Online vs CD-ROM - the Impact of CD-ROM Databases upon a Large Online Searching Program

Article excerpt

The Evans Library of Texas A&M University has had an online searching program since 1976 when it began offering mediated searches to the faculty, staff, and students of the university. Since 1982, the Library has offered one no charge search, per academic year, to any graduate student actively working on a proposal, thesis, dissertation, or record of study. In 1983, Evans began experimenting with enduser searching, a program that has grown to such an extent, that end-users now consume nearly 75% of the total online time for all searching done at the Library


In 1986, Evans Library installed its Wiley Laser Disk Service (see the March 1988 and May 1988 issues of ONLINE for detailed information about the installation) which now includes 23 databases on compact disk or optical disk. The databases are available to patrons on a first come first served basis. We do not take appointments for the CD-ROM databases, although we do have waiting lists for the most popular ones, and will limit users to 30 minutes when others are waiting. ERIC SilverPlatter was one of the first databases added in late November 1986. Dissertation Abstracts OnDisc was also available early as a test product. The production model was added in March 1987. A month later PsycLIT SilverPlatter became available, and Agricola SilverPlatter was added in June 1987.

These four databases are the basis for comparisons to online searching because they have been available long enough to show some effect on online searching, and because the CD versions are comparable to the time periods covered by their online equivalents, with the exception of Agricola which covers 1981 to the present on CD-ROM, as compared to 1969 to the present as an online database.


Figure 1 shows the monthly percent of online connect time for each of the four databases. Together, these databases accounted for 31% of all connect time in October, but the percentage dropped to a low of 7.5 in November 1987. For the purposes of this article, 16 months were selected for review, representing at least two months before the installation of the first Laserdisk database through the 1987 calendar year.

The number of online users for each database follows the same trend. In Figure 2, it can be noted that the total number of online users for the four databases decreased considerably after the laserdisks were installed. In the case of all four databases, the users who migrated in the largest numbers from online to Laserdisk products were the users of the BRS/After Dark and Dialog's Knowledge Index services. The people who had mediated searches conducted seemed to still prefer mediated searches, albeit in smaller numbers.

Figure 3 indicates that the number of ERIC online users decreased significantly after the installation of the Laserdisk product in November 1986. This pattern is not so evident with Dissertation Abstracts, as shown in Figure 4. The peak months of Dissertation Abstracts usage (November, January, and September) correspond to those months in which graduate students are most heavily engaged in proposal writing.

Figure 5 presents the number of online users by month for PsycLIT. Again, shortly after the Laserdisk product was installed in April 1987, the number of online users decreased from a high of 60 in April to a low of 17 in November and 6 in December. December figures are generally atypical, because of the limited hours of library service during the Christmas holidays. Figure 6 shows that Agricola use followed a similar pattern, with online usage decreasing dramatically after the introduction of the CDROM version.


Figure 7 presents for each type of database service available in the Library, the percentage of the users who are:

1) graduate students;

2) faculty and staff,

3) undergraduate students; and

4) other

The other category includes non-university users such as local community residents or visitors from nearby colleges and junior colleges. Only information on three databases was available for presentation in this chart, as ERIC is located in a different department whose method of collecting data is somewhat different, so the comparison would not be valid.

The bar graph clearly shows that faculty and staff make up the heaviest users of the mediated searching service. Our experience has been that a small number of the faculty become enthusiastic users of the Laserdisk products, as a few faculty members have in the past utilized the end-user searching service. Not as many faculty and staff will use either, however, because of limitations upon their time. While previous research projects at the Evans Library have shown that faculty like the concept of conducting their own searches, in practice they prefer to make an appointment for a mediated search or to send a student or staff member to have a mediated search done for them. They do not like to tie up evening and weekend hours to conduct end-user searches, nor to wait in a queue to use laserdisk products.

As indicated in Figure 7, 51.7% of those who conducted mediated searches of Agricola during the time period of the study were faculty and staff, 35.5% were graduate students, many of whom were taking advantage of the opportunity to have the one free search related to their thesis or dissertation topic, and 12.8% were other users. The absence of undergraduates from the population of Agricola mediated search users probably is not unique to our library.

In the case of Dissertation Abstracts, the utilization of the mediated searching service was approximately equally divided between graduate students (44%) and faculty/ staff (43%), with other users accounting for 13%, and undergraduates 9%.

Faculty and staff were the heaviest users of Psychological Abstracts mediated searches (46.2%), with other users representing 46.1% of the use. In contrast, graduate students represented only 7.7% of the users of Psychological Abstracts as a mediated search.

In the case of Agricola and Dissertation Abstracts, graduate students made the greatest use of the end-user searching program, accounting for 52.9% and 68.4%, respectively, of the use of these databases. The pattern was slightly different for Psychological Abstracts; graduate students were slightly edged out by undergraduates who used the end-user searching program 52% as contrasted with 46.8% for graduate students.

Undergraduates make up a significant portion of the population of CD-ROM database users, even specialized ones such as the three being presented here. For example, 38% of the users of Agricola on CD-ROM were undergraduates, as were 26.3% of the users of Dissertation Abstracts, and 63.6% of those who used Psychological Abstracts. Graduate students were the second largest group of users, representing 48.9% of the Agricola CD-ROM users, 56.3% of the Dissertation Abstracts users, and 32.5% of the Psychological Abstracts users. Our experience for these databases and others has been that graduate students and higher level undergraduate students will make heavy use of the databases, with faculty and staff having some sort of funding generally preferring the mediated searches.


Part of the downward trend in percent of total connect time and in number of users can be explained as follows: in September 1987, at the beginning of this academic year, the Library's offer of one no charge search to graduate students specifically excluded those databases which the Library owned on compact disk. We would search online the most recent months not covered by the compact disk, but no more. This naturally caused a reduction in the number of no charge searches in these databases, because the graduate students tend to select databases likely to produce a great number of citations rather than limit themselves to the latest three or so months of a database, For example, last academic year, 59% of all mediated searches in the ERIC database were no charge searches for graduate students; whereas, this year with access limited to the latest months, only 31% of mediated searches in ERIC were no charge searches.


A pronounced upward trend in Knowledge Index and BRS/After Dark connect time occurred in early 1988. Beginning January 17, 1988, the Library received a gift from the Association of Former Students to pay the connect costs of all student end-user searching. Prior to this time, the Library paid half the connect costs and the user paid half the connect costs plus all printing costs. The Association of Former Students' gift essentially meant that all Knowledge Index searches were free to the students, and BRS/After Dark searches were reduced from an average of $5.11 to $2.16. The effect of the gift revealed itself in two ways. First, students who previously used BRS/After Dark immediately switched their allegiance to Knowledge Index, if the database they wanted was available on KI. Second, our total end-user connect time nearly doubled, and would have increased more if we had had the staff and terminals to accommodate the demand, It should be noted, however, that the percentage of connect time for the four databases studied remained in line with previous months. Our total connect time increased, not merely the demand for these four databases.

The staff members who monitor the After Dark end-user program are careful to explain to users that ERIC, PsycLIT, Agricola, and Dissertation Abstracts are:

* available to them as CD-ROM databases

* free

* available far more hours than the After Dark online databases

In spite of this explanation, the majority of these end-users go ahead and search the database online. Many of them express the opinion that online is better. In a way, this is true. Until we increased the clock speed on our CDROM microcomputers, online was certainly fasten Some techniques, such as limiting a term to title and descriptor occurrences, can be done in one step online, whereas, three search statements are required on the SilverPlatter databases (TERM in ti; TERM in de; combine the two statements).

Other students are under the impression that ERIC online and ERIC on CD-ROM are two different databases, and it takes a long explanation about "search services" and "packaging of information" to convince them otherwise. Some are searching a database not available on CD-ROM, and in the same appointment period they also search one of the CD-ROM databases, "because they still had time left over." But the majority of them say they prefer the convenience of being able to make an appointment, of being sure that they can access the database at the time they choose. The most frequent complaint we hear from CD-ROM users is "I had to wait 15 minutes/30 minutes/an hour before I could use the database."


Our experience with the Association of Former Students money, and with users still wanting to conduct online searches of products we have on CD-ROM, has let us know that users really would like to have a variety of options, if given the choice. Yet for libraries, this is not always going to be a viable option. For example, had we been more strict ,and not extended the use of a subsidy for searching the online version of a database which was also available on Laserdisk, our library would have saved approximately $196 per month. In some ways it is duplicative to offer the same product in more than one way. One could argue that the Library money is not being used wisely. If we have money to subsidize searches the next academic year, it is likely that the Library will not pay any portion of the cost for those databases which are also available as a Laserdisk. For example, if a student wanted to conduct an online end-user search of ERIC, the full cost would be charged. However, we have discussed subsidizing the updating of a CD-ROM search, provided the administrative details can be worked out.

Also, it is apparent that we have not done enough to promote the CD-ROM databases or to educate our users to the fact that a compact disk database is just as good as the online database. Perhaps we need to promote CD-ROM at the expense of online, by charging our end-users more when they access a duplicated database. We have recovered approximately 45% of our online charges for these four databases from our fee-based searches and end-user charges, but we should be recovering 100%. We do not feel that our library can afford to pay twice for the same information.


While we consider our CD-ROM program to be highly successful, particularly since it has accomplished the introduction of computerized database searching to far more users than we could ever reach through online services, we do not see CD-ROM as the answer to all of our information delivery needs. The limitations have become all too obvious, making the Laserdisk products less attractive to knowledgeable users. Limitations include:

* only one user at a time is able to access each database, so that there are waiting lines for most databases at all times during the day;

* files on the Laserdisk are limited in size; and

* the response time is not as fast as that of most online services.

We expect to continue to offer a hybrid service, including:

1) mediated searching for those who have the funding and prefer to have an expert conduct the search;

2) end-user searching for patrons who want to do their own online searches, particularly of databases not available as Laserdisk products; and

3) CD-ROM and other Laserdisk databases, especially databases that are narrow in scope and thus have a limited number of users.

We hope someday to be able to offer distributed access to the most popular online databases. This could be accomplished by loading databases, such as those available via BRS/OnSite, or by using a gateway so that patrons could be assigned passwords for use from any terminal or microcomputer to remote databases with the cost either being paid by the University or the Library or with the user being billed.

We have found that the various kinds of services can co-exist, and, in many ways, they enhance each other. Users who are first attracted to a Laserdisk database only to discover that it does not cover their topics then ask questions and learn about the BRS and Knowledge Index programs. Many new users learn of the existence of the Wiley Laser Disk Service via word of mouth from friends, which seems to be the most effective advertising for the service.

Staff can effectively deal with the different services, since many of the searching principles are the same. The administrative details are different, and our end-user and mediated services have designated staff, while all of the staff of the Reference Division - librarians, classified staff, and students - work in the Wiley Laser Disk area. The individuals who service the Laserdisk area do not have to know all of the administrative and billing details that are required for someone to be able to work with the online end-user or mediated programs.

It is very clear that price is extremely important to students and will have a great effect on what services they will use. Thus, if someone adds a Laserdisk database and charges for it, students will be less likely to use it than if there is no charge. This seems to be less important to faculty and staff, but will make a difference in what students elect to do.


What have we accomplished by adding CD-ROM databases to our mediated and end-user searching programs? We have introduced far more users to the concept of computerized literature searching than could be accommodated by our online programs. Over a hundred people a day use the compact disk databases in the Wiley Laser Disk Service compared to a maximum of 25 appointment slots available to mediated or end-user searchers per day. Furthermore, the CD-ROM users are working in a less threatening environment because there are no clocks to watch and no fees for online connect time or citations; this is an excellent environment for learning the techniques of computerized literature searching.

The higher volume of users and the teaching/learning opportunity are very good reasons for adding CD-ROM databases. Although the online usage of these four databases has come down as a percentage of total connect time, usage of the online equivalents remains high and the total mediated and end-user searching programs have not been affected. Because users introduced to CD-ROM databases frequently are interested in trying their new skills in an online environment, the overall market for computerized services should expand.

Communications to the authors should be addressed to Vicki Anders, Head, Automated Information Retrieval Service, and/or Kathy M. Jackson, Head, Reference Division, Sterling C. Evans Library, Texas A&M University, College Station, Texas 77843; 409/845-5741.