Is It Challenging PsycINFO?
The knee-jerk reaction to e-psyche’s launch was predictable: Does it really hope to compete with the American Psychological Association’s (APA) venerable PsycINFO database (http://www.apa.org/psycinfo)? PsycINFO had been among the first online databases, and has earned an excellent reputation for its extensive coverage of a well-chosen mix of journals, books, book chapters, conference papers, and dissertations within the discipline. Its CD-ROM version (a subset of the online version) has been on college libraries’ top-10 lists for a long time, and it’s been one of the most widely hosted databases by online information services. PsycINFO is not inexpensive, but it’s certainly a far better value than the Mental Health Abstracts (MHA) database that it was running neck and neck with (until MHA was bought by IFI/Plenum and turned into the epitome of a database that’s withering into extinction, as I discuss in my column in the December 2000 issue of Information World Review).
APA could have rested on its laurels, but it didn’t. It not only updated
PsycINFO regularly with about 5,000 records a month, but also increased
its update frequency; revised and chiseled its thesaurus; produced good
user guides; linked the analytic records to the monographic records and
back; and recently added a substantial historical subset, going back to
the late 1880s. Then came e-psyche, which threw down the gauntlet to challenge
More and Less Than PsycINFO
This kind of challenge is quite unusual, although not unprecedented. Such a venture is not in the same league as the ones from entrepreneurs who challenge the online information service providers by offering competing products at discounted prices. They don’t create abstracting-and-indexing or full-text databases, but license them, enhance them with other content, and integrate them with smart software. Northern Light did this a few years ago, and, to a much lesser extent, Contentville did it earlier this year.
The e-psyche challenge is more comparable to what Dick Harris (of Responsive Database Services) did when he created high-quality and unique content (especially in the TableBase database) and entered into the overcrowded market of business-journal databases. His competence and dedication made his databases fierce competitors over-night in an area where it was hard to imagine having a new kid on the block.
It seems that e-psyche also has good backing. Its co-founders, Dennis Auld and John Kuranz, who launched ABI/Inform and Management Contents nearly 30 years ago, are veterans of the industry. The team also includes Marjorie Hlava and her Access Innovations company, which has impressive experience in the software aspect of creating and converting databases, especially with its powerful, new Data Harmony tool set. The selection of Cambridge Scientific Abstracts as e-psyche’s first online information service partner was also a good one. CSA’s software is among some of the most intuitive, and it makes good use of the Web. Since it’s not too large a service, e-psyche won’t be lost among the other databases.
Still, the ambitions of e-psyche are very high. It promises that nearly 100,000 records will be added every year on a weekly update basis, which is about two-thirds more than what PsycINFO does. It also promises to add records for articles within a month of their publication. Eventually, e-psyche will have a nearly three-times wider journal base than PsycINFO, which processes around 1,400 journals. In terms of the record content, the most significant extra feature is the addition of the bibliographic data of all the works cited by the source documents, similar to ISI’s citation indexes. There is one area where it offers far less than PsycINFO: Its records only go back to 1996, while PsycINFO recently enhanced its database with a large subset of records with articles dating back to the 19th century.
As I was testing the e-psyche database in late October, I couldn’t corroborate
its promises because there was only a small subset of 13,500 records from
just a few journals. However, the journal list is available for free on
the producer’s Web site at http://www.e-psyche.net/journals.html,
and the database subset gave enough samples to ponder a little about the
decisions and implications from the perspective of a back-seat driver who
went for a test drive.
Thoughts of a Back-Seat Driver
As for the retrospectivity of the e-psyche database, it’s not bad. Most users want current information, not historical. True, the cited half-life of journals in psychology and psychiatry in the most recent edition of Journal Citation Reports (JCR) would warrant a retrospective time span at least twice as long. But as it’s well-known from published citation analyses (not to mention relaxed conversations), a significant part of the older literature is cited for convention and courtesy, not for its direct relevance to the article being written. I wouldn’t lose much sleep over the issue of retrospectivity.
The width of the planned journal base, however, may be of concern. The list includes 3,875 titles—that’s about the same journal-base size that MEDLINE has for the entire field of medicine. The JCR covers 480 serials for psychology and psychiatry. MHA is not relevant at all, since its journal coverage this year has plummeted to an abysmal 23 titles.
As I scrolled up and down e-psyche’s journal list, I didn’t sense any padding—the titles seemed to be relevant for psychology and psychiatry. For example, the titles that have the words “social psychology” or “social psychiatry,” and are not covered by PsycINFO, included such journals as Advances in Experimental Social Psychology, Asian Journal of Social Psychology, Contemporary Social Psychology, Current Research in Social Psychology, Journal of Experimental Psychology, and Social Psychology of Education. These were clearly relevant for the scope of the database. In this simple test of journals, e-psyche had twice as many titles (26) than PsycINFO. (Remember, at this stage few of them actually appear in the database.)
Tripling the journal base of PsycINFO seems a tad excessive. Having subscriptions to that many journals, paying for the grossly overpriced back issues, and processing all these back issues would devour an enormous amount of venture capital. It’s more important to make absolutely sure that the most significant journals are there. For example, the list didn’t include the unquestionably essential title in this category, Social Psychiatry, and my test search confirmed its absence. You should not merely check the journal list as it may omit a journal that’s indeed covered. For example, I was surprised not to see the journal European Addiction Research on the list. However, a search revealed that it’s already covered in the preview set. I think e-psyche needs to scale back its journal base to a more feasible 2,000 titles, and spend the money on the promising citation data. (See Figures 1 and 2.)
The citation data were present in the sample record (except for the odd-looking book reviews). Even in this passive, view-only mode, they’re useful. They would be more useful if they were hotlinked for jump-searching from the displayed record. It wasn’t clear to the naked eye how deeply structured the cited items are, but because the raw data is in XML format, chances are good that searching by cited author, journal, year, and work all would be possible. If CSA—or e-psyche in its own version—would implement a software feature like DIALOG’s RANK command, it would allow interesting bibliometric analy-ses, and that alone would attract quite a few users.
In many records the author’s e-mail address is hotlinked, and so are the descriptors and authors. It would be useful to do the same for identifiers and journal names. The Resource locator is hotlinked too, but it takes users to the publisher’s home page. Linking to the journal itself would be better, because this would allow users to quickly check the table of contents and, in many cases, also the abstracts for issues that are not yet in the database, but are available at the publisher site free of charge (as is the case with Karger). (See Figure 3.)
In the CSA version I’d like to see some additional features, such as more sort options (perhaps by journal and author), user-defined print format, and most importantly, browsable indexes for most of the searchable fields. This effective software could be made perfect via such enhancements plus multiple-database searching with de-duplication possibilities, as in DIALOG and Ovid.
The jury is still out about the extent to which e-psyche will challenge the incumbent king, but its efforts are worthy. One big advantage of PsycINFO is that it’s available for the casual user on a pay-as-you-go basis on the HealthGate site, where you pay only when you ask for the full record with abstract. The e-psyche database is meant to be licensed for the academic and corporate market, which may be an unnecessary limit. It should partner with medical portals to offer access on a pay-per-view basis. The PsycINFO deal with HealthGate may have been triggered early this year by the progress of the e-psyche project. It’s too early to say if e-psyche is a better mousetrap (especially because its price was not yet announced), but it’s good to have one more option.
[Editor’s Note: For more on PsycINFO, see the news
story on page 45.]
Péter Jacsó is associate professor of library and information
science at the University of Hawaii’s Department of Information and Computer
Sciences, and a columnist for Information Today. His e-mail address is
|Table of Contents||Information Today Home Page|