Vanguard University of Southern California
[The primary target audience for this “Literature Search Essentials” document is the undergraduate student confronting the need to do a literature search either as a specific assignment or as part of a larger research project. My particular focus is on psychology students, and key search resources (PsycINFO, SSCI) for research in psychology are described; however, because access to the key psychology search resources is restricted, this document uses an educational resource (ERIC) that is available to anyone with access to the Web.]
The purpose of a literature search for a research article is to identify the existing information sources (including books, journal articles, and Web documents) most relevant to the research question being studied. The results of a thorough literature search will help answer questions like these:
1. What research has already been conducted that is relevant to my research question?
2. What theories have guided the research that has already been conducted?
3. What hypotheses have been tested?
4. How have other researchers defined their variables?
5. What kinds of research designs have other researchers used?
6. How were participants tested in these previous research studies?
7. What participant populations (children, undergraduate students, chronic schizophrenics, etc.) have been studied?
8. What mistakes have other researchers made?
9. What suggestions have been made for future research?
Typically, the sources found in a literature search will be cited in the introduction section of the research article to achieve several goals: first, to identify the area of research by selectively illustrating existing research findings and/or theories; second, to identify the need for additional research; and third, to identify how the research being reported will advance understanding within the specified area of research.
The purpose of a literature search for a research article is not to identify every existing resource related to the topic of the research article, but rather to identify the most relevant resources. For the undergraduate student faced with an assignment, this is both good news (“I don’t have to find everything that’s ever been written about this topic”) and bad news (“How do I know whether a particular resource is relevant enough to include? Relevant in what way?”). The judgment to use or discard a particular resource requires some familiarity with the general area of research and the methods typically employed in that area. This process can begin by reading summaries in textbooks and by reading key articles (including “review” articles that summarize research in a given area). Even students new to the research process can acquire sufficient depth of understanding to properly summarize the key issues in a given field of research.
What are the key search resources in psychology?
1. PsycINFO is a database containing all the information from Psychological Abstracts, the key secondary reference journal, containing nonevaluative summaries (abstracts) of the literature in psychology and related disciplines (sociology, anthropology, education, law, psychiatry, nursing, pharmacology, business, medicine, linguistics, and physiology) from over 1300 periodicals. It contains more than 1.5 million references and adds over 55,000 references each year. PsycINFO does not provide full-text access to the sources within it. However, the abstracts retrieved through PsycINFO are extremely helpful, often providing sufficient information to discard sources that are clearly not relevant to your topic. To get a feel for the PsycINFO search process, visit http://www.psycinfo.com/demo/. Most universities provide full access to the PsycINFO database. Check your university library for access directions.
2. The Social Sciences Citation Index (SSCI) has the unique ability to conduct “cited reference searching” that allows you to start with a specific source (for example, a key article on your research topic) and quickly identify all articles since the publication of your key article that cited your key article. The assumption is that if an article cited your key article, it may be relevant to your research topic. The use of SSCI and PsycINFO together greatly reduces the possibility of missing valuable research. Most universities provide full access to the SSCI database (often through ISI Web of Science: http://www.isinet.com/isi/products/citation/ssci/). Check your university library for access directions.
3. ERIC (Educational Resources Information Center). Because not everyone using this “Literature Search Essentials” document will have access to PsycINFO or SSCI, a third database (ERIC) that is freely accessible will be described and used to illustrate common search procedures. According to the ERIC Website (http://www.eric.ed.gov/), “the ERIC database is the world’s largest source of education information, with more than 1 million abstracts of documents and journal articles on education research and practice.” As with PsycINFO, ERIC provides abstracts, rather than full-text documents.
“So … how do I get started?” For the purposes of this document, I will assume that you have a question of interest that implicitly defines a research topic. Let’s use the following example: “How does praise by a teacher affect the self-esteem of high-school students?” For help in narrowing or broadening a topic, visit the following site: http://www.library.ucla.edu/libraries/college/help/topic/
Let’s try a search using ERIC! Examine the ERIC search form at http://www.eric.ed.gov/. Click on the “Advanced Search” button. You will describe what you are looking for by entering information into “term” search fields. There are three term search fields provided, although you will often use only one or two. Note also that the default search is “Keywords.” You can also search by author or title (among many other search parameters). Note also that in the “Publication Type” box, you can limit the search to journal articles. For the following exercises, check the box that limits results to journal articles. Consider the research question mentioned earlier: “How does praise by a teacher affect the self-esteem of high-school students?” Take a minute to think about possible search terms that could be used to find potentially relevant research about this topic. Don’t worry whether every term you come up with is good or relevant; simply generate a list of 5-10 possible search terms. Compare your list of terms to mine . There is no one “correct” and complete list of terms. By trial-and-error you will discover what terms are productive for specific searches within a specific database. Now, perform a simple search for “self-esteem,” limiting your search to journal articles. Enter “self-esteem” (don’t include the quotation marks), check the “journal articles” box, and click “Search.”
What do I do with the search results? First, notice how many “hits” your search returned. This search should have returned over 4000 hits! The first 10 hits are listed, with a link at the bottom of the list to proceed to the next 10. Find a title that sounds interesting and click the title link to get additional details. Note that all of the bibliographic information necessary to find the article is provided: title, author(s), publication date, and journal citation. Read the abstract, keeping in mind your research question. Categorize the article as “clearly relevant,” “clearly irrelevant,” or “unknown relevance.” Finally, examine the “descriptors” that are listed and ask yourself whether any of them are good descriptors that should be added to your list of possible search terms.
4000+ hits? Help! In scanning the titles of the first 10 hits returned from your search, it should be apparent that not all hits are equally relevant. Some are clearly relevant, some are clearly irrelevant, and others may or may not be relevant. Let’s see if we can better focus the search on our research question, hopefully resulting in fewer (less than 100) hits that include a higher percentage of clearly relevant hits. One way to focus a search is to use two search terms and the “AND” operator. Enter “self-esteem” as Term 1 and “praise” as Term 2. Notice that “AND” is selected as the default operator. The use of the “AND” operator restricts the outcome of the search to articles that include both “self-esteem” AND “praise.” In other words, the use of the “AND” operator is used to narrow searches. Perform this particular search (continue to limit your search to journal articles) and notice that the number of hits is reduced to a much more manageable number of articles. In addition to the narrowing of the search, the results obtained should be more likely to be relevant to your research question.
Organizing information from a search. Whenever the title of an article returned from a search sounds potentially relevant, examine the abstract. If after reading the abstract the research still sounds potentially relevant, you should decide how you will organize the information you are gathering. You may choose to print the abstracts of all the articles you find that are potentially relevant. Alternatively, you may create a Word document into which you paste the abstracts of relevant studies. It is essential that you retain the journal citation information so that you will be able to find the full article later.
Great! So now I’m done? Au contraire, Pierre! To this point, you’ve performed a couple of searches in one database using a couple of search terms. You’ve found some good resources that are clearly relevant to your research topic. Remember that your goal in performing a literature search is to identify all the resources that are clearly relevant to your research question. There are at least two reasons why you are likely to not yet have achieved that goal: (1) There may be relevant research that is indexed in the database by different descriptors. Therefore, you need to try many searches using different combinations of descriptors, noting which descriptors return relevant results; (2) Using the same search terms in a different database may result in additional relevant articles. For example, although there is some overlap in coverage between ERIC and PsycINFO, they have different foci. It is for this reason that researchers use multiple search strategies (for example, using both PsycINFO and SSCI) to reduce the likelihood of missing relevant research.
Try some more searches! Now that you’ve performed a search for “self-esteem AND praise,” try some more searches: “self-esteem AND adolescents,” “self-esteem AND teachers,” “self-esteem AND teacher,” “self-esteem AND reinforcement,” “praise and adolescents,” “self-esteem AND praise AND adolescents.” You may notice that some of the same articles appear in multiple searches. This is good! They are likely to be relevant to your research topic. Also, some searches are likely to return zero hits. Not to worry! Zero hits may result from misspelling a search term, narrowing the search too much (“self-esteem AND praise AND adolescents AND teachers”), or using a search term that is not used by the database. Don’t conclude from a search with zero results that there is no relevant research! Try other search terms.
I want online full-text articles! More and more full-text articles are becoming available through various databases; however, it is important to emphasize that the purpose of a literature search is not to identify the available online full-text information sources most relevant to the research question being studied. It is essential that you identify all of the resources most relevant to your research question, regardless of whether those resources are easily available online or not. It is almost certain that some of the articles that you identify will not be easily available online in full-text format. Your task is to obtain all of the most relevant articles. This may involve trips to local university libraries, requests for articles through interlibrary loan, and requests for reprints from authors.
If the full-text articles aren’t available online, why can’t I just use the abstracts? The abstracts are extremely useful in identifying potentially relevant research and in identifying research that is not relevant and therefore does not need to be pursued. Abstracts, however, do not include sufficient detail about the theoretical rationale for the research, the methodology employed, the results obtained, or the implications of the findings for future research. Therefore, you should seek to obtain the full articles of the resources that you have identified as clearly relevant to your research question. Details from the articles you obtain through this process will likely help you design a better study. You may learn from the mistakes of others; you may learn what populations have yet to be studied; you may discover specific measures used by others that you too can use.
O.K. Back to the search process. It’s time to practice! For the following items, use ERIC to do whatever is needed to answer each question. When you’ve completed your work for each item, click “Check your work” to compare your response to mine!
1. Research question: Is physical appearance related to self-esteem? Using ERIC, find the abstract of a relevant article that studied Japanese adolescents. Check your work.
2. By looking at the abstract of the article found in question #1, what additional keywords would you consider using for future searches (perhaps in other databases)? Check your work.
3. Research question: What is the relation between prayer and physical health? Using ERIC, identify the two most relevant articles. Check your work.
Where can I go for additional information on literature searches?
1. Beginning Research in Psychology: University of Melbourne http://dozer.infodiv.unimelb.edu.au/SuRe/subject.php?subject=Psychology&field=beginning
2. Writing a Psychology Literature Review: University of Washington http://depts.washington.edu/psywc/handouts/pdf/litrev.pdf
3. Summarizing a Research Article: University of Washington http://depts.washington.edu/psywc/handouts/pdf/summarizing.pdf
4. Library Research in Psychology: American Psychological Association http://www.apa.org/science/lib.html