How to get started with systematic literature searching

As you transition from basic to more advanced search techniques, understanding Boolean logic and advanced search syntax becomes essential for refining search queries and obtaining more precise results. Boolean searching is the basic unit of almost every search query and involves the use of logical operators such as AND, OR, and NOT to combine search terms and create complex queries and advanced search syntax is additional operators used in searches, such as proximity operators, wildcards, truncation

When you enter a search query into a database, the system processes the words or phrases you provide to find matching content within its indexed database. The system looks for documents or records that contain the exact words or related variations of the terms you've entered. It uses indexing and search algorithms to match your query with the indexed content and retrieve relevant information based on your search criteria. It will also retrieve results based on your use of a number of different search operators and syntax.

In the initial stages of planning the search for your work, it's crucial to define its scope. Consider the following aspects:

1. Determine the length of your literature review.
2. Specify the number of sources or citations needed.
3. Establish publication date requirements, if any (e.g., including sources published between 1990 and 2010).
4. Identify the types of resources to consult, such as trade journals, peer-reviewed journals, books, or websites.
5. Decide whether your literature review will focus on theoretical issues, topical subjects, or qualitative/quantitative research.
6. Develop clear inclusion and exclusion criteria to guide your selection of relevant literature for the review.

At the very start of your work, it's advisable to start by exploring an overview of your topic. Background information offers:

1. Keywords and search terms to expand your understanding and locate further information on your subject.
2. Insights into the fundamental theories and predominant themes related to your topic.

UWS Library Services offer a multitude of resources, including electronic and print materials, to help you explore various subjects and possibilities. Encyclopedias, for instance, offer concise summaries of topics and serve as gateways to additional scholarly works on the subject matter.

When searching academic databases and the Library's catalogue, using general keywords can help you cast a wide net and uncover a variety of relevant resources across different disciplines and subjects.

Boolean logic serves as the fundamental framework for constructing search queries. It enables users to combine keywords and phrases using logical operators such as AND, OR, and NOT to refine search results. By employing Boolean logic, you can specify the relationships between terms and control the inclusiveness or exclusiveness of their search queries, thereby enhancing the precision and relevance of the retrieved information.

Search syntax refers to the rules and conventions used to construct search queries within a database or search engine. It involves the use of specific symbols, operators, and formatting techniques to refine and specify search criteria. Syntax may vary from one database to another. Additionally, databases may have their own unique set of search syntax and features, so it's essential to consult the database's help documentation for guidance on using syntax effectively within that specific platform.

When you combine search syntax with keywords, you create structured search queries that help databases and search engines understand your information needs more precisely. No matter how complicated or simple your search strategy, it is still based on these basic building blocks. Here's what happens when you combine syntax with keywords:

1. Boolean Operators: Using Boolean operators like 'AND,' 'OR,' and 'NOT' allows you to create both simple or complex relationships between keywords. For example:

   • 'AND' narrows down your search results by requiring both keywords to be present.
   • 'OR' broadens your search results by finding records containing either keyword.
   • 'NOT' excludes certain keywords from your search results.

2. Wildcards (also known as Truncation): Wildcards like asterisks (*) or question marks (?) enable you to search for variations or partial matches of keywords. For instance (operation of Wildcards will vary from one database provider to another, so check before you begin your search):

   • 'lyric*' would retrieve results containing 'lyric,' 'lyrics,' 'lyrical,' 'lyricism,' etc.
   • 'wom?n' would retrieve results containing 'woman' or 'women.'

3. Phrase Searching: Quotation marks (" ") indicate that words should be searched together as a phrase. This helps retrieve results where the words appear together in the specified order. By enclosing a phrase within quotation marks in a search query, you instruct the search engine to look for the precise phrase as it appears within the text:

   • "Born in the USA" the database's search engine will look for documents or records containing this exact phrase within the selected fields (title, abstract, etc).

4. Proximity Operators: Proximity operators like 'NEAR', 'adj', or 'WITHIN' allow you to specify the distance between keywords in your search. For example, 'child NEAR/3 education' would retrieve results where 'child' and 'education' appear within three words of each other. They imply a more semantic relationship between the concepts being searched for. For example, using a proximity operator can indicate that the terms should appear close to each other in the document, suggesting a stronger association between them. This helps narrow down search results to documents where the terms are contextually related or appear in close proximity to each other. It's important to note that not all databases support proximity operators.

Controlled vocabularies are standardised sets of terms that are used to describe a particular topic or subject in databases. They are often subject specific, consist of preferred terms and associated synonyms or non-preferred terms, and are designed to capture the unique terminology and concepts within a particular field or domain. In MEDLINE EBSCO, the field is MH Exact Subject Heading. When a record is added to a database, it may be assigned one or more terms to describe its subject matter, this makes it easier to find when searching for records about a specific topic. 

Every record is assigned one or more of these standardised terms which makes it easier to search for records on a particular topic because all records that are about the same topic will use the same terms.  Some databases rely more on keyword searching, where users can input free-text terms to search for relevant content. It's important to be aware of the indexing and search capabilities of the specific databases you are using and adapt your search strategies accordingly.

When crafting a complex search that involves various sets of terms related to specific aspects of your research question, consider saving each component as a separate search (referred to as 'blocks'). These individual searches can then be run independently or combined with other searches as needed.

Boolean logic serves as the fundamental building blocks of search queries. In this case, we are addressing the following research question: 'How does physical fitness and exercise impact stroke recovery, as evidenced by randomised controlled trials (RCTs)?'. We will be looking for keywords and terms for 'stroke', another for physical fitness interventions, and a subsequent collection to identify randomised controlled trials. We want to retrieve documents which contain all these criteria.

Think of it like a Venn diagram; each set of terms functions as a collection of keywords combined using the OR function, expanding the search to include any of the terms within a set. The overlaps represent the AND operator, requiring all terms to be present simultaneously, thus leading to more precise and targeted information retrieval.

The above search contains three collections of terms. In a triple intersection, each element or item must meet the criteria specified in all three sections of the query. Therefore, it contains at least one word from each of the sections, satisfying the conditions set forth in the query's criteria.

Using a combination of both free-text/keywords and controlled vocabulary terms where available. This will maximise the retrieval of potentially relevant records. Let's use the National Library of Medicine's Medical Subject Headings (MesH) in the MEDLINE EBSCO database in the following example.

When the the Query containing both controlled vocabulary and keywords is entered into a database, it might look as follows:

The final line integrates the terms from each preceding step into a single line which represents the your query. Line-by-line searching involves breaking down the search process into smaller, more focused steps. This method helps researchers adjust their search terms based on initial findings, ensuring a thorough and targeted exploration of relevant literature.