also: article databases, research databases
Literature databases are typically composed of records that describe articles that were published in scholarly journals, trade journals, newspapers and magazines, as well as resources like conference proceedings and papers, reports, and government or legal publications.
Databases may be multidisciplinary/multipurpose, or have a specific focus that determines which sources get indexed into records for the database, based on subject focus, date coverage, and material type. For example, see the table below for scope information related to some major health sciences databases.
Some databases may be entirely bibliographic, meaning they contain only records that describe literature. Other databases are considered full-text, which means that in addition to bibliographic descriptions, they contain the entire texts of the associated document. Other databases may contain full text content for a subset of their records.
Databases are typically made available electronically on a commercial basis by public or private database producers (sometimes known as vendors), who provide a platform or interface to access and search the database.
|
Database |
MEDLINE |
CINAHL |
|---|---|---|
|
Platform/Vendors |
>Ovid (Wolters Kluwer) >PubMed (NLM) |
>EBSCO |
|
Subject Coverage |
biomedicine and health, broadly, including:
|
Nursing & allied health, broadly, including:
|
|
Publication Type Coverage |
|
|
|
Full Text Access |
No; records may contain link outs to full text on PubMed Central or publishers’ pages |
Partial; some journals are held in full text in CINAHL |
International Encyclopedia of Information and Library Science, edited by John Feather, and Paul Sturges, Taylor & Francis Group, 2003. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.action?docID=178969.
Medline overview. (n.d.). [Product, Program, and Project Descriptions]. Retrieved September 27, 2022, from https://www.nlm.nih.gov/medline/medline_overview.html
Database platforms may be provided by:
also: database record, descriptive record, reference, citation
A bibliographic record contains the data elements necessary to help users identify and retrieve that resource, as well as additional supporting information, presented in a formalized bibliographic format. Bibliographic records are composed of metadata fields, fields that contain information about the resource that aids in discovery and organization (e.g., title, abstract, author, publication details, resource type, and subject headings).
Bibliographic record. (2022). In Wikipedia. https://en.wikipedia.org/w/index.php?title=Bibliographic_record&oldid=1100147395
In bibliographic records, important descriptive metadata includes title, author, source type, and subject descriptors.
Merriam-Webster. (n.d.). Metadata. In Merriam-Webster.com dictionary. Retrieved September 27, 2022, from https://www.merriam-webster.com/dictionary/metadata
also: free text terms
Keywords are flexible and variable, often drawn from a researcher’s natural language descriptions of the core concepts of interest. Keywords can be used as search terms in a database interface and used to ask the database to retrieve records that contain those terms. Ideally, keywords should be significant, meaningful terms that describe or relate to a concept of interest.
Because databases generally search for the exact terms that are entered, certain types of keywords should be avoided, including relationship words (e.g., cause, contrast, correlation), and judgment words (e.g., advantage, disadvantage, benefit). For more guidance, see Crafting a Search.
Keywords can be searched for in specific fields of a database record (e.g., title or abstract) using field tags or in an unqualified search, where the search will likely retrieve records that contain the keyword in any field of the database record.
also: index terms, controlled subject vocabulary
When a record gets created for a database, subject headings are included in order to better describe the information resource. Those subject headings are usually drawn from standardized lists (i.e., thesauri) that establish definitions and preferred usages.
Subject headings are typically based on a controlled vocabulary, meaning that for any given concept, only one term is used (e.g., articles about “cars”, or “motor vehicles” or “sedans” would all be tagged with the same subject heading: “automobiles”). Subject headings may differ from the natural language text words or keywords that are used to describe a concept.
Subject headings are also typically hierarchical, or exist within a tree structure; narrower, more specific subject headings sit under broader, more general subject terms. For example, in the MeSH hierarchy, the term “Mass Vaccination” is a more specific term under “Vaccination”, which itself is a more specific term under “Immunization”
Tree View for MeSH Term “Vaccination”:
Different databases employ different thesauri of subject headings in the indexing process. For example,
Field tags can be used to conduct “subject searching”, where only records with specific headings in the subject field are retrieved.
For more information, see Crafting a Search-Identifying Subject Headings
also: “exploding a heading”
By default, when searching PubMed using MeSH terms, the terms will be searched with ‘autoexplode’, and the search will automatically include records tagged with the more specific headings underneath that term. For instance, a search for Vaccination[MeSH] will automatically also retrieve records tagged with the MeSH term “Mass Vaccination”, because that term is under “Vaccination” in the subject tree.
In other databases, subject explosion does not happen automatically; the user must specify explosion, if needed, using specific search syntax. For instance, CINAHL uses the + sign to denote explosion (e.g., MH "Vaccines+"), whereas databases on the Ovid platform (EMBASE, MEDLINE, PsycINFO) add exp to the subject term (e.g. exp vaccine/).
For more information see: Strategies for Broadening a Search - 2. Explode Your Subject Headings
also: qualifiers
For example, the MeSH thesaurus contains qualifiers that can be used with MeSH terms for conditions, drugs, organs, and other subjects (e.g., “Breast Neoplasms/genetics” for the genetics of breast cancer, or “Uterus/surgery” for surgery performed on the uterus). CINAHL subject headings can also be made more specific with available subheadings.
also: Boolean connectors
Connecting search terms with AND tells the database to retrieve records only if each of those terms appear in the record. The operator AND is typically used to connect conceptually distinct terms. The more concepts that are connected with AND, the narrower the search results become.
Connecting search terms with OR tells the database to return records if any of the given terms appear in the record. Connecting synonyms and alternate terms with OR expands the search results ("OR retrieves MORE").
Connecting search terms with NOT tells the database to exclude records from the search result if the record contains the specified term. Adding an excluded term with NOT will create a narrower, more specific search.
A word of caution: the NOT operator should be used sparingly and carefully; because the database will simply exclude records in which the specified word appears, it is possible to accidentally exclude relevant records that happen to contain that term.
For more information about Boolean operators and examples, see: Crafting a Search-Employing Boolean Operators
also: adjacency search
For example, in the EBSCO platform the following query will retrieve records if the word “gender” is found within 3 words of the word “bias”.
gender N3 bias
Proximity searching is useful in cases where there are multi-word variations of a keyword that might be relevant to include. For instance, the example above will locate records if they contain the phrase “bias based on gender”, “gender related bias”, “bias due to gender”, etc.
Different databases platforms use different operators for adjacency searching, as well as different rules for counting the number of intervening words.
Many databases use quotation marks as the punctuation for phrase searching. For instance, searching “metabolic syndrome” will only retrieve records if that exact phrase is used in the record (i.e., the results won’t contain records if that specific phrase is not used in the record).
Different database platforms may allow for different phrase searching functions. See these resources for the specifics related to common platforms:
Also: wildcard
Many databases use an asterisk * as the wildcard punctuation. For example, the query allerg* would retrieve records containing the word allergy, or allergies, or allergic, or allergist, or allergen, etc
Different database platforms allow for different truncation and wildcard punctuation. See these resources for the specifics related to common platforms:
also: field code, search tag
For example, the field code TI can be used in CINAHL to specify a phrase should be found in the title field: TI “environmental racism”
For more information, see Crafting a Search - Field Tags
Different databases have different default search fields, but title, abstract and subject heading fields are commonly included. For specific information about what fields are included in unqualified searches by database, see:
also: search string(s)
Search queries can range from relatively simple to more complex. Simple search queries often rely on keyword searches, and allow the database to run the search based on default settings. Complex search queries might include more operators, specific punctuation, database-specific field tags, and may make use of a ‘building block’ style of searching.
misinformation AND nutrition
("Disinformation"[Mesh] OR "Communication"[Mesh:NoExp] OR "Information Literacy"[Mesh] OR misinformation[TIAB] OR "fake news"[TIAB] OR infodemic[TIAB]) AND ("Nutritional Sciences"[Mesh] OR "Nutrition Therapy"[Mesh] OR "Diet Therapy"[Mesh] OR "Food"[Mesh] OR "Dietary Supplements"[Mesh] OR nutrition[TIAB] OR food[TIAB] OR diet[TIAB])
A building block style search can often be created using the tools within a database platforms’ advanced search or search history.
|
Search Number |
Search Query |
Results |
|---|---|---|
|
#4 |
#1 AND #2 AND #3 |
31 |
|
#3 |
conspirac* OR conspiratorial OR distrust* authorit* |
1,680 |
|
#2 |
misinformation OR disinformation OR “fake news” OR infodemic |
677,444 |
|
#1 |
nutrition* OR food OR diet* OR supplement* |
2,522,212 |
(nutrition* OR food OR diet* OR supplement*) AND (misinformation OR disinformation OR “fake news” OR infodemic) AND (conspirac* OR conspiratorial OR distrust* authorit*)
also: database limits or limiters
For example, a publication date limiter will tell the database to only retrieve records if the publication date listed falls within a specified range.
Different databases offer different filtering capabilities, depending on the metadata that is present in the database’s records. For instance, because health-focused databases (like MEDLINE, CINAHL and PsycINFO) include metadata that describe the characteristics of the participants in a study, it is possible to limit a search by elements like participants; age, sex, or species.
For instance, the search query burns AND dressing will retrieve records about dressing burn wounds. But the results set might also contain records for articles about dressing as an activity of daily living, by author Robert Burns.
also: broad search
A sensitive search can be built by integrating a flexible set of keywords and subject headings, and minimizing the database filters that are applied.
In capturing most of the relevant literature, there is a low risk that a sensitive search omits relevant literature
By design, a sensitive search will also capture a large number of ‘false hits’, so more time must be spend manually screening through search results to identify truly relevant results
Database searches that are performed for evidence synthesis projects (e.g., systematic reviews, scoping reviews, integrative reviews), should err on the side of sensitivity.
also: narrow search
Specific searches are often built by using more specific search terms, phrase searching, and database limits.
Specific searches will capture some portion of the existing relevant literature, and by design, the results will contain fewer irrelevant records (“false hits”), leading to less time spent manually screening through records.
A search that prioritizes specificity holds a greater potential for missing relevant literature.
also: reproducible search
Such details include:
To save a reproducible search string, it may be useful to create a database account and save the exact search that was run, in the native database environment. For more information, see: Saving Searches.
also: citation pearl growing
Strategies for snowballing based on a relevant source include:
also: hand search, hand-search
Handsearching is a process employed in systematic literature searches that is meant to complement structured searching of electronic databases in order to identify any additional eligible sources that were not located by the database search. Handsearching may also involve ‘snowballing’ searches, where the reference lists and citations of included studies are checked for additional sources.
For more information, see Rutgers’ LibGuide - Hand Searching
also: citation searching, citation chasing
Several online tools and citation indexes allow the user to track the citation of a given publication (forward in time). In other words, these tools will generate a list of references that have cited a given text since it was published.
For more information, see: Guide to Cited Reference Searching
also: PRISMA flowchart, search flow diagram
PRISMA diagrams map out the number of records identified, included and excluded, and the reasons for exclusions.
PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) is a set of reporting standards designed to ensure transparent reporting of systematic reviews. PRISMA diagrams are most appropriately used in conjunction with the PRISMA Checklist, a 27-item checklist addressing the introduction, methods, results and discussion sections of a systematic review report.
Creation of a PRISMA diagram requires careful record keeping during the screening phase of an evidence synthesis project, which requires the use of citation management software (e.g., Zotero or EndNote), and/or systematic review screening software (e.g., Covidence).
For more information, see: Tutorials & Tools for Literature Reviews
