Skip to Main Content

How to search like a librarian (Dahlen, AAA 2025)

This guide provides supplemental resources for the workshop "How to search like a librarian: Systematic search skills for professionals and students" presented by Sarah Dahlen at the 2025 meetings of the AAA.

Keywords and controlled vocabulary

You're probably familiar with using keywords to search for information. This is called "natural language searching," because we use the words that come to our minds when thinking about our main concepts. We want to include keywords in our searches. 

We also want to use controlled vocabulary, which are standardized terms that provide a consistent way to describe concepts. Each article in a database that uses controlled vocabulary has been assigned these terms (sometimes known as "subjects" or "subject headings") that identify the concepts that are central to the topic of the article. 

If I search using the controlled vocabulary term "academic persistence" in the subjects field, my results will only include those sources that have been tagged as having "academic persistence" as central to their topic, not articles that merely mention that term anywhere in the full text. 

Boolean searching

We use Boolean operators (AND, OR, NOT) to construct our search strings. If you're not familiar with Boolean operators, this video gives a nice overview. 

Choosing databases

Not all databases use controlled vocabulary. For those that don't, you would search with keyword strings only. 

Here is a partial list of databases and database platforms that include controlled vocabulary: 

  • EBSCO (platform)
  • ProQuest (platform)
  • PubMed
  • Embase
     

These databases do not use controlled vocabulary: 

  • JSTOR
  • AnthroSource
  • Google Scholar
  • Sage
  • Science Direct (Elsevier)
  • Wiley
  • Web of Science
  • Scopus
     

If you are using a database that isn't on either list, look at whether it allows you to search a "subjects" field or whether subjects are listed for the articles in the results list. These are signs that the database uses a controlled vocabulary. 

Populating your concept table

A concept table provides a structure for generating keywords and controlled vocabulary that relate to each of your main concepts. These terms will be used to create your search strings. 

Concept table template

Example of a completed concept table

This video will walk you through how to populate your concept table, including generating keywords, controlled vocabulary, and search strings. 

Once the first tab ("search strings") of your concept table has been finalized, the next step is to put all of your keywords and controlled vocabulary that are phrases (more than one word) in quotation marks. We do this so that the databases search for the phrase "student parents," for example, rather than searching for "students" and "parents" as separate words. 

You will also delete any punctuation and put a capitalized "OR" between each term. 

Example:  "postsecondary education" OR colleges OR universities OR "higher education" OR "state universities"

Notice that the search terms that are single words (colleges, universities) do not need to be in quotation marks. 

Wrapping text in your table

If you can't see all of the your terms in the cells of your concept table, you may need to format the cells so that the text wraps.

Select the cells and find the text wrapping icon in the menu at the top of the table (you may have to click the three dots for it to appear): 

text wrapping in menu

How the concept table organizes your search

To be able to check that your concept table is put together properly, you need to understand how it will translate into your database search in the next step. 

What we want is for the database to return articles that include all of our main concepts. 

In each column, the different terms representing that main concept are connected by "OR," meaning that any one of those terms represents that concept. When we search in the databases, each column will be connected by "AND," meaning that we're searching for sources that include at least one term from each column. In other words, the we are looking for sources that include all of our main concepts, we've just used a lot of different terms to represent what our main concepts are. 

If there's a term in one of your columns that doesn't really represent that concept, consider removing it, as it will produce some irrelevant results.