|Year : 2020 | Volume
| Issue : 2 | Page : 193-197
Building blocks: The art and science of searching the literature
Professor, College of Nursing, CMC, Vellore, Tamil Nadu, India
|Date of Submission||30-Dec-2020|
|Date of Decision||31-Dec-2020|
|Date of Acceptance||03-Jan-2021|
|Date of Web Publication||19-Feb-2021|
Dr. Sheela Durai
College of Nursing, CMC, Vellore, Tamil Nadu
Source of Support: None, Conflict of Interest: None
The universal phenomenon of 'search' plays a significant role in the process of review of literature. When review is carried out as part of research, the search for literature itself is to be done methodologically and systematically, turning the process of search into a science. Formulation of the search strategy is the most significant step which will direct the course as well as the content in the review of literature and needs to be given a lot attention. This article introduces scientific literature to the nurse researcher and elaborates on the building blocks of the stages of search, such as identification of search approach, databases, terms/phrases, areas/fields, search operators and search limits. When a nurse researcher gets familiar with the search styles and tactics of scientific literature, he/she realises that literature search is an art and science of building search blocks effectively with much strategising.
Keywords: Database, literature, review, search strategy, search
|How to cite this article:|
Durai S. Building blocks: The art and science of searching the literature. Indian J Cont Nsg Edn 2020;21:193-7
| Introduction|| |
The search for the best has been a universal phenomenon for ages. People spend a lot of time surfing the net, trying to buy the best camera phone or identify the best diet for weight loss. It is no different in the healthcare system. It can rather be stated that this search for the best answer is more prevalent in the medical field. Every person associated with health and illness is looking for the best option or solution, be it a health team member, a patient or a family member. Nurses, as advocates of patients and their families, always look out for the best interest of the patient and inevitably end up searching for answers. This search turns out to be a scientific exercise when it is pursued as part of a research. Literature review, an integral part of every research, begins with a search, though the way the search is carried out depends on the type of review.
| Scientific Literature|| |
Literature is the word used to describe all written material relevant to the topic selected. This term is derived from the Latin word 'litaritura/litteratura' meaning 'writing formed with letters'. It is frequently used in academic circles to indicate information that has been published on a topic. Scientific literature consists of scholarly publications that report original theoretical and empirical work in the natural and social academic fields. Scientific publishing consists of the process of placing results or findings of one's research into literature.
| Classification of Scientific Literature|| |
Scholarly publishing entails three different levels of literature depending on the origin of the content.
It is the literature that contains findings of original research, written by researchers who performed the study. Earliest material that was created from the time period involved and not filtered through interpretation or evaluation are also referred to as primary sources., Examples include journal articles, conference proceedings, patents, dissertations, diaries and newspaper articles of that time.
Literature that synthesises information from primary sources to draw new conclusions, derive new data or provide criticism on topics discussed is called a secondary source. They help in understanding what is known and not known about a topic, and what additional information is required for research., Textbooks, encyclopaedias, reviews, commentaries, criticisms and monographs are secondary sources.
When literature is primarily used as a source to locate primary and secondary sources, and does not provide any new information, then it is called as tertiary source. A tertiary source may consolidate or organise primary and secondary sources together into one source in order to facilitate easy, quick access to data. Indexes, bibliographies, directories, handbooks, manuals, etc., are tertiary sources.
A scholar delving into research needs to know how to search, identify, gather, read, synthesise and write the content that is relevant to the research at hand, giving due importance to primary, secondary and tertiary sources.
| Review of Scientific Literature|| |
Review of literature is the art of meaningfully reading and organising available literature. It is also a science when the entire process of search, analysis, synthesis and presentation of literature involves a methodological course. It is a piece of discursive prose meaningfully constructed and presented and not just the list of summaries. The purpose is to convey to the reader the existing knowledge and ideas that are already established in the topic being studied. A thorough review of literature is essential to be aware of all known concepts related to the phenomenon and also identify the gaps in knowledge, though the research methodology determines the type of review.
In quantitative research, review of literature takes on an important role as it directs the path the researcher needs to pursue. It is through the review of literature that the researcher justifies the purpose of his study, establishes a conceptual framework, finds the gaps in literature, identifies the aspects to be studied, adopts methodology from and also obtains evidence that supports or contradicts the findings of the study. In qualitative research, each methodology involves a different kind of review. Review of literature is done after data collection in phenomenology. There is limited review in grounded theory. Early review is followed in ethnography, and historical research is all about only review of literature.
Review of literature itself turns into a research when it is used systematically following scientific rigour. Scoping review, integrative review, narrative review, rapid review, umbrella review and systematic reviews including Cochrane reviews, Campbell Collaboration reviews, meta-analyses and meta-synthesis are the different types of reviews. Whatever be the type of review, every review begins with a plan for searching the literature.
| Search – The Cornerstone of a Good Review|| |
Searching the literature can take on different paths depending on the type of review. Even if the purpose of the review is to provide a framework for research, every search is to be done in a methodological manner by the scholarly nurse to provide a firm knowledge base.
Stages of literature search
The process of search itself is graded into five stages, starting from formulation of search strategy to classifying findings of search.
- Formulation of search strategy
- Implementation of search strategy
- Retrieval of sources
- Screening of sources
- Classification/categorising/grouping/grading of findings/preparation of themes.
The nursing researcher may find herself involved in different stages of the search simultaneously, but the first stage is to be definitely chalked out before embarking on the other stages.
| Formulation of Search Strategy|| |
The search stage is most crucial in review of literature, but most textbooks do not provide in-depth information on how to go about it, partly because every database has its own unique techniques and terms related to search. A systematic, methodologically organised search of the literature with effective use of available resources ensures quality yield of relevant literature and thus it is imperative that nurses are aware of the principles and steps in planning search strategy. The first aspect to be determined in the search strategy is the plan for location of relevant literature. The decisions taken during search strategy planning can determine the answers, conclusions and quality of the results. The nurse researcher is to make decisions regarding three important ways of locating literature before starting the search [Figure 1]. The first building block in the construction of a good search strategy is the block of locating literature.
| Block 1 – Location of Relevant Literature|| |
The foundation of a good search strategy is the determination of the inclusion and exclusion criteria of the content that is to be included in the review.
The literature that is unpublished and mostly found in dissertations and theses may contain information that is very relevant to the research topic. If it is important for the research question in hand, then the researcher can decide to pursue grey literature by identifying specific university libraries.
A decision is also to be made whether ancestry or descendency approach is to be used, as very often they provide the most specific content required. Ancestry approach deals with tracing the articles cited by the referenced article, whereas descendency approach identifies the articles which have cited the referenced article. The ancestry approach can also be termed as 'backward chaining', the process of looking at good resources of the past cited in the article being referenced. As the cited lists increase, the network of resources will improve too. The descendency approach is also known as 'citation tracking' or 'forward chaining'. The researcher can look forward at research conducted and published after the referenced article and see how the information has given way to new information or how this information is used to support findings of other researches. A highly influential article or landmark article is cited by many recent related articles, and the researcher can find more information from these articles. It is important to remember that author names are cited differently based on the citation style and hence there is a high chance of missing the original author.,
Choosing basic or advance search option
Most databases provide the option of 'basic' or 'advanced' search. Broad and general topics require only a basic search. Advanced search option opens up a whole lot of limits to narrow down research from language, year of publication, full text, search areas, operators and so on.
After these aspects are sorted out and decisions for search approach made, then the researcher starts on the nest building block and identifies the search databases.
| Block 2 – Search Databases|| |
A database is a searchable collection of information. Databases cover either a single subject or multiple subjects, one disciplinary area or multidisciplinary areas. Databases may be general or subject focussed. They may be subject specific with subsections for different fields within that subject. Before the computer era, databases referred to bibliographic compilation of articles and journals from recognised publishers such as CINAHL. With the advent of online databases, such printed bibliographic databases have become obsolete.
Difference between search database and internet
Databases may be electronic sources, but are different from 'internet' or 'open web'. They may use internet technology to transmit the information they contain, but are accurate and credible compared to internet, and the term 'database' should not be confused with 'internet'. The internet is a worldwide system of computer networks with search engines such as Google and Yahoo. Databases and internet both yield search results, but the internet is unstructured whereas databases focus on specific subject areas and specific types of information with peer-reviewed and referred resources. There is no control over the organisation, level, authenticity or quality of information in the internet. Results consist of a list of websites, and none can ensure accuracy and reliability of the internet. Databases, on the other hand, are sold to libraries by vendors or companies such as EBSCO and ProQuest, who manage the databases and ensure that they are updated and provide accurate information.
Difference between search database and search engine
Both are tools for finding answers, but database is a searchable collection of data describing a set of concepts. Each concept is described by an electronic record, made up of a number of rich descriptive records or fields, such as 'title', 'author', 'year' and 'publisher'. They cost money as they contain the best of research and cannot be accessed via a search engine. A search engine is also a database, but it searches a very large database of websites, which is automatically updated by a program to look for new websites. As there are no quality controls over websites, the information may not be complete, and the content may be unreliable. They are great source for quick answers and checking sources not covered by library databases. Common databases used by nurses are EBSCO, PubMed, ProQuest, Ovid, etc.
Types of databases
There are three levels of content in databases, namely indexing, full-text and combination databases.
- Indexing databases – They provide a citation and an abstract with no direct access to full text of journals. Citation details found in these databases can be used to search for full text in library search
- Full-text databases – They provide access to complete articles and usually contain the publications of just one publisher
- Combination databases – They hold a mixture of both citations with abstracts, and full-text journal articles, covering a range of different subject areas.
The choice of the type of database will determine the effectiveness of the search as well as the output of the search and so it is important that the nurse researcher takes some time to reflect and record the chosen databases. Once this decision is made, the next building block would be to dwell on search questions.
| Block 3 – Search Questions|| |
The second step of the search strategy is to determine the search questions. It is easy to miss something you are not looking for, and hence, it is important to formulate a relevant search question. Typical questions for evidence generation contain only those terms that are useful for the search. The patient-intervention-comparison-outcome format [Table 1] is a favourite among many researchers and is a must in evidence-based research. Ex: Is breakfast good for weight loss?/Is consuming breakfast effective in causing weight loss among obese post-menopausal women?
| Block 4 – Search Terms/Phrases|| |
Search terms are keywords or phrases that reflect the key concepts of the research topic. The best way to ensure an effective search is to first list down all the terms or phrases associated with the research concept, including synonyms and alternate spellings. They are instrumental in providing us with all the relevant articles. Creating a tabular column with all relevant search keywords will help in ensuring all content related to the topic is retrieved as different authors may use different terms to describe the same concept. The central keywords can be considered as facets and synonyms can be listed next to them or below them. Ex: Facets in a study maybe breakfast (breakfast, first meal and morning meal), weight loss, obese (overweight and fat), etc. A simple search of one or two words may yield over a million results. Sometimes, a group of two or more words may convey the idea. To get fewer, more relevant results, double quotes may be used in certain databases and this is called as phrase searching.,, Ex: 'weight loss', 'obese menopausal' and 'menopausal obese'.
| Block 5 – Search Areas/Fields|| |
When a search term is entered in the search box, the search will be done across many fields automatically. In order to narrow down the results, the researcher can specify expectations. If the key term is expected to be in the title, then 'Title' alone is to be chosen as the search area. Many such areas are given in different databases and they vary from database to database. It is up to the researcher to choose what would be efficient in yielding effective results. Ex: title, abstract, title/abstract, title/abstract/full text, author, publisher, language, volume, ISSN and DOI. Explicit search can be done by using the syntax: 'field: (query)' in few databases. This will turn up only those specified terms in the specified fields. For nurse researchers, awareness regarding the Medical Subject Headings (MeSH) terms is mandatory. It is the vocabulary or thesaurus of terms of the US National Library of Medicine and is used in all major databases such as MEDLINE, CINAHL, PubMed and Cochrane Library. There is a hierarchy of terms from top broad category to broader terms and then narrower terms. When MeSH search is used, the publisher supplied citations, and in process, citations are automatically excluded. The next building block of search operators is most crucial and differs from database to database.
| Block 6 – Search Operators|| |
There are different types of search operators, namely Boolean/positional operators, wildcards, truncations marks and plurals.
These include article title, author and journal. Ex: 'Cooper' under 'author' will yield all articles authored by anyone identified as Cooper.
Boolean operators are used to maximise the effectiveness of searches by formatting and combining keywords into searches. AND, OR and NOT are recognised Boolean operators. AND is used to combine words for different concepts. Ex: Obese AND post-menopausal will retrieve results where both obese and post-menopausal are found. OR helps in adding synonyms and similar concepts and tells the database to turn up results where one of the words appears. NOT is used to exclude terms to narrow the search. Ex: weight loss NOT obese will retrieve results that include 'weight loss', but do not include 'obese'. These operators must be written in ALL CAPS. All search terms will usually be combined with AND operator by default.,,
When researchers want to locate records where search terms are in close proximity to each other, then positional operators may be used. NEAR, ADJ, WITH and SAME are positional operators. NEAR operator is used to locate records in a field where all specified terms are next to each other. Ex: If 'breakfast NEAR weight loss' is used in title field, the database will retrieve only articles where the title contains 'breakfast' and 'weight loss' next to each other. ADJ operator helps in locating records in a field containing all search terms adjacent to each other, in the order that is entered. Ex: If 'Breakfast ADJ weight loss' is entered in title field, only records with 'breakfast weight loss' will be retrieved and not 'weight loss breakfast'. WITH operator helps in locating records within a field such as title, where a sentence with all specified terms will be retrieved. Ex: If 'breakfast' and 'weight loss' are entered in the field of abstract, only records which have both 'breakfast' and 'weight loss' in the same sentence in the abstract will be retrieved by the database.
In certain databases, search terms are enclosed in quotes and then tilde (~) is used followed by a number indicating distance to be allowed between search terms. Ex: 'obese menopause' ~10 will yield results where 'obese' and 'menopause' appear within 10 words of each other.
Wild cards/truncation marks
Different databases have different specifications for wild cards and truncation marks. A wild card in one database may be called a truncation mark in another. In general, when there are various forms of a word, it can be cut back to the root word with a truncation symbol. The asterisk (*) will match zero or more characters within or at end of a word. Ex: Educat* would tell the database to look for all possible endings to that root such as educate, educated, education, educational or educator. Sometimes, there may be variations in spellings, especially between British English and American English. The question mark (?) will yield results with different alphabets or numbers in its place., Ex: colo?r would yield results with both color and colour. #, ! and ? are usually considered to be wildcards with their usage slightly differing between websites. They cannot be used as the first character of a search, and they cannot be used within double quotes.
How databases search for single/plural version of key concepts can have a huge impact on search results, as some databases will automatically search for plural version of a singular term. The symbol + helps in obtaining the results with plurals too. Ex: subject + will yield 'subject' and 'subjects'.
| Block 7 – Search Limits|| |
Most databases provide a range of options that enable you to retrieve relevant search results by manipulating specific elements such as population, age, year of publication, full text, language, article type such as case studies, reports, review articles, conference papers and type of research such as review, randomised control trial, etc. The inclusion and exclusion criteria determined during the stage of planning the location of articles play an important role here.
| Principles Guiding Literature Search|| |
No computer can understand the human brain the way humans themselves can. If nurses want computers to yield results, the concepts and related terms need to be sorted out in the human brain before they can be fed into the computer. From databases, key terms, range of years to population, every aspect is to be thought of, deliberated on and recorded down. This is the crucial part of the search process.
Moving from general to specific
It is well known that the more specific you are in your search terms, the lesser the yield of content. When general terms are used, the yield may be too many that the nurse may not have the time to skim through all the yields.
Recording the search
It is essential to record search strategy to enable retracing of steps if needed. Every aspect of search from databases to terms to limits ought to be recorded.
Building on what's known
Every new information yielded through research or formulation or brainstorming helps in finding something else. Nurse researchers will find it helpful to fist gather known concepts about the research concept before delving into the unknown.
| Conclusion|| |
The task of searching, being the very first step in the process of review of literature, directs the entire course of the review of literature. The review can only be as good as the sources retrieved during the search and so this step is to be given utmost importance during the review process.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Williams JJ, McGowan J, Finke LA, Sharpley-Whiting TD, Leitch VB, Cain WE. The Norton Anthology of Theory and Criticism. USA: W. W. Norton & Company; 2010. Available from: https://wwnorton.com/books/9780393602951
. [Last accessed on 2020 Dec 16].
Snyder H. Literature review as a research methodology: An overview and guidelines. Bus Res 2019;104:333-9.