RESEARCH METHODOLOGY
Module II Guide
Literature review.
Instructor: Prof. D. K. Muriithi, PhD.
1. Literature Review
Definition: A critical synthesis of existing research on the topic to identify gaps, trends, and
theoretical frameworks. (The literature review critically examines existing research relevant to
your topic)
Purpose:
• Clarify what is already known and unknown.
• Refine research problem and objectives.
• Identify theoretical frameworks and models.
• Discover suitable data sources and methods.
• Justify your choice of methods and contribution.
• Establish context
• Avoid duplication
• Support hypothesis formulation
Steps
• Define scope: concepts, population, time frame, geography.
• Search systematically: databases (e.g., Web of Science, Scopus, PubMed, EconLit),
institutional repositories, government reports.
• Use keywords and Boolean operators: “adoption of technology” AND “logistic
regression” AND “Kenya”.
• Screen and select: using inclusion/exclusion criteria.
• Read critically: note methods, data, findings, gaps, limitations.
Organizing the review
• Thematic: group by themes or variables (e.g., socio-economic factors, institutional
factors, methodological approaches).
• Methodological: group by study designs or models (e.g., logit vs probit, panel data
models, mixed models).
• Chronological: useful if the field has evolved over time.
For each key study, highlight:
• Objectives and context.
• Data and design (cross-sectional, panel, experimental).
• Statistical methods used.
• Main findings.
• Strengths and limitations.
Conclude with:
• Identified gaps.
• How your study will address at least one gap (scope, data, methods, or context).
Structure:
1. Introduction: Purpose of the review
2. Themes/Concepts: Organize by key concepts or theories
3. Gaps: Identify areas needing further research
4. Conclusion: Justify your research based on the review
Example:
In a study about survival analysis in cancer patients:
• Review classical parametric models (Weibull, Exponential)
• Discuss semi-parametric approaches (Cox PH model)
• Evaluate recent machine learning alternatives (Random Survival Forests, Gradient
Boosted Models)
• Identify need for hybrid models combining parametric and ML methods
Tools:
• Use databases like Google Scholar, PubMed, Scopus, Web of Science
• Manage references using Zotero, Mendeley, or EndNote
•
NB: Use PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses)
guidelines if conducting a systematic review.
2. Bibliography / References
Definition:
A list of all sources cited in the research proposal or report, following a specific citation style
(APA, Chicago, IEEE, etc.).
Key points/Guidelines:
• Use credible academic sources (peer-reviewed journals, books, conference papers)
• Ensure consistency in formatting
• Use a consistent referencing style (e.g., APA, IEEE, Chicago, Vancouver) as required by
your institution.
• Include all and only sources cited in the text (for a thesis, you may sometimes also list
key background readings).
• Typical components for journal articles: authors, year, title, journal, volume, issue, pages,
DOI.
• For books: authors, year, title, edition, publisher, place.
• For reports/web sources: author or institution, year, title, URL, and access date (if
relevant).
For applied statistics:
• Cite statistical methodology sources (textbooks, method papers) used to justify models
and assumptions.
• Cite software packages appropriately (e.g., R packages, Stata manuals).
Example (APA Style):
Ntzoufras, I. (2009). Bayesian Modeling Using WinBUGS. John Wiley & Sons.
Therneau, T. M., & Grambsch, P. M. (2000). Modeling Survival Data: Extending the Cox Model.
Springer.
WHO. (2022). World Health Statistics 2022: Monitoring Health for the SDGs. Geneva: World
Health Organization.
Tool Recommendation:
Use reference managers (e.g., Zotero, Mendeley, EndNote) and LaTeX/BibTeX or your word
processor’s tools to generate automatic citations and manage references systematically.
AI tools for literature reviews
• Elicit: A powerful tool for systematic reviews that can extract specific details (e.g.,
methodologies, findings) from hundreds of papers into a customizable table.
• SciSpace: Includes a "Chat with PDF" feature to ask questions about specific
papers and an AI Literature Review tool that synthesizes findings from multiple
sources.
• Scholarcy: Automatically breaks down dense academic articles into concise,
structured "summary flashcards" that highlight key objectives and results.
• [Link]: Uses "Smart Citations" to tell you if a paper has been supported or
contradicted by subsequent research.
• Consensus: An AI search engine that extracts evidence-based answers directly
from peer-reviewed papers. It is particularly useful for verifying scientific
agreement on specific questions using its "Consensus Meter".
• Semantic Scholar: A free, AI-powered academic search engine that provides
"TLDR" one-sentence summaries of papers and tracks citation influence.
• AnswerThis: [Link]
How to Use AI Ethically for Your University Coursework
Framework: The guidelines are structured using the acronym ETHICAL.
E – Embrace Awareness
• Understand AI tools’ capabilities and limitations.
• Stay informed about how AI is evolving in academic contexts.
T – Transparency is Key
• Clearly disclose any use of AI in your coursework.
• Be honest and accurate about how AI contributed to your work.
H – Highlight Your Creativity
• Ensure your own intellectual and creative input is evident.
• AI should support—not replace—your original thinking.
I – Integrity Always
• All submitted content must be your own unless properly cited.
• This includes AI-generated material, which must be referenced appropriately.
C – Cultivate Understanding
• Use AI to enhance learning, not bypass it.
• Focus on demonstrating your personal comprehension of the subject.
A – Append AI Usage
• Include an AI Appendix with every assignment.
o Detail the AI tools used, their purpose, and prompts given.
o If no AI was used, explicitly state that.
L – Learn to Discriminate
• Develop judgment about when AI use is appropriate.
• Avoid unethical or prohibited applications of AI in academic work.