horne, Bob Kibbee, Holly Mistlebauer, and Adam Smith. "MyLibrary Personalized Electronic Services in the Cornell University Library". D-Lib Magazine Volume 6, Number 4, April 2000. Available at <http://www.dlib.org/dlib/april00/mistlebauer/04mistlebauer.html>.
Ghaphery, James and Ream, Dan. ITAL. Volume 19, Number 4, December 2000. Available at <http://www.lita.org/ital/1904_ghaphery.html>.
Morgan, Eric Lease, Guest Editor. ITAL. Special Issue: User-Customizable Library Portals Volume 19, Number 4, December 2000. Selected articles and full table of contents available at <http://www.lita.org/ital/ital1904.html>.
(Corrected coding on navigational link to "In Brief" at the bottom of the page 8/31/05.)
Copyright © James Ghaphery