Research Skills Tutorial

Main Page | Modules: 1 | 2 | 3 | 4 | 5 | 6 | 7 | Outcomes

Contents of this Module:

Learning Outcomes
Setting Realistic Expectations
Selecting Appropriate Databases
Developing Search Terms
Combining Search Concepts
Advanced Search Features
Interpreting Search Results

Learning Outcomes

By the end of this module, students should be able to:

1. Recognize content differences among database types: bibliographic, full-text and numeric.
2. Identify the main concepts of a research topic.
3. Develop a list of appropriate search terms.
4. Develop a search strategy appropriate to the selected database.
5. Use Boolean operators successfully to combine search concepts.
6. Understand phrase searching and truncation.
7. Recognize arrangement of search results. (alphabetic, date, relevancy ranking)

Introduction

Understanding basic database content, structures and functions will lead to more effective database searching. Many databases used for research function in similar ways, even though the content may vary widely from one database to another. The experienced researcher knows what features to look for and can approach new databases intuitively, resulting in confident and successful database searching.

Setting Realistic Expectations

It would be most convenient to be able to search only one database to conduct all research. We would be able to enter a single search query and have all the results delivered to the computer screen at once. More importantly, we would need to learn how to use only one "search engine." (A search engine is the software that determines how and where a search query is entered, how search results are refined and, to a lesser extent, how the database screens appear; how results are displayed, printed or downloaded, etc.) But no single, comprehensive database of all information in the universe exists today. And, different databases may use one of a variety of different search engines. So the researcher must align expectations for the types of information needed (as discussed in Module 1) with the types and range of appropriate databases available.

Selecting Appropriate Databases

Just as there is no single, comprehensive database, not all databases are equal in terms of their content. There are three basic types of databases useful for research: bibliographic, full-text and numeric.

• Bibliographic databases are used to identify sources of information the content of which can be found elsewhere, not in the database itself. These databases simply list citations or references to books, periodical articles, government publications, and videos. They may or may not provide specific information on where to find the item. An on-line library catalog is one example of a bibliographic database. It lists the books, periodical subscriptions (not the articles), government publications, videos and other materials that an individual library owns and it includes detailed location information.
  
  A periodical article index database is another example of a bibliographic database. This type of database lists articles published in a group of periodicals selected by the producer of the database. The periodicals themselves may or may not be owned by any individual library. This type of database simply identifies the articles; it does not specify location or availability.

• Full-text databases provide more than citations; they provide the complete text of the information, and increasingly often include graphics, as well. Full-text databases may make historical documents available on-line, such as the US Constitution. Others supply current information, such as periodical articles or daily entries from the Federal Register. Not all full-text databases are comprehensively full-text, however. Few periodical article databases provide the full-text for every article listed. The primary reason for this is that some periodical publishers are reluctant to relinquish their copyright ownership rights, by allowing their articles to be made available on-line, for any fee.

• Numeric databases supply quantitative data. A few examples of the data available in numeric databases are company financial information, census statistics and national labor data.

The experienced researcher knows that not all pertinent information can be found on-line. All researchers should resist the temptation to rely solely on that which is conveniently found on-line. The articles or other information found on-line may or may not be the most useful. The information should support the topic focus and assignment parameters, as outlined in Module 1, regardless of the physical format. Periodical articles identified but not available in full-text on-line might be found in a local library, and otherwise will likely be available through Interlibrary Loan. (see Module 4, "Finding Periodical Articles")

The successful researcher, having used the research strategy of Module 1, will have identified the type(s) of information needed. Now, the researcher must select the database(s) or print materials that will provide the type(s) of information needed. One would not search a library catalog database to identify periodical articles, because library catalogs do not list individual articles; similarly, one would not search a numeric database to identify books on a topic. To determine the appropriateness of a database, the researcher should identify its content. There are several ways to identify database content: ask a librarian, professor or colleague who is familiar with the database; seek written descriptions of the database, such as information found in the "About" or "Information" section of a database's front page; or, simply experiment with the database.

Developing Search Terms

The choice of appropriate search terms is crucial to creating an effective database search. The first step in developing a list of search terms is to break the research topic into all its constituent concepts. For example, the topic, "the role of college-level leadership classes in developing leaders," can be broken into the following concepts:

• leadership
• education
• higher education
• effectiveness

Notice that the words representing the constituent concepts do not simply mirror the words in the topic. The researcher who remains stuck on specific words to describe a topic or an assignment severely limits chances of developing an effective search.

The next step is to make a list of alternate terms for each concept. The easiest way to do this is to consult the database thesaurus, or list of subject headings for the database, if one exists. For example, libraries use the Library of Congress Subject Headings (LCSH) to assign subject index terms that describe items listed in library catalog databases. All the items in a library catalog on the same topic are assigned the same subject heading word or phrase. So, the researcher who performs a subject search using an LCSH subject heading identifies all the items in that library catalog on that topic.

It is also advisable to note any additional terms listed in the database thesaurus. These terms may represent broader, narrower or related subjects, all of which may be useful in refining a search. In the event that a search query has zero results, a broader subject heading may be used with the expectation that a chapter or section of an item on that broader topic might discuss the more specialized topic. For example, if a search on the topic, "attention deficit hyperactivity disorder," resulted in too few items, perhaps a search on the broader topic of "behavior disorders in children" would provide some useful information. Conversely, if the first search results are too numerous, a second search using a narrower search term might produce fewer and more relevant results. "Related" subject terms have less direct relevance, but they may be used if all initial search queries are unsuccessful.

Finally, the researcher may have encountered other alternative terms or synonyms in the background readings or in other earlier experiences with the topic. These alternate terms will be needed for searching those databases that use no subject headings. They can also be useful in constructing "keyword" searches. A keyword search is often the next option when a subject search has zero results. Keyword searches generally match on words appearing in titles, abstracts or in the content itself. Keyword searches, it should be noted, may produce less relevant results, particularly when the words that create the match appear in different parts of the item. For example, if one matched word appears in the title and another matched word is in the abstract, these words may not have the expected relationship, and the identified item may have nothing to do with the topic being researched.

Combining Search Concepts

Armed with lists of subject headings, keywords, related subjects and synonyms, the researcher must now determine how best to combine these terms to produce the desired results. One cannot simply assume that the terms can be listed one after the other in a single search query. Different databases combine search terms in different ways, but there are common underlying structures at work.

There are three basic ways to combine search concepts. Taken together, they are called "Boolean operators". The three Boolean operators are represented by the words, "and", "or", and "and not". The "and" operator combines terms in the sense that "this term and also that term" must be present. Using the topic cited above, "the role of college-level classes in developing leaders," the search query could appear as:

leadership AND education AND college AND effectiveness

All four concepts are expected to be present.

Boolean operators can be graphically demonstrated by the Venn diagram of two slightly overlapping circles. The "and" operator appears as:

AND

The results of a database search combining terms using the "and" operator are represented by the intersection of the two circles, where one circle represents one concept or term and the second circle represents the other concept. As can be seen, the "and" combination results in fewer items.

Some databases use the word "and", others may use a plus sign (+), and still others may use the ampersand (&) to perform the "and" operation. Even more confounding, databases using the CARL search engine, such as the Regis University Libraries catalog, imply the "and" operator between words listed in a single search query or with words added to the initial search query. The researcher who realizes this can build a more effective search. Consulting either librarians or database help screens can often shed light on the question of whether and how Boolean operators can be used in the database at hand.

The "or" operator combines terms in the sense that "either this term or that term" may be present. One effective use of the "or" operator is in combining synonyms or alternate terms. Again using the sample leadership topic, the "or" operator could be used to combine:

college OR university OR higher education

The "or" operator widens the net of the search, so to speak. The Venn diagram representing an "or" operation appears as:

OR

The results of a database search combining terms using the "or" operator are the totality of both circles. The "or" combination obviously produces larger results. Most Internet search engines implicitly perform "or" combination searches of words listed one after another in a single search query.

The "and not" Boolean operator eliminates one search term entirely from the search results. The "and not" search combination can be used effectively to delete one small subset of a larger topic, e.g., "computers AND NOT laptops". The Venn diagram representation of the "and not" operation appears as:

AND NOT

Advanced Search Features

Many, but not all, databases will offer the opportunity to indicate that the words entered together in a single search query constitute a phrase; i.e., these words are to be taken together, in this particular order, with no intervening words. This feature is particularly useful when the database implicitly performs a Boolean operation between words. It should be clear from the earlier discussion that there is a significant difference in results between these two searches:

leadership AND management
(leadership management)

Some databases create phrase searching by placing parentheses around the terms. Internet search engines signify phrase searching by enclosing the terms with quotation marks. Again, help screens or librarians can provide guidance on constructing a phrase search.

Another advanced searching feature is called "truncation". This feature allows the searcher to pull together into one group all words that begin with or contain the same word root. This is an advanced way of accomplishing a complicated "or" operation. For example, the researcher who wanted to combine the words "educate" or "educated" or "education" or "educator", could enter: "educat?" instead of combining all four terms with the "or" operator. In this example, the question mark indicates that any letter(s) and any number of letters may follow the "educat" root of the word. Different databases may use different symbols to indicate truncation. The symbols themselves are sometimes referred to as "wildcards". Some examples of wildcards include the question mark (?), dollar sign ($), exclamation point (!), and asterix (*). Again, consulting a librarian or the database help screens may help.

Interpreting Search Results

Knowing or being able to recognize how search results are arranged can assist the researcher in interpreting the results of a database search. Some databases arrange search results in alphabetical order, either by author or title. Others use a reverse chronological or publication date listing, generally with the most recent item at the top of the list. And, some databases now offer a "relevancy ranking" arrangement option.

A relevancy ranking is derived from an analysis of how often and in which parts of the database item the search term(s) appear. An item in which the search term appears in both the title and first paragraph would have a higher relevancy ranking, for example, than an item in which that term appeared only once in paragraph five. Researchers should exercise caution in relying too heavily on relevancy rankings, however, as the "relevancy" is determined by the database developers and their search algorithms rather than what the researcher may consider to be relevant.

But, knowing how and why search results are arranged as they are can be very valuable in interpreting the results of a database search quickly and efficiently.

Go to Module 3