Research Tips for
SL&IT School Library Students

• For books and other library materials at Mansfield University and beyond check PILOT (MU Library Catalog), EZ-Borrow (consortium of PA and neighboring state libraries), Access Pennsylvania (another consortium of PA libraries) or WorldCat (access to the world, through the Library of Congress). Online students may request Inter-Library Loan (free) if needed. If you are new to the library check out their tutorial, Search & Find it!

• Many states provide their citizens access to union catalogs and/or subscription databases like PA's Access Pennsylvania, EZ-Borrow, and POWER Library. Popular examples include BadgerLink (WI), INFOhio, NOVEL (NY), NJ State Library/Electronic Resources, MEL (MI), and TexShare. For other states search the web for "virtual library" and the name of your state.

• For education articles, reports, guides, and classroom materials go no further than the Educator's Reference Desk (formerly ERIC): http://www.eduref.org.

• For the greatest coverage of education and library science topics check out the following EBSCOHost databases, available through the Databases page of the North Hall Library:

  • Education Full-Text

  • Education Research Complete

  • ERIC

  • Library Literature & Information Science Full Text

• For articles cited in popular library databases you can often generate a correct citation from the bibliographic record (though note that the default cite is NOT MLA). The trick is to go through the motions of saving or printing the citation, specifying the desired citation style, then clicking OK to create the citation. For a quick demo of MU's most used database vendor see our short video for EBSCOhost.

• NoodleBib—subscription service available via the MU Library. Create bibliographies in various styles, including MLA. Note that you must first create an account via the Library (requires login with your 18-digit barcode number).

• Citation Machine —free online citation creator (you get what you pay for)

• MLA Website (official page)

• MLA Formatting and Style Guide —online guide from Purdue University's world famous Owl resource center for writing and research support

1. In developing a strategy consider how you can limit results to a specific grade level, age, or exceptionality; learning style; date range; or publication type (e.g. research report, lesson plan, article, speech, case study, etc.).

2. Consider using controlled vocabulary in databases that support it (e.g. ERIC descriptor or Identifier). Check for a database-specific Thesaurus. Originally printed as a reference book, the ERIC thesaurus is built into EBSCOhost's ERIC database. This official guide to education terminology identifies descriptors used to index articles, and includes related terms and definitions.

3. Constantly develop sophisticated search skills with advanced search features or alternate sources of information, such as database-specific thesauri; pathfinders (bibliographic essays); or specialized web sites (e.g. US Census Bureau or the National Center for Education Statistics for data).

4. Use advanced search syntax like quotes for exact phrases (e.g. "school library") or Boolean operators (and, or, not). Some databases and most Internet search engines by default search multiple word queries as separate search concepts broadened with the Boolean or. Consequently, a search on school library will find a million items with any instance of the word school (to include vocational, graduate, special, and even the school of hard knocks) PLUS another million items related to any kind of library (to include college, law, special, the Library of Congress, etc.). And keep in mind a "school library" (a place where K-12 students learn) is different from a "library school" (a place to train librarians of all types).

5. To squeeze Internet search engines of all their power click on the Advanced Search. Many of the more popular engines provide simple forms to add limits or filters to produce more accurate and relevant results. If you are familiar with the tricks of a specific search engine use it on their basic search. For example, some Internet search tools (e.g. Google) allow you to mandate inclusion of specific words with a plus sign or exclude terms with a minus sign (e.g. -amazon) to filter out hundreds of hits linked to Amazon advertisements to a book.

6. As of spring 2011 Google no longer supports the plus sign (+); instead, use quotes to force inclusion of specific terms or to prevent the search from automatically looking for synonyms. For example, the quoted term, "cat" would find only the that specific term, not variations like the plural "cats" (either the Broadway play or the animal) or synonyms, like kitty, kitten, feline, etc. (to find synonyms prefix the ~ character before the term; e.g. ~cat).

7. If the goal of your research is to find the best articles possible, then a search strategy that relies heavily on the free Internet and only one all-purpose commercial database (e.g. Academic Search Complete) is not very effective. For education-related topics you should always search ERIC, along with the more comprehensive Education Full-Text. For specialized topics you may wish to consider PsycINFO (covering psychology and learning theory) or Library Literature.

8. An effective search strategy can include seemingly logical approaches that nonetheless lead to unexpectedly poor results. As a coach to novice searchers you should model expert behavior by sharing experiences of what does NOT work, in addition to successes. A good search strategy could consequently include advise on what not to do.

9. Part of a search strategy involves describing which terms and particular tricks are used for which databases. For example, in PsycINFO the preferred terminology for multiple intelligence is cognitive style and intelligence (possibly qualified with grade or age level). You could then add specific intelligences or learning styles to further restrict results. How do you know which term to use? Think Thesaurus!

10. In developing resources for a research project be sure to examine a wide variety of sources, to include a good reference work (encyclopedia, handbook, bibliographic work, etc) or basic informational web site. ERIC Digests are especially excellent starting points, and they are very brief. One excellent and equally succinct online resource for learning theory is the Theory into Practice Database (now integrated with instructionaldesign.org).

11. In evaluating search results, consider the number and quality (relevance) of retrieved hits. This is how you can compare the reliability and accuracy of one resource over another. How would you compare results if the same query resulted in 50 hits in one database and 846,743 in another? What if the tool that yielded only 50 hits consisted of 45 that were highly relevant, while the other displayed page after page of advertisements and low-quality blogs? Think quality, not quantity.

12. If you cite a database family (e.g. EBSCOhost, Gale, Wilson, etc.) be sure to specify the precise database title. For example, the MU Library subscribes to dozens of EBSCO products. So, please cite the particular title (e.g.  Academic Search Complete, Middle Search Plus, EBSCO Animals, ERIC, etc.). Also keep in mind that there are many other resources available to both MU students and teachers in general via state consortia (e.g. Power Library in PA, InfoOhio in Ohio, New York's Novel, etc.).

13. When providing a link for commercial resources note that generic addresses (e.g. http://search.ebscohost.com) neither identify the article nor even the specific database (see above). If a persistent link is available that goes directly to the article, use it. EBSCOhost conveniently provides this link in each citation. But note that this link will work only on a school campus that utilizes IP address recognition. That means you can use the link only if your computer's Internet address belongs to a registered domain on the vendor's customer list.

    If the database can not identify the user's location (i.e. campus) it will not be able to authenticate the user and consequently deny access. In that case, either use a Proxy server if available, or the generic URL to the database family. Similarly, large vendors like EBSCO use multiple servers, so addresses that identify a specific server (e.g. web12.epnet.com) will rarely work, unless that particular server is available (each server can handle only so many connections at a time, and the network automatically assigns the current search session to a specific server).

    For convenience, the MU library provides custom URLs that link a user to a specific product's login. Upon entering these pages you will either go directly to the database, if accessed from the MU campus, or type your MU ID to identify yourself as an authorized user.

14. Note that Wilson's Education Full-Text database (bought by EBSCO in late 2011) identifies many materials that are not cited in either ERIC or related databases. Likewise, the US Dept of Education maintains its own free version of ERIC, at: http://eric.ed.gov (which you might use if your school has no access to a commercial version of ERIC).

Though Internet search tools are popular and easy—and for many information seekers, even sexy—they are not always the most accurate or reliable tools for the job. True information professionals rely on the best tool, not necessarily the easiest or most convenient. Besides accommodating more powerful search strategies and procedures, proprietary databases (i.e. subscription-based) offer many advanced features.

Two of the most important concepts in database searching are recall and precision. Google will gladly retrieve a million items for you ("hits"), but how many of those are any good? On the other hand, an expensive subscription database might find only one or two items with the same search, but those few items might be exactly what you were looking for. Recall in this case is a measure of quantity. Of all the possible relevant articles available, how many will a database or search tool find?

A search for "Harry Potter" in Google, for example, will find over a million hits, most of which revolve around the commercial aspects of the series (fan sites, bookstore advertisements, movie reviews, etc.). If you wish to focus exclusively on the effect of the popular book series on the motivation or reading skills of middle school children, you will find far more relevant materials in a specialized database devoted to education or library science.

In contrast to recall, precision is an indication of how relevant the retrieved items are to your information need. A search that retrieves hundreds of articles about fine plates and saucers will do your students absolutely no good if their class project is about Chinese culture (get it, "Fine China"). A similar example is a search on the sea mammal, dolphins, that yields hundreds of news items about the Miami Dolphins football team. Generally speaking, as precision increases, recall decreases and vice versa (i.e. the more stuff you find, the less likely that it will be very relevant or useful).

Perhaps more important than any other feature, proprietary databases (though not all) provide abstracts and notes about the contents of the material they index. These are brief summaries of articles or other materials, thus offering the searcher more significant criteria for evaluating whether the item might be appropriate. Depending on the database and your specific subscription, you might retrieve either a citation only for the material, a citation plus abstract, or ideally the citation, abstract and link to the full-text. Articles, depending on availability and copyright provisions, may display in one or more formats (PDF, html, or plain text).

Another difference between proprietary (i.e. fee-based) indexes and free search engines is the source for included materials. Internet search engines look everywhere and anywhere in cyberspace, which is why they find so many items (most of which are not very useful). Results can also be skewed in favor of paying customers who spend thousands of dollars to guarantee that their sites or products appear high in the list of "relevant" findings. Library databases and indexes, in contrast, review only a select number of journals and/or scholarly publications. There are far fewer items to be found, thus increasing precision, and they exclude advertisements and content of questionable value (e.g. propaganda, porn, hate-speech or highly provocative sites).

One database that is available as both a free and fee-based subscription service (with value-added features) is ERIC. Developed and maintained for the US Department of Education, ERIC is the premier source for education articles, research, and free classroom materials. As an example of the power of a specialized database, ERIC provides its own Thesaurus of search terms and limiters to specify the types of material available (research reports, data, speeches, tests, classroom guides, etc.). Other databases with their own specialized thesauri include PsychLIT & PsychINFO (psychology), CINAHL (nursing) and Medline (medicine).

To increase both precision (quality) and recall (quantity) library databases offer several special features that free Internet search engines generally do not. The best way to learn about the specific features of databases is to review their help page and example searches. Below are the most common features available in subscription (fee-based) library databases:

Advanced Database Features

The advantage of subscription (also known as fee-based or proprietary) databases is that they search only a limited number of pre-selected sources. The possible universe of information for recall is restricted to mostly scholarly journals and other carefully reviewed publications. To increase precision proprietary databases often utilize their own controlled vocabulary, that is, subject headers or descriptors that remove the element of ambiguity or double meaning with search terms.

These carefully chosen terms are specific or unique to the users of the database, especially in the disciplines of nursing & medicine, psychology and most significantly, education. For example, the descriptor "School Libraries" in EBSCO (or "Libraries, School" in Wilson's Education Full Text) can not be confused with the free-text term, "library school" (or consider the word "china" to refer both to a country and also fine dishes and plates).

Internet search engines generally look for free-text, that is, they indiscriminately search every single word on a web page or in a document for the specified query words or phrase. These terms mirror everyday usage or natural language, the opposite of the precise (some would say arcane) language of indexers and catalogers, those folks who categorize people, places and things to bring order into the universe of search terminology (see the excellent articles by Marydee Ojala and Leise, Fast, & Steckel in your Readings). Searching free-text is less precise than controlled vocabulary, but is far more convenient and flexible since you need not learn obscure terminology or look up terms in a specialized Thesaurus, the dictionary of specialized subject headers or descriptors.

As a very general rule (not always followed) subject headings describe books whereas descriptors refer to articles and non-book materials. One the most famous thesauri in the school library world is the Sears List of Subject Headings, a simplified version (with less technical vocabulary) of the more famous Library of Congress Subject Headings (LCSH), both of which describe and define tens of thousands of terms used to catalog or search for books.

Free-text searching is also necessary when searching for terms that are new, relatively unknown or are otherwise yet unclassified in a thesaurus. Consider, for example, a new fad that just became all the rage within the last month, or the change of a country's name. Print thesauri generally take months to prepare, and only after professional indexers and catalogers debate the incorporation of new terms or agree to revise old ones.

Free-text certainly increases recall but at the expense of precision, since random instances of words out of context in a long document can easily lead to the false impression that a document is relevant simply because 2-3 keywords happen to appear in the same document.

To search by field (e.g. author, title, descriptor, file format, etc.) you can either enter terms in the appropriate drop down option or type the field code in the command line (see the screen shots below). As an example, the search "TI weeding and DE School Libraries" looks for articles where the word weeding appears in the title of articles along with the official descriptor, school libraries, to narrow down the number of possible retrievals. As a good rule of thumb do not search for descriptors unless you are very certain that the term you use is the official descriptor for the specific database you are searching, as a descriptor in one database probably is not the same in another (we'll discuss the use of thesauri further below).