Searching by Topic

We will begin this section with a brief overview of the strategy. Then we will illustrate these principles with the example outlined in the previous section. This is the most extensive section in this tutorial, because this is the most commonly used search technique.

Identifying Key Terms

How do you search through the mass of information in the various indexes to find what you need for your project? Like all research, library research needs a clear problem statement. Before you begin your library research, be sure you know what you are looking for. You will need both a clear statement of the problem and a list of the key terms that would identify relevant research papers. These key terms will be used in your search to enter the various indexes and to guide you through journals and annual reviews. 

The Thesaurus of Psychological Index Terms, published by the American Psychological Association, can be very helpful here. It not only lists the index terms that are used in Psychological Abstracts, but also lists the cross-referenced terms that might also identify relevant literature on a topic. Look up your topic in the Thesaurus, and it will list the index terms under which you will find appropriate references. Use those as your key terms to conduct your search through the various citation indexes. Similar publications listing index terms are available for other abstract services, such as ERIC. 

You can also find likely index terms just by looking through the article or articles that you initially discover on a topic of interest. Index terms are not mysterious. They tend to be the main terms that show up in article titles and in the abstract. For example, if you are studying memory loss in Alzheimer's disease, the key terms would be "memory loss" and "Alzheimer's." The Thesaurus may suggest additional terms, such as amnesia, memory dysfunction, memory performance, or memory, or terms like dementia, pre-senile dementia, and so on. 

Some of these terms will be useful, whereas others are unlikely to be useful. For example, memory might work well, although it is a broad category and may identify too many articles. We will see examples of that shortly. Amnesia involves memory loss, but the type of memory loss found in Alzheimer's Disease would not be found under the key term of amnesia. Of course, you may not know that until after you have conducted your library research, but that is OK. Library searchers are, by there very nature, ongoing. You start with some basic terms and ideas and refine both your ideas and the later searchers as you learn more about the topic.

Computer Searches

Fortunately, many libraries now have extensive facilities for computer searches, saving a great deal of time and making your search more complete and systematic. PsycINFO, Books in Print, ERIC, Medline, and dozens of other indexes can be accessed through computer terminals in the library and elsewhere. 

Consult your reference librarian to learn how to use these systems. The procedures for doing computer searches vary from one system to another, but most systems operate under the same general principles. All systems have records (containing all of the information about a publication) and fields within each record. Each field contains specific information about a publication (e.g., title, author, journal, keywords, and abstract). You can search a specific field or all fields to find information. 

For example, if you know that relevant research was published by "John Smith," you can search the author field for that name, although for your sake, we hope that the name is more distinctive than John Smith. Searching for authors with common names like smith can be very tedious. If you are searching for an author with a common name, make sure that you get the author's full name and middle initial to help in the search.

There are many strategies for computer searches of the research literature, but most people start by searching for keywords. Most indexes will have a keyword field, where a small number of descriptive terms summarize the main content of the paper. Searching this field for certain topics is adequate for many searches. But, because computers are so powerful and fast, it is easy to search everything (title, keywords, abstract, etc.) to find potentially relevant papers. Most computer databases automatically search the entire record for key terms.

Entering keywords to narrow a search is an art that requires considerable logic and some practice. In most systems, entering the term "child" will identify any article that has the terms child, childhood, children, and child's. Similarly, entering the term "schizophren" will identify articles that have the terms schizophrenia or schizophrenic. A term like "schizo" will identify the same articles, but will also include articles that have terms like schizotypal, schizotaxia, or schizotype. 

Make sure that you know the syntax of the system that you are working on. In some systems, you must put in a wildcard character (e.g., schizophren$; the $ is a commonly used wildcard character) to tell the system that you want any article that has that keyword as a stem.

Some keywords will identify several thousand potential articles, while other keywords identify no articles. If the list is too large, you will have to reduce it to make it manageable. You might refine the list by limiting it to only those papers published in a given time period--perhaps the last five or ten years. But you probably want to also narrow the list by further restricting the topic. 

To narrow the topic you must understand two Boolean operators--AND and OR. If you are interested in childhood fears, you could use the term "childhood fears" in your search. However, such a search would miss any article that did not use that specific phrase (perhaps using a phrase like "fears common in childhood" instead). Using the Boolean operator AND to search for "fear AND child" would likely give you a more complete search. Any article that had both the term "fear" and the term "child" in its record would be identified. The operator AND will narrow a search by requiring that two conditions be met. 

In contrast, the operator OR will broaden a search by identifying articles that meet one or more of several specified conditions. For example, searching for "frontal OR parietal OR occipital OR temporal" would identify any article that mentions one or more of these lobes of the brain.

Many databases give you the option of specifying complete searches using the AND and OR in a single step. When you are familiar with complex database searches, this is the simplest way to get what you want right away. However, most databases offer you the option of searching step by step. 

For example, you can search for "child" as the first step. If you want to include children from 5 through 15, you might want to search for "teen" and "adolescen" in two more steps. Then you can use a COMBINE function, which allows you to combine two or more searches using either the AND or OR functions. If you combine the searches for "child" OR "teen" OR "adolescen," you will get a set that includes any one of these terms. You could then search for terms like "fear," "phobia," and "terror," which you could then combine using the OR function to get a set that includes any mention of these various terms for fear. Finally, you want to use the AND function, combining these two large sets to get a set that includes both a mention of children or teens and fear. This kind of process will be a lot easier to follow and understand when we walk you through an example shortly.

As you will see in our example, it is often a good idea to do the Boolean operations as separate steps, so that you can see how effectively you are narrowing the search. For example, if you were interested in negotiation styles in labor contracts, you might do separate searches for both of these terms. The separate searches would give you an idea if you are too narrow. If you search for "labor contracts" and only get three articles, chances are others are being missed, because they are using other terms. In this case, you can search for alternative terms and then use the Boolean operator OR to combine them, which will produce a list that uses any of the terms referring to this topic. When you are ready to narrow the topic, you can use the Boolean operator AND to merge the searches, identifying only those papers that discuss both topics.

Once you have narrowed the search to an appropriate size list, you can enter commands to display the records (titles, authors, sources, and abstracts). By reading the titles and/or abstracts you can select those that appear to be most appropriate for your topic and eliminate the rest. At this point, you will have a screen display of a fairly refined list of appropriate articles. Using the print command will give you a printout of your selected references and their abstracts. 

Now the real work begins. Your list of references is just that--a list. What remains is locating each of the papers, books, and chapters, reading them, integrating the information, and writing your paper. The locating part gets easier every year. When this tutorial was first written, about 5% of the research articles needed by the authors for their papers could be accessed online. Now it is approaching 90%. It is literally possible to complete an entire library research project from home.

As you begin to read the material you have identified, you will eliminate more of the cited work as not appropriate or useful, and you will also find more references in the various reference sections of each paper you read. Listed in the next two sections are useful strategies for identifying other relevant papers.

Our Example: Explicit Memory in Anxiety Disorder

As you will see in our example, the process of library research uses all of the techniques that are outlined in this section and in the next two sections. We do not always use the techniques in the same order, and we typically will use one technique to identify relevant work, use another to identify still more useful information, and then use the most recent information to go back and use an earlier technique in a more refined way.

In our example, we started with an article that caught our eye and our interest (Becker et al., 1999). Our first task is to carefully read the article. Don't worry about the fact that parts of the article will be over your head. Unless you are already an expert on the topic of a research area, parts of any article are likely to be over your head. Do not ignore what you do not know. Keep track of your questions, because your later research and reading will likely answer many, if not all, of them.

You have two external resources and one internal resource when you start your research on this topic. The internal resource is your interest in the topic, which can make the task of learning about this new area an adventure. Your two external resources are (1) the paper, and (2) the reference list for the paper. The paper will give you some insights into the topic, although no single paper will give you all of the possible insights. The reference list provides a wonderful starting point for your library research.

The reference list from the Becker et al. (1999) article on explicit memory in anxiety disorders is a wonderful starting point for your library research. Because the article caught our eye, the ideas expressed in the article are inherently interesting to us. 

As you read the article, you will find references to particular theories, studies, or procedures that you will want to know more about. By noting those references and consulting the reference list, you can identify several relevant articles right away. Many students stop here, which is a terrible mistake. The references identified in this manner are only a subset of the relevant references, and more importantly, references obtained in this way are an inherently biased subset.

After you have marked the references you know that you already want to read, look over the other references. You are looking for two things--the words that are typically used to describe particular research in the titles of the articles, which will be used in your search of the research literature, and major contributors to the research and theory in an area. These are people who are quoted frequently in the article. 

Be aware that most articles have multiple authors and the most relevant author from the standpoint of being a major contributor to the field will not always be the first author. Many faculty publish extensively with their students, listing the students as the first author. Our brief review of the reference list suggests several people who have contributed numerous papers on the topic and given us an idea what key terms we might use in our search.

We will rely on our computerized databases as much as we can, primarily PsycINFO, which catalogs the works of greatest relevance to psychology. Be aware, however, that some topics in psychology will require the use of additional databases to get a truly representative list of background reading. 

At the library at our university, we access the library resources through a single initial screen. Some of the library resources are available to anyone with access to the web. Other resources will be restricted to faculty and students from the sponsoring institution. For example, computerized databases are often provided through a subscription, and the library can only afford to provide those resources to students and faculty at the school. 

By clicking on the phrase "Databases by title" on the right of the screen, we are directed to a second screen, which lists the available resources. If instead we click on "Resources by subject," we get this screen listing to general subject areas. Clicking on "Psychology" provides us with a list of databases relevant for psychology. Note that many databases are relevant to more than one subject area and are thus listed in EVERY relevant subject area. 

Finally, we click on PsycINFO to get the primary database for psychology. Of course, the way this library is organized is not likely to be the same way your own library is organized. But librarians are masters of organization, and it is likely that you will be able to find the relevant information just as easily at your university.

Clicking on PsycINFO brings up a screen asking for our name and password. Again, most universities require password information so that only students and faculty have access to these costly resources. After we enter them, this search screen appears. Check with your librarian to find out how to get access to similar databases at your library. 

Be aware that this is the search screen for a particular subscription service--in this case, OVID. Your university may purchase resources through another subscription service. Therefore, the screen may look different. In fact, depending on when you are viewing this tutorial, even the OVID search screen may look different. Most search services update the looks of their search screens every few years. However, the search strategies are likely to be very similar. In fact, if you are skilled at computer library searches with one database, you can often search other databases with little more than a brief orientation. 

The Help button in the upper right hand corner can provide a brief orientation to a database or answer any specific questions you might have about how the database is organized. [NOTE: The screens that will be presented in this demonstration are just images of the screens that you would see if you did the search. They are NOT functional. Therefore, clicking on icons will not produce the same result you would obtain working with the real product.]

We will briefly walk you through an initial search of this topic, showing you about 30 screens from the process to help you follow along. These are not the only relevant screens, and our approach is certainly not the only way to attack this problem, but we think it will give you a good idea of how to find information relevant to any topic you might be interested in. After we have slowly walked you through the logic of a search, we will provide you with a series of animations that illustrate every step of the process.

• We begin by entering the phrase "explicit memory" in the box under the command Enter keyword or phrase and then click on the Search button. We are searching first for keywords, and that is the default option. If we wanted to search by author, for example, we would have clicked on the author icon near the top of the screen and then entered the author's name in the box provided. Alternatively, we could enter the author's name in the box and then click the author button on top. This tells the program to search for an author's name rather than for a keyword. We will see an example of this type of search shortly.

• Our request for information on "explicit memory" gives us this next screen. The program is asking us how specific or general we want the search. We will just start with the topic area of "explicit memory" by clicking on that subject as shown, although we could have opted for a broader search if we wanted to. We may end up going back and broadening our search later, which is easily done in a library search. Note that the program is giving you hints at the bottom of the screen. You can play with these hints to see how they affect your search when you are doing your own library research. Clicking on the continue button gives us this next screen.

• The box at the top of the screen notes that 494 articles were found, the first 10 of which are listed on the page. We can look at more of the 494 articles by clicking the Next Results button near both the top and bottom of the page, but we will want to narrow the search before we do that. 

• Entering the keyword "anxiety disorder" and clicking the Perform Search button gives us this screen. We select just the keyword "anxiety disorders" and click continue. The result of this search suggests that there is a lot of research published on this topic. There were 34,000+ articles with the keyword of anxiety disorders, but probably few of them looked at explicit memory.

• Our next step is to look for articles that include both explicit memory and anxiety disorders, which we can do with the combine option. Clicking on Combine Searches near the bottom of the Search Results box will give you this screen. 

• To find articles that have both keywords we (1) select both #1 and #2 by clicking in the box in the Select column and (2) use the AND operator. The various Boolean operators are in a drop-down box just to the right of the words "Combine a search using." Clicking on the arrow of the drop-down box shows you the operators available. Clicking on the Continue button produces this next page.

• Only 22 articles had both of these terms in their keywords, and the first 10 of those articles are listed at the bottom of the page. If we wanted to, we could look at the abstract of any of these articles by clicking on the abstract hyperlink for the article.

• Twenty-two articles on the topic is a start, but certainly there must be more relevant articles than that. The Becker et al. paper referenced dozens of articles. Therefore, we want to try a different strategy to find more articles. We will now search for articles by people who have published extensively in this area. Remember, that was something that we noticed when we looked at the Becker et al. reference list.  

• To search by author, we click on the author icon at the top of the screen, which will give us this screen. The format of the screen is very similar to the previous screen except that now we are entering an author's name. Note that the format for entering the name is listed just above the box for the name (last name, a space, first initial if known). We are going to start with the name "Mathews A." Dr. Mathews was coauthor of several relevant papers in the Becker et al. reference list. We enter the name in the box provided and click the Search button, which gives us this screen.

• Now we have to do some guessing. Reference lists only give the initials of authors. There are several possibilities for the name "Mathews, A" listed. If we happen to know that his name is Andrew Mathews, then we can eliminate Alice, Andre, Angela, Anne, and Audrey. Actually, in this case we can probably rule out all of them, because they have all published very few articles, and we know that our A. Mathews published several articles. Chances are A. Mathews, A. M. Mathews, Andrew Mathews, and Andrew M. Mathews are all the same person, so we would check each of those names. If we are wrong at this point and include the publications of another author, it is not a big deal. Those incorrectly identified articles will likely drop out later when we narrow the search more. At this stage, you want to be over-inclusive, checking any names that look like they might be the person you are looking for. Clicking on the Search button will give you this screen.

• We have 131 articles by Dr. Mathews. You might note that we implicitly used the Boolean operator OR in compiling this list of 131 papers when we selected all four variations of his name in the previous screen. 

• A quick glance at the titles of the first 10 of the 131 articles verifies that we have papers on various cognitive processes in anxiety disorders, including at least two papers on memory in anxiety disorders. 

• We could look at all 131 of the articles, but it is easier to narrow the search again at this point. We click on Combine Searches to get this screen. We will combine anxiety disorder (2) and Mathews (4) by clicking each and clicking the Continue button, which gives us this screen. 

• There are 41 articles by Mathews with a keyword anxiety disorders. We can view the first ten on this page and the next 10 by clicking on the Next Results, and the next ten by clicking on the Next Results again, and so on. Several of these articles look promising, and we would likely get copies of them at the library and start reading. First, however, let's find some other possible background papers to read.

• We click on the Main Search Page icon at the top of the page to go back to the search page. Another author whose name came up repeatedly was McNally. We enter "McNally R" and click on the author icon, which gives us this page. There are several people with the name McNally, but there are only six R. McNallys and only one with the publication record that we would expect. We would select Richard, Richard J., and R. J. McNally and click on the combine search button to get this screen. [Note that if you perform this same search exactly as we outline it, you are likely to get different numbers of articles, because the databases are constantly being updated with new published articles.]

• While we are at it, we will search for another prominent author from the Becker et al. reference list. We click on the author icon, enter "Borkovec T," and click on the Search button to get this page. There are only three Borkovec entries and they all look promising. We select all three and click on the perform search button to get this screen.

• Let's do one more author (C. MacLeod). Entering that name and clicking on the author button gives us this screen. There is only one MacLeod with a publication list extensive enough to be the person we are looking for. We select Colin and Colin M. MacLeod, and just to be on the safe side, we also select C. MacLeod and C. M. MacLeod. Clicking on the Search button gives us this screen.

• You may notice that part of our search history (the listing of what we have done so far) is now hidden on this screen. That is done to save space on the screen, but clicking on the expand tab on the side of the search history boxes will give us the entire history at any time.

• As you can see, Andrew Mathews, Rich McNally, Tom Borkovec, and Colin MacLeod each published a substantial number of papers. Each appears to publish many of their papers on anxiety disorders and several on memory functioning in anxiety disorders. We could, if we wanted to, obtain the 700+ papers from these four prominent authors and read them all. We could screen the abstracts of each paper and read only those that seem relevant. But we are going to try to narrow the search before we do either of these. 

• Our original search for "explicit memory" gave us a number of articles. We have noted in our search so far that there are many papers on memory functions that do not appear to involve explicit memory, but may well be of interest to us. Therefore, we want to do an expanded search of the memory keyword. 

• We enter "memory" in the keyword box and click on the Search button, giving us this screen. We want to gather into a set as many articles about memory as we think might be useful. We decide to select explicit memory, memory, semantic memory, spatial memory, verbal memory, and visual memory. Some of these are unlikely to be relevant to the topic, but if they are irrelevant, they will drop out when we narrow the search again. Clicking on the Continue button gives us this screen.

• Well, we certainly expanded our set of memory articles (52,000+). We definitely do not want to read all of those, but that was never our intention. We want to use this set to compare it to other sets of articles we have identified. We click on the Expand tab to get this screen, which shows our entire search history. 

• When we had narrowed the search early in our efforts by combining explicit memory (#1) with anxiety disorder (#2), we only got 22 articles. We suspect now that the term "explicit memory" was too narrow for our purposes, so we identified a much larger set of memory articles (#9). We want to combine this larger set of memory articles with the set of anxiety disorders articles, so we click on Combine Searches to get this screen. The sets we want to combine are #2 and #9, so we select them and click Continue to get this screen.

• Now we have identified a set of 603 articles on memory in anxiety disorders. We can try to narrow it down further, but we might be better off scanning the abstracts of the articles or even the titles to get a set that will serve our needs. 

• We have also obtained manageable sized sets of articles from four prominent researchers in the area. We could narrow those sets by combining them with the memory set by using the Boolean operator AND. Scanning those articles would likely give us a useful set of core articles. We can quickly scan these articles and their abstracts online. 

• Let's say that we want to start to screen the 603 articles on memory and anxiety disorders. We would click on the Display link at the end of that line, which will give us this first screen. Clicking Next Results gives us this next screen. Continuing to click on Next Results will give us the third screen, the fourth screen, the fifth screen, the sixth screen, and so on. If an article catches our eye, we can click on the Abstract link for that article, which will produce a screen like this. At any point, we can print a screen for our later reference.

• We can also mark the articles that seem more relevant by clicking in the box at the beginning of the article's reference. After we have screened the articles, we might end up identifying 35 that we may want to explore more. By selecting them, we can use the resources at the bottom of the screen (labeled Citation Manager) to print the list of relevant articles, save it to a file, or even email it to ourselves or someone else that we are working with. We can easily download the abstract as well as the reference for later review.

• To this point, we have been thinking in terms of a broad concept of memory distortions in anxiety disorders. As we scan articles, we will see that there are many different types of anxiety disorders, so we may find ourselves wanting to focus on a specific anxiety disorder (e.g., panic disorder or PTSD). 

• As we start to narrow our focus, we may want to go back to the library and try another search with the more specific focus that we need. Also, as we read the articles that seem relevant to our more narrow topic, we will identify different terms that seem to represent the same general phenomenon. We will want to use all relevant terms in our search to broaden the topic appropriately. For example, we may find that terms like "memory bias" and "memory distortions" and "distorted memories" are all used to describe the concept in which we are interested. When we go back to the library for a second search, we will want to create initially a set that uses any of these terms. To create a set that uses any of several terms, we will search for each term separately and then combine them using the Boolean operator OR.

• The work so far (probably about 30 minutes at the library or on your computer at home) is just the first step. It is best to do some background reading now to familiarize yourself with the concepts and identify issues that seem central to you. In that way, when you return to the library, you will be able to do a much more specific and detailed search.

One thing that we did not mention in this search is that it is common to find that certain areas of research assume that you have knowledge about particular theories, procedures, or strategies, which you may not be familiar with. Sometimes these elements will be referenced in the papers you read, giving you a source to pick up this critical information. Sometimes they will not be referenced. It is not uncommon to have to do a specific literature search for these elements of existing research and theory in order for you to build the background understanding required.

Animations

Now that we have walked you through the logic of a literature search, we will perform many of the same steps just described in flash animations, so that you can see every step and how each step is performed. We have broken this series of animations into discrete steps. Just click on each step in turn to see the animations. [NOTE: Although these animations are very helpful in seeing how each of these steps are performed, they take a long time to download unless you have a high speed Internet connection. We do not recommend these animations if you are relying on a dial-up connection.] 

To see each of the animations, just click on the titles in the table below. Use the back arrow key on your browser program to return to this page.

 

---

Next Section

Return to the Library Research Main Menu

 

APA Style Tutorial | Authors | Chapter Resources | Copyright | Errata | Flashcards | Glossary
Help | History | Home | Index | Instructor's Resources | Internet Links | Library Research
Manual Computation of Statistics | New Material | Practice Quizzes | Random Number Program
Research Design Checklist | Research Examples | Site Map | SPSS for Windows | Statistical Concepts
Statistical Flowcharts | Statistical Tables | Study Guide/Lab manual | Tutorials | Web Browser Tutorial

Copyright © 2020 Graziano & Raulin. All Rights Reserved.