library / information services

About SMS
Midwifery Education Program
Extension and Doula Education
California Challenge Process
Library and Information Services
News and Events
Marketplace
Research/Advocacy
Job Announcements
Search

Contact Us

Skills List:
Using Databases, Online Indexes, And Search Engines



This page reviews skills that are useful for searching a wide range of online databases.   These skills are essential for complex, difficult, or in-depth searches and can also come in handy for quick searches in databases or on the Internet.

This is not a comprehensive list of skills, nor are the discussions of these skills comprehensive.


 
  • WHERE TO FIND INFORMATION
  • EVALUATING INFORMATION SOURCES
  • FINDING TERMS TO DESCRIBE YOUR INFORMATION NEED
  • DATA FIELDS
  • CONTROLLED VOCABULARY
  • USING A THESAURUS
  • BOOLEAN LOGIC
  • NESTING
  • PHRASE SEARCHING
  • TRUNCATION
  • LIMITS
  • ITERATIVE SEARCHING
  • TROUBLESHOOTING
  • RECORD KEEPING
  • GETTING HELP

  • PLACES TO LOOK FOR INFORMATION

    Please note that this list includes some of the most important sources of information but it is not comprehensive.
  • Journal databases See http://www.seattlemidwifery.org/library_journaldatabases.htm
    for a list of databases and access information.
  • Book catalogs, including library catalogs and commercial sites such as Amazon.com. See http://www.seattlemidwifery.org/library_booksandvideos.htm.
  • Clinical guidelines. See http://www.seattlemidwifery.org/library_clinicalguidelines.htm.
  • Conference abstracts.
  • People: colleagues; librarians; and researchers or experts working in a subject area.
  • Your public library: check the library Web page to see what resources are available.  The King County and Seattle Public Libraries, for example, allow members Internet access to a number of databases and other resources that are not freely available on the Web.  In addition to offering access to resources, libraries also collect and organize links to information sources you may not otherwise find.
  • Data repositories: toxicology information; laws; statistics; etc. Many of these are available on the Web.  See the Reference Tools section of the SMS Library Resources page.
  • Citation databases: Has anyone cited your 1993 article in the past 2 years? Check with Science Citations and Social Science Citations.
  • Browsing: library shelves; journal tables of contents; web pages; and other sources
  • Internet search engines: searching the Internet can alert you to new angles on an issue and sources of information you may not have considered.
  • Organizations: Is there an organization devoted to studying this issue?
  • Textbooks.
  • SMS Senior Papers and vertical file.

  • EVALUATING INFORMATION SOURCES

    Two sites that cover this issue:
  • DISCERN is a brief questionnaire that assists users with assessing the quality of written information on treatment choices for a health problem. The instrument can also be used by authors and publishers of information on treatment choices as a guide to the standard which users are entitled to expect.
  • ACP Journal Club criteria for article selection
  • An informative article on "Missing Information" in the biomedical literature is available in the SMS Library:
  • Aronson, Barbara. Missing Information. WHMS Liaison: Newsletter of the WHO Library and Information Networks for Knowledge. 10(1):4-7.
  •  Some major elements to consider when evaluating the credibility and usefulness of an information source:
  • Currency: how recent is this source?
  • Objectivity: is only one viewpoint considered? Are the claims justified by the evidence presented?
  • Attribution: does the author clearly identify sources?
  • Clarity of presentation: are the methods and goals stated?
  • Validity: are standard research protocols adhered to?
  • Authority of author/s: what is the author's affiliation, background, previous publication history, educational history?
  • Authority of publisher: reputation of sponsor or publisher; if a journal, is it peer-reviewed? li>Conflict of interest: what are the affiliations and possible motives of the authors, publishers, website hosts, etc.
  • Approach: is this a scholarly publication? A publication geared toward consumers?

  • WHERE TO FIND TERMS TO DESCRIBE YOUR INFORMATION NEED

    Words are the keys that unlock the information in bibliographic databases.   Your search will be much easier and more effective if you identify the most useful search terms for your topic and have insight into how to best use them in online databases.   Unfortunately, for most searches, it is very difficult to know whether you have found all the important terms or have used them in the right way.   It is also easy to lose track of which words you've used when searching a particular database.   Here are some strategies to help.
    1. Establish a place to keep records of the words you use in your search. (See the Keeping Records section of this page).
    2. Have a list of search terms written down before you begin searching in a database.   Have an idea of which terms may be most useful as key terms and which as subject terms
    3. As you proceed in your search, keep a record of new terms that you find.
    4. Topic aspects.   Most topics have several aspects. Analyze your topic to identify these aspects.   For example, the topic, educating families on genetic testing during pregnancy, has at least three main aspects: education (or perhaps informed consent), parents or pregnancy, and genetic testing.  
    5. Think about the relationship between your topic and the database you are about to search.   For example, if you were doing the above search in a bioethics database, you may not need to use the word, "ethics" -- or you might need to make it more specific.
    6. While searching, make a note if certain words tend to appear in records that are off your topic.   For example, you may be searching with the term "teaching aids" and notice that many irrelevant records on HIV appear.   The presence of the word, HIV, in a record is a signal that it is an article on AIDS.   Therefore, you will may want to use the NOT feature of the database when searching on that term.   "hearing aids" NOT HIV.   Along with a list of key and subject terms, then, you may also want to keep a list of "not" terms.

    Here are some ideas for where to find search terms:

  • Bibliographies of articles and books.
  • Book indexes.
  • Thesauri. Some databases, such as Medline and CINAHL, have their own.
  • Medline's Medical Subject Headings (MeSH) list is a thesaurus that deserves special mention. It is searchable through PubMed.
  • Subject terms assigned to articles in PubMed and other databases.
  • The Internet: do a search and see what you find.
  • Ask a librarian.
  • Brainstorm with the terms you already have:
  • Synonyms
  • Related terms (for example, the terms belief and perception and trust.)
  • Broader terms (for example, metabolic disorder is a broader term than hypothyroidism).
  • Narrower terms (for example, midwife is a narrower term than healthcare provider).
  • Geography, population, and time (for example, pregnant adolescents in U.S. urban areas in the 1980's)

  • DATA FIELDS
    AKA: Search Fields What are data/search fields?
    Search fields are like cross sections of a database index. When you limit your search to authors, or to article titles only, for example, you are searching within specific fields.

    Most bibliographic databases offer a user-friendly "default" search box that shows up on the first page you see when you open the database. Typically, the search engine or database will look for the terms in several fields at once, often the title, abstract, and subject fields. For simple searches, this user friendly default can be all you need. In complex searches, it can be essential to search specific fields.

    Why are search fields used?
    In a perfect world, you would be able to enter your terms into a search box and retrieve citations for all the relevant articles on your topic with no irrelevant articles included. Unfortunately, finding all the relevant articles is not so easy!

    If your terms are general enough to collect all records on a topic, you are likely to also retrieve many off-topic records. On the other hand, if you make your terms specific enough to exclude irrelevant records, you will also miss many that are relevant. Searching specific fields helps compensate for this Catch-22. Using fields allows more precision because it allows you to access subsets of the index. For example, if you were searching for an article authored by a person named Black, it would be much easier to search for the word, black, in the author field only so that you would not also retrieve citations of articles that have the word, black, in the title.

    Check with the help file in each database for instructions on how to use search fields.

     

    CONTROLLED VOCABULARY

    A controlled vocabulary is an official list of terms that indexers use when assigning subject heading terms or in order to maintain consistency in the way authors or other data are cited.

    For example, the term used for St. John's Wort in Medline's subject headings is hypericum. Using this term allows the person searching to pick up all articles on this substance regardless of how the authors referred to it.

    Some reasons to use a controlled vocabulary field:

    • When you are doing a comprehensive search and want to pick up all the articles you can, regardless of which synonym the authors use to refer to your topic.
    • Each word in a controlled vocabulary is assigned only one meaning. This can be invaluable when the words describing your concept can have several meanings.
    • When you want to understand your topic or formulate your search better. Controlled vocabulary thesauri often provide clarification of terms and may provide additional terms related to your concept that you hadn't thought of.

    Some fields, such as subject and author fields, commonly use a controlled vocabulary. Other fields, such as title and abstract fields, cannot because the information included in those fields originates from the authors and cannot be changed by the indexers. The search words that you enter into a field which is not controlled are sometimes referred to as key words.

    Internet search engines like Google do not use controlled vocabulary because they are searching the Internet.

    To use a controlled vocabulary, look at the thesauri for that database (PubMed and Cochrane both use the National Library of Medicine's MeSH, which can be accessed from PubMed.) or look at the subject headings on articles you have already found on a topic from that database. A short-cut strategy is to do a quickie search in the opening search window of a database and then to look at the subject headings listed for the resulting articles.


    USING A THESAURUS

    A thesaurus for an index or bibliographic database differs from the thesaurus you may be familiar with using as a writing tool. Both are resources that allow you to identify alternative terms for a given concept. However, they have different emphases. A standard thesaurus focuses on helping the author widen her choice of words. In contrast, a database thesaurus can be most useful because it narrows the users choice of words for a concept so that she is able to identify the official subject term used for that concept in the related database.

    For example, there are several words that refer to the condition of pregnancy, including parity, gravity, pregnancy, expecting, etc. There may also be alternate spellings for words, including foreign language equivalents. A database thesaurus will list all related terms and alert the user to the subject term that is always assigned in that database to documents that deal with the concept of pregnancy regardless of the language used by the authors of the documents.

    The terms given in a database or index thesaurus are subject terms. You use them when you are searching for documents in the subject field of a database.

    The thesaurus used in PubMed is called MeSH, which stands for Medical Subject Headings. This is probably the most important thesaurus to know about, as it is used in a number of biomedical databases, including some Internet medical search engines. The MeSh Browser in PubMed is a good place to start becoming familiar with this tool.


    BOOLEAN LOGIC

    Boolean logic, when it is used in literature searching, means that you are using the words "and" and "or" to connect concepts. For example, the search statement [fetal AND positioning], will find only those articles that include both of these terms. The search statement [fetal OR positioning] will find articles that have either or both of these terms.

    Using the boolean connector, AND, narrows the results of your search. Using the connector, OR, broadens your search and gets you more results. Some people find this counterintuitive. The word, "and" seems as if it would be additive and, therefore, more inclusive. However, in literature searching, the opposite is true.

    Imagine that you want to find a job that is both close to home AND well-paying. That would be more challenging than finding a job that is either close to home OR well-paying because there are fewer jobs out there that fulfill both these criteria.

    Some search engines, such as PubMed, require the operators or connectors between words to be capitalized. If you do not capitalize in PubMed, this feature will not work! Read the Help section of the database you are using to find out what is required in your search statements.

    The word, "not" is sometimes considered a boolean operator, although technically it is not.

    Some search engines use symbols for these operators, for example + and - .


    NESTING

    Nesting refers to the use of parentheses to force a search engine to evaluate your query in the order you specify.

    Using parentheses forces a search engine to find articles that match the specifications within the parentheses before it goes on to the rest your query.

    This is useful because search engines have a default order for evaluating queries (usually from left to right). The following is an example of how this might throw off a search.

    Most search engines evaluate queries from left to right.   So, if your input is the following:
            pregnancy AND lactation OR breastfeeding
    you may find articles that are either about pregnancy AND lactation or that are about breastfeeding alone.

    If you include parentheses in your statement as follows:
            pregnancy AND (lactation OR breastfeeding)
    The search engine will evaluate the portion of your search that is in parentheses first.   It will look for all articles that include either word, lactation or breastfeeding.   Then it will find all articles within that set of results that also contain the word or concept, pregnancy.

    The use of parentheses to nest the various portions of your query is standard across most databases.

    Nested statements can get complex.   Here is an example of a more complex nested query:
            ((pregnancy OR pregnant OR gravid) AND (lactation OR breastfeeding OR "breast feeding"))
           AND siblings


    PHRASE SEARCHING

    A phrase is two or more adjacent words that stay in the order in which you typed them.

    The term "vaginal birth after cesarean" can be searched in two ways:

  • as a string of words that can be rearranged into any order. Such a search will pick up any document
    that contains both of these words in whatever order they appear.
  • as a phrase.
  • If you search for the words, "vaginal", "birth", "after", "cesarean" on the Internet or in a large database and do not specify that you want them searched as a phrase, you are likely to retrieve many documents that are not about the concept, VBAC. If you search for this concept as a phrase, however, your results are likely to be much more manageable and on-target.

    Check the help files in the database or search engine you are using to find out how to tell it that you want your terms to be searched as a phrase.

     

    TRUNCATION

    Truncation, also known as stemming, directs a search engine to find all words containing a specified part of a word and to ignore other letters that may be in that word.

    Truncation is helpful when you want to search for all forms of a word. For example, the search term, "midwi*" in a database (like PubMed) that uses the * for its truncation symbol, will search for midwife, midwives, and midwifery.

    Truncation symbols vary among databases. Common symbols include "*" and "?" Check the help files of the database for its conventions.

    When using truncation, be aware that some word stems can be parts of words that you may not want to retrieve.   For example, plan*, will find instances of the words plan, plans, and planning.   It will also find plant, plantar, plane, planet, and planus, among other words.


    LIMITS

    Limits are criteria applied to a search that narrow the results of that search. In a broad sense, anything that narrows your search, including the use of search fields, is a limit. In common use, limits are criteria, such as the following:

  • Date of publication
  • Type of publication (case study, review article, letter, etc.)
  • Population group studied (i.e., age; whether human or animal)
  • Language of article
  • PubMed allows all of the above limits in addition to others. Some search engines have limits capabilities that will allow you to specify a date range, to ask for only web pages with images or videos, to specify geographical location of the pages, or to ask for only those pages of a certain domain (.com, .org, .gov, .edu, etc.)

    Limits can be an invaluable part of a complex search. Check with the help files of the database you are working with to learn what limits are available.


    ITERATIVE SEARCHING

    Information you gather while you search can be cycled back into that search. For example, while you search, you may find:

  • New subject terms and key words. (Look at the subject terms assigned to articles that are on-topic; they may lead you to other articles.)
  • A more in-depth understanding of your search topic and how it is related to other topics.
  • Authors who are prolific on that subject.
  • Articles that are not retrieved by your searches in databases can sometimes be found in the bibliographies of books and other articles.
  • In addition, some search engines, like PubMed, include a "find more articles like this one" feature. And citation searches in Science Citation Index and Social Science Citation Index allow you to identify articles that cite those you have already found.


    TROUBLESHOOTING

    Too few results?

  • Use "or" to search for synonyms of a term
  • Choose broader (more general or inclusive) terms
  • Use truncation to pick up varying forms of a word
  • Search more fields (don't limit search to just the subject field, for example.)
  • If you are searching in a controlled vocabulary field make sure you are using approved terms from the thesaurus.
  • Check to see if you are using the correct conventions for that database. For example, PubMed requires boolean operators to be all in caps. Methods for indicating truncation and phrase searching vary from database to database.
  • Check your spelling.
  • Use iterative searching.
  • Too many results?

  • Use "AND" to narrow your search.
  • Use "NOT" to exclude terms.
  • Search only specific fields, such as subject heading fields.
  • Use limits, for example, date, population, language, type of article, etc.
  • Search for phrases.


  • RECORD-KEEPING

    Recording your search steps in complex searches can save you a lot of time. The most important information to record during a complex search is:

  • What databases you used
  • What terms you used in those databases (i.e., subject headings, key words, authors, etc.)
  • How you used those terms in your queries. In other words, it is best to record your actual search statements. For example, "pregnancy complications AND social support".

  • GETTING HELP

    Use the help files of a database if you are not familiar with its conventions. It is also recommended to ask for a librarian's help.

    lady_sms logos

    TOP