SUBSCRIBE NOW!

Vol. 33 No. 4 — Jul/Aug 2009

Early in my days as director of the National Science Foundation (NSF) Library, a program officer asked me for help filling out a panel of reviewers. On a short deadline, he had worked for 2 days and only found two possible reviewers. In desperation, he wondered if there was any assistance I could give. The topic was “vibrational spectroscopy.” In 2 hours, I had a list of several hundred possible reviewers, with information about each reviewer to help him choose those most appropriate. He was astounded that I could do this so quickly and easily, and I was surprised that he could not.

As a result, I developed a workshop that all new program officers at NSF would attend on how to use standard library resources to quickly and easily find and evaluate possible peer reviewers, even if the subject area is unfamiliar. These program officers come to NSF as highly sophisticated users of standard databases, having used them throughout their careers as scientists. But what they have not done is viewed these tools through this particular lens.

Web of Science and Scopus both offer a wealth of incredible features for this task. These databases are so powerful, flexible, and elegant that they have abilities you never even dreamed were available. Let me give you a quick walk through some of the special features that can make this task simple and effective for you and your patrons.

WEB OF SCIENCE

Both databases have many similar abilities, but each also has unique features. I will start with ISI’s Web of Science (WOS). To get the most from its elegant features, run your search only on Web of Science since other databases on the WOS platform have fewer features and will limit your ability to manipulate the data. For my initial search, I will look for the exact phrase vibrational spectroscopy in article titles only. I will also require that USA be in the address field so that at least one author from each result will have been resident in the U.S. at the time of publication. And I will limit my search to the most recent 5 years. Note that in this database, if I limit my initial search to the most recent 5 years, that limitation will be carried forward through all my subsequent steps.

I am going to first manipulate my results by changing the sort. WOS automatically lists the results in reverse chronological order. Changing the sort to “Times Cited” brings the most heavily cited search results to the top of my list. Since I am looking at a 5-year time horizon, this is not a completely “fair” evaluation; older articles have had more time to be cited. But if you keep that in mind, sorting by “Times Cited” is a “quick and dirty” way to bring some of the most interesting articles to your immediate attention.

You can now choose several articles and check the article records. Links will allow you to go forward and backward in time from a single, highly relevant article. Notice also that each record gives the contact information for all of the authors, including email address if this was published with the article (although only the “corresponding” author will be specifically “tied” to his or her contact information). This information may not be current since it is given at the time of the publication of the article, but it is not updated by WOS. Still, it gives you a starting point.

REFINING RESULTS

You can also choose to look at the “Related Articles.” These will be articles that share references with the particular record being examined, on the theory that if two authors are citing the same papers, they must be addressing the same topic, even if they use different keywords. If I have done a very tight search and have only retrieved a few search results, I can quickly expand my results by following the links in “Related Articles,” without having to restructure my search or come up with additional keywords.

On the other hand, if I have too large a pool of results from my initial search, I can use the “Refine Results” feature (to the left of the search results) to elegantly narrow the result pool. Notice that I have the ability to refine by “Subject Areas.” If I need reviewers, for example, that are not simply knowledgeable about “vibrational spectroscopy” but who have particular expertise in the instrumentation aspects of this field, I can target likely groups of articles, perhaps the groups “spectroscopy,” “instruments and instrumentation,” and “optics.”

These subjects are assigned not to the individual articles but to the coverage of the journals in which the articles are published; therefore, they are independent of individual article keywords. In a single, easy step, I have identified authors with the very specific expertise I need without wracking my brain for additional keywords that may or may not have given me clean search results.

I now have a rough idea of the top authors in the field of “vibrational spectroscopy,” by virtue of having sorted my initial search results. I also have a list, if I need it, of authors who have a very specific expertise in this discipline—all obtained in a matter of minutes.

I can also use the “Refine Results” feature in another way that will yield specific lists of authors who have essentially been already “peer reviewed” for me. I can refine my initial pool of search results by “Document Type,” choosing “Review Article.” I know that, generally, people are asked to write review articles because they are well-known in their fields, so all the authors of these review articles are of probable interest as peer reviewers. I can examine the list of review articles and know that each one of the references listed in each of these articles has authors who have been peer reviewed for me by the authors of the review articles—who I already know are tops in their fields.

I have quickly built a considerable list of possible peer reviewers. Next, I need to eliminate those reviewers with possible conflicts of interest (COI).

CONFLICT OF INTEREST

Look again at the search results from the initial search. I will now use one of my personal favorite features of WOS, the “Analyze Results” button. You will find it at the top of your search result list. You will love this button! You are first asked to set the specifications for your analysis. I have chosen to analyze by institution name.

I now have a list of the institutions that are the “major players” in the field of “vibrational spectroscopy.” Perhaps I know that, for my purposes, I will have conflicts of interest with authors from, for example, the University of California–
Berkeley. I can go down this list and choose the institutions that are least likely to present COI problems. I choose a dozen or so of these non-California institutions; I now have a list of 93 results to examine. I have not completely eliminated UC Berkeley because the institutions’ authors may have co-authored some of these papers with UC Berkeley authors. But my short list should be enriched for non-UC Berkeley authors compared to my initial search results.

I can manipulate this list of search results in any of the ways I manipulated the initial list: I can change the sort, refine by subject or by article type, “Review,” and reanalyze a set of search results as many times as I want. If I “Analyze Results,” yet again, using the list generated by the institutional analysis and analyzing this time by “Author,” I have now determined who the most prolific authors from the chosen institutions are.

ZEROING IN ON AN AUTHOR

I see that Z. Chen is at the top of this list. Perhaps I am familiar with this author’s work and know that Dr. Chen is of particular interest, but I need to know more. I can choose Z. Chen from this list and “View Records.” I have now isolated the body of work authored by Z. Chen that have the exact phrase “vibrational spectroscopy” in the article title and that have been published within the past 5 years. (I could do a straight author search on Z. Chen in order to pick up the author’s complete body of work, but for now I am happy to work with this particular subject-oriented, limited set of results.)

I can find Chen’s co-authors by analyzing my set by author. Note that I have a set of articles all written by the same author, and I am also analyzing this set by author. This seems redundant, but it actually gives wonderful results. I now see that 100% of the papers written by Z. Chen have Z. Chen as an author, which is as it should be! But I also have a list of the authors with whom Z. Chen has co-authored on this topic within the past 5 years—because 5 years was the limit I set for my initial search way back at the beginning of this article. Those limits are still in effect.

The list will be complete for all of the journals covered by Science Citation Index. (Note that in order to get a list of all authors—rather than just those with articles that include the title phrase “vibrational spectroscopy”—I should start with an author search instead of using the “Analyze Results” button from a subject search.) In any case, I can quickly look down this list to see if there are any apparent conflicts of interest.

Additionally, if I have decided that Z. Chen is of particular interest but I also suspect Chen may be unwilling or unable to do peer review for me at this time, I now have a list of people that Chen has essentially peer reviewed for me—Chen’s partners in publication.

I could continue to use the “Analyze Results” button to get more information about Z. Chen. For example, analyzing a search by country reveals if Chen has participated in any international collaborations on this research. Analyzing by institution reveals with which institutions Chen has had collaborations. Analyzing by Source Title reveals in which publications Chen prefers to publish, which is very useful since I know the most prestigious journals in this field.

Analyzing by subject area shows other subject areas. It may be interesting to search for other peer reviewers in these extra subject areas. This would allow me to expand my search in ways I might not have considered. For each of these questions, I only have to use a single, simple button to harvest a wealth of knowledge about this author and this field!

WOS also allows me to examine the effect of the body of Z. Chen’s work in this particular area by using the “Create Citation Report” (this button is at the top right of your list of search results). This information has long been used for purposes such as tenure evaluations.

SCOPUS

Now let’s look at another product. As mentioned earlier, Scopus has many of the same features as WOS, but it also has many unique features.

I am going to use the same search on Scopus, vibrational spectroscopy in article title, united states in affiliation, and limited to the most recent 5 years. On the results screen, you will notice that the “Refine Results” area at the top of the screen actually does work similarly to the “Analyze Results” button and the “Refine Results” feature in WOS. It shows you the major authors and journals in your result pool and allows you to break your results down into narrower subject groupings, although these subject groupings are not, at this time, as specific as those in WOS.

You can also limit your results by article type to get those useful “Review” articles. You can change the sort by clicking on the column heading “Cited By”; the most highly cited articles will then come to the top of your list. With a few clicks, you can do some very fancy manipulations of your data. Note that as Scopus searches its database, it also searches the web and the patent literature. (If your search statement included affiliation, you will likely not get any web results, however.) It is important to remember that scientists from commercial entities (rather than academic institutions) might be found in the patent literature more readily than in the journal literature.

If you examine an individual article record, the authors and their contact information (at the time of publication) will be listed. But unlike WOS, each author will usually be tied to his or her specific contact information, which eliminates some guesswork! Notice that each author’s name is associated with a “person” icon.

On the right-hand side of the screen, you will see the citation history, allowing you to go forward and backward in time as in WOS. You can also look for related records—and Scopus does this in a most elegant fashion! Notice that Scopus allows you to look for related records by references, authors, or keywords. Notice too that you can click in the boxes in front of particular articles to limit your search to records related to only those articles. This gives you the possibility of getting very clean and specific results. You can also choose to look for related articles just in the Scopus database, or you can expand out to the web using the wonderful Scirus search engine.

Additionally, if the target record has been cited by a patent application, there will be links to this application. This allows you to find scientists who are affiliated with commercial enterprises that may emphasize patents more than the production of journal articles. As a funding organization, it also allows you to possibly track the practical applications of pure research funding you may have provided. This is a very interesting resource.

Now, let’s look at that author icon I noted earlier. Clicking on this icon yields detailed information about the author and the body of the author’s work. This will include the familiar and useful citation analysis features (use the “Citation Tracker” button). It will also include a list of co-authors (but these will not be, as in WOS, limited to the time period of your search statement). It will also include a history of affiliation, which is helpful in determining if “this is the same author who was previously at …” This information is not 100% accurate (nothing ever is!), so you will want to double-check—but it gives you a starting place. There is also a unique author ID which generates the complete list of the author’s articles. This list, like your original search results, can be manipulated using the “Refine Results” feature.

CHOOSE BOTH IF POSSIBLE

Which of these databases do I like best? Both! Many institutional budgets force you to choose one or the other because they are often perceived as being duplicative resources. It is true that they accomplish many of the same tasks. But each also has particular—and extremely valuable—features absent in the other. I love the precision and scope of the “Analyze Results” button in WOS and the ability to limit the author information to a particular time period. I also love the precision of the “Refine Results” choices in Scopus and its easy entry into the patent literature.

Look at the breadth of what I have accomplished using these two resources. I have harvested a huge amount of detailed information in a matter of minutes—all by using a single subject phrase and two databases. Without even knowing exactly what “vibrational spectroscopy” might be—with no subject knowledge—I can easily compile an accurate list of some of the most influential authors, a list of authors of review articles, and a list of authors of papers on particular narrow aspects of the topic.

I can eliminate some possible conflicts of interest. I can list authors having collaborations with other institutions and with international collaborations. I can determine which journals are favored by a particular author or even by authors in a particular institution. I can measure the extent of the influence of the body of work of a particular author in this field. I can analyze the entire field to see if it is increasing or decreasing in prominence. I can look for patents and track the influence of individual articles into the field of practical and applied applications. I can work miracles for the scientists and the students who depend on my information expertise.

None of these database tools were developed for this particular purpose. But once I know that these features exist and how they can be used, I can be alert to the possible use of similar features in other databases. I am looking at my standard tools through a new and different lens. Instead of a purely subject search lens, I now have a peer review or colleague search and evaluation lens. And now, so do you.