Unit 1: Literature Review Techniques and Tools - Subjective Questions

A literature review is a systematic, critical, and comprehensive survey of existing scholarly work relevant to a specific research topic or question.

Significance in dissertation research:

• Establishes context: It situates the researcher's work within the existing body of knowledge.
• Identifies gaps: Reveals unexplored or under-researched areas that justify the new study.
• Avoids duplication: Ensures the researcher does not repeat work that has already been done.
• Informs methodology: Helps in selecting appropriate methods, tools, and theoretical frameworks.
• Builds credibility: Demonstrates the researcher's command over the subject.
• Refines the research question: Sharpens the focus and scope based on prior findings.

In summary, a literature review forms the intellectual foundation upon which the entire dissertation is built.

The literature search follows a structured, iterative procedure:

1. Define the research topic and objectives: Clearly articulate what you are searching for.
2. Identify keywords and search terms: Break down the topic into core concepts and synonyms.
3. Select appropriate databases and sources: e.g., IEEE Xplore, Scopus, Web of Science, Google Scholar, PubMed.
4. Formulate search queries: Use Boolean operators (AND, OR, NOT), truncation, and wildcards.
5. Apply filters: Restrict by date range, publication type, language, and subject area.
6. Screen results: Review titles and abstracts to filter relevant material.
7. Retrieve full-text documents: Access and download relevant papers.
8. Organize references: Use reference managers like Mendeley, Zotero, or EndNote.
9. Iterate and refine: Refine keywords and queries based on initial findings (e.g., citation chaining).

This systematic approach ensures comprehensive coverage and reproducibility of the search.

Key point: A narrative review offers a broad overview based on the author's expertise, while an SLR uses a transparent, replicable methodology (often following PRISMA guidelines) to reduce bias and ensure completeness.

Boolean operators are connecting words used to combine or exclude keywords in a search query, refining the results.

The three main operators:

• AND — Narrows the search; retrieves records containing all terms.
  • Example: machine learning AND healthcare → results must contain both terms.
• OR — Broadens the search; retrieves records containing any of the terms.
  • Example: neural network OR deep learning → results contain either term.
• NOT — Excludes terms; removes unwanted results.
  • Example: cancer NOT lung → results about cancer but not lung cancer.

Additional techniques:

• Truncation (*): comput* retrieves computer, computing, computation.
• Phrase searching (" "): "artificial intelligence" retrieves the exact phrase.
• Nesting ( ): (diabetes OR obesity) AND diet groups terms logically.

Proper use of Boolean logic increases precision and recall in database searches.

Literature sources are classified based on their proximity to the original research:

1. Primary Sources

• Original, first-hand research or data.
• Examples: research articles, conference papers, theses, patents, experimental reports.
• Highest credibility for original findings.

2. Secondary Sources

• Interpret, analyze, or summarize primary sources.
• Examples: review articles, textbooks, meta-analyses, commentaries.
• Useful for gaining context and understanding trends.

3. Tertiary Sources

• Compile and index primary/secondary sources.
• Examples: encyclopedias, dictionaries, bibliographies, databases, handbooks.
• Useful for initial orientation and locating other sources.

Summary: Primary sources provide original evidence, secondary sources offer analysis, and tertiary sources aid in discovery and navigation of the literature.

Reference management tools help collect, organize, cite, and share research references efficiently.

Popular tools:

• Zotero — Free, open-source; browser integration; excellent for web capture.
• Mendeley — Free with cloud storage; PDF annotation; social networking features.
• EndNote — Commercial; powerful for large libraries; strong journal formatting.
• RefWorks — Web-based; institutional subscriptions.
• JabRef — Open-source; ideal for LaTeX/BibTeX users.

Key features these tools provide:

• Automatic citation generation in multiple styles (APA, IEEE, MLA, Chicago).
• PDF organization and annotation.
• In-text citation and bibliography insertion via word processor plugins.
• Duplicate detection and library synchronization.
• Collaboration and sharing of reference collections.

These tools save time, reduce citation errors, and ensure consistency across the dissertation.

PRISMA stands for Preferred Reporting Items for Systematic Reviews and Meta-Analyses. It is a standardized, evidence-based framework for reporting systematic reviews transparently.

Main components:

• Checklist: A 27-item list covering title, abstract, methods, results, and discussion.
• Flow diagram: Visualizes the flow of information through the four phases of study selection:
  1. Identification — records found through databases.
  2. Screening — removal of duplicates and irrelevant records.
  3. Eligibility — full-text assessment against inclusion criteria.
  4. Included — final studies included in the review.

Importance:

• Ensures transparency and reproducibility.
• Reduces selection and reporting bias.
• Improves the quality and credibility of reviews.
• Provides a clear audit trail of the search and selection process.

PRISMA has become the gold standard for reporting systematic reviews across disciplines.

Citation chaining is a technique for discovering additional relevant literature by following the citation links of a known, relevant paper.

Two types:

1. Backward citation chaining (Backward snowballing):

   • Examines the reference list of a paper.
   • Identifies older, foundational works the paper builds upon.

2. Forward citation chaining (Forward snowballing):

   • Identifies newer papers that have cited the known paper.
   • Uses databases like Google Scholar ("Cited by") or Scopus.

Usefulness:

• Uncovers papers missed by keyword searches.
• Traces the evolution of ideas over time.
• Identifies seminal works and research trends.
• Complements database searching for comprehensive coverage.

Citation chaining is especially valuable in systematic reviews to ensure no important study is overlooked.

Inclusion and exclusion criteria are predefined rules that determine which studies are considered relevant and eligible for a literature review.

Inclusion criteria — Characteristics a study must have to be included:

• Relevant topic or research question.
• Specific time period (e.g., 2015–2025).
• Publication type (peer-reviewed journals, conferences).
• Language (e.g., English).
• Methodology or population of interest.

Exclusion criteria — Characteristics that disqualify a study:

• Off-topic or tangential content.
• Non-peer-reviewed sources (blogs, opinions).
• Duplicate publications.
• Insufficient methodological rigor.

Importance:

• Ensures relevance and quality of selected studies.
• Reduces bias and subjectivity.
• Makes the review reproducible and transparent.
• Keeps the review focused and manageable.

Clearly documented criteria are a hallmark of a rigorous systematic review.

Precision and recall are metrics that measure the effectiveness of a literature search strategy.

Recall measures completeness — the fraction of all relevant documents that were retrieved:

Precision measures accuracy — the fraction of retrieved documents that are actually relevant:

Trade-off:

• A broad search (using OR, truncation) increases recall but lowers precision (more noise).
• A narrow search (using AND, specific terms) increases precision but may lower recall (missing relevant items).

Combined metric — F-measure:

Researchers aim for an optimal balance depending on whether comprehensiveness or accuracy is more critical for the review.

Academic databases are curated repositories of scholarly literature, essential for comprehensive searching.

Major databases and search engines:

• Google Scholar — Free, broad coverage across disciplines; includes grey literature.
• Scopus — Large multidisciplinary abstract and citation database (Elsevier).
• Web of Science — Citation indexing; strong for impact analysis (Clarivate).
• IEEE Xplore — Engineering, computer science, and electronics.
• PubMed / MEDLINE — Biomedical and life sciences.
• ACM Digital Library — Computing and information technology.
• ScienceDirect — Full-text scientific journals (Elsevier).
• SpringerLink — Books, journals across sciences.
• JSTOR — Humanities and social sciences archive.

Selection considerations:

• Subject coverage relevant to the research area.
• Access rights through institutional subscriptions.
• Search features (filters, citation tools).
• Currency and comprehensiveness of indexed content.

Using multiple databases ensures broader and more reliable coverage.

Critical evaluation involves systematically assessing the quality, validity, and relevance of a paper rather than merely summarizing it.

Key evaluation criteria:

1. Authority: Are the authors credible? Is the publication peer-reviewed and reputable?
2. Objectivity of purpose: Is the research question clearly stated? Is there bias?
3. Methodology: Are the methods appropriate, valid, and clearly described?
4. Data and analysis: Is the sample adequate? Are the statistical methods sound?
5. Findings: Are conclusions supported by the data? Are limitations acknowledged?
6. Currency: Is the work recent and relevant to the current state of the field?
7. Contribution: What is the novelty and significance of the work?

Techniques:

• Use a critical appraisal checklist (e.g., CASP).
• Compare findings with other studies.
• Note strengths and weaknesses explicitly.

Critical evaluation transforms a review from a descriptive summary into an analytical synthesis.

A research gap is a question, problem, or area that has not yet been adequately addressed or answered in the existing body of literature.

Types of research gaps:

• Evidence gap: Conflicting or inconclusive findings.
• Knowledge gap: Unexplored topics or theories.
• Methodological gap: Limitations in methods used by prior studies.
• Population gap: Under-studied populations or contexts.
• Practical/application gap: Theory not yet applied in practice.

How a literature review helps identify gaps:

• By synthesizing existing studies, patterns and inconsistencies become visible.
• Authors often state limitations and future work directions.
• Comparing studies reveals unanswered questions.
• Mapping the literature (e.g., in a concept matrix) exposes areas with little coverage.

Identifying a clear research gap justifies the need for the dissertation and defines its contribution to knowledge.

Synthesis is the process of integrating findings from multiple sources into a coherent, thematic narrative rather than listing them individually.

Steps in synthesizing literature:

1. Read and take structured notes on each source.
2. Group sources by themes, theories, methods, or chronology.
3. Identify relationships — agreements, contradictions, and trends.
4. Create a synthesis matrix (concept matrix) mapping authors against themes.
5. Write thematically — organize the review around ideas, not around individual papers.

Common organizational structures:

• Thematic: Organized by topics or concepts (most common).
• Chronological: Traces the development of ideas over time.
• Methodological: Groups studies by research approach.
• Theoretical: Organized by theoretical frameworks.

Goal: Produce an integrated analysis that highlights what is known, what is debated, and what is missing — leading naturally to the research gap.

Keywords are the core terms that represent the main concepts of a research topic and drive the search process.

Importance:

• Determine the relevance and comprehensiveness of search results.
• Poor keyword selection leads to missed relevant studies or irrelevant clutter.

Strategies for effective keyword selection:

• Break down the topic into key concepts.
• Identify synonyms and related terms (e.g., car, automobile, vehicle).
• Consider spelling variations (e.g., behaviour / behavior).
• Use abbreviations and full forms (e.g., AI / artificial intelligence).
• Extract keywords from relevant papers' titles, abstracts, and author-supplied keywords.
• Use controlled vocabulary such as MeSH terms (in PubMed) or database thesauri.

Refinement:

• Combine keywords using Boolean operators.
• Iteratively refine based on initial search results.

Well-chosen keywords are the foundation of an effective and efficient literature search.

White (published) literature refers to formally published, peer-reviewed material, while grey literature refers to material produced outside traditional commercial or academic publishing channels.

When to use grey literature:

• To reduce publication bias in systematic reviews.
• To access recent developments not yet formally published.
• In emerging or applied fields where reports dominate.
• To include practitioner and policy perspectives.

Caution: Grey literature requires careful critical appraisal for reliability since it lacks formal peer review.

Plagiarism is the act of using another person's ideas, words, or work without proper acknowledgment, presenting them as one's own. It is a serious academic and ethical violation.

Types of plagiarism:

• Direct plagiarism: Copying text verbatim without quotation or citation.
• Paraphrasing plagiarism: Rewording without attribution.
• Self-plagiarism: Reusing one's own previously published work without disclosure.
• Mosaic plagiarism: Mixing copied phrases with original text.

Strategies to avoid plagiarism:

• Always cite sources for ideas, data, and quotations.
• Use quotation marks for direct quotes.
• Paraphrase properly and still cite the original source.
• Maintain accurate notes distinguishing your ideas from others'.
• Use reference management tools for consistent citation.
• Run work through plagiarism detection software (e.g., Turnitin, iThenticate).

Maintaining academic integrity is essential for the credibility of the dissertation.

Citation styles are standardized formats for acknowledging sources within a document and in the bibliography.

Common citation styles:

• APA (American Psychological Association):

  • Used in social sciences, psychology, education.
  • Author–date in-text: (Smith, 2020).

• MLA (Modern Language Association):

  • Used in humanities, literature, arts.
  • Author–page in-text: (Smith 45).

• IEEE (Institute of Electrical and Electronics Engineers):

  • Used in engineering, computer science.
  • Numbered in-text: [1], references in citation order.

• Chicago/Turabian:

  • Two systems: Notes-Bibliography (footnotes) and Author-Date.
  • Used in history, arts, and sciences.

• Harvard:

  • Author–date style, widely used across disciplines.

Key differences lie in:

• In-text format (numbered vs. author-date).
• Reference list ordering (alphabetical vs. order of appearance).
• Punctuation and element order.

Consistency in applying one style throughout the dissertation is essential.

Snowballing is a systematic technique for finding additional relevant papers by using the reference lists and citations of an initial set of known papers (the start set).

Complete procedure:

1. Define the start set: Identify an initial group of highly relevant papers through database searching.
2. Backward snowballing:
   • Examine the reference lists of the start-set papers.
   • Screen references by title, then abstract, then full text.
   • Add relevant papers to the review set.
3. Forward snowballing:
   • Find papers that have cited the start-set papers (via Google Scholar, Scopus).
   • Screen and add relevant ones.
4. Iterate: Repeat backward and forward snowballing on newly added papers.
5. Stop: When no new relevant papers are found (saturation).

Advantages:

• Reduces dependence on keyword accuracy.
• Captures papers using different terminology.
• Effective for interdisciplinary topics.

Limitation: Depends heavily on the quality of the start set. Wombining snowballing with database searching yields the most complete coverage.

Conducting a thorough literature review presents several challenges:

1. Information overload

• Challenge: Overwhelming number of publications.
• Solution: Use precise keywords, filters, and clear inclusion/exclusion criteria.

2. Access limitations

• Challenge: Paywalled articles.
• Solution: Use institutional access, open-access repositories, interlibrary loans, or contact authors.

3. Keeping the search current

• Challenge: New papers published continuously.
• Solution: Set up database alerts and RSS feeds.

4. Managing references

• Challenge: Tracking numerous sources.
• Solution: Use reference managers (Zotero, Mendeley).

5. Assessing quality

• Challenge: Distinguishing credible from weak sources.
• Solution: Apply critical appraisal checklists; prefer peer-reviewed sources.

6. Avoiding bias

• Challenge: Selectively including supporting studies.
• Solution: Follow systematic, documented protocols (PRISMA).

7. Synthesis difficulty

• Challenge: Integrating diverse findings.
• Solution: Use synthesis matrices and thematic organization.

Addressing these challenges systematically ensures a rigorous and comprehensive literature review.