- Prepare a one-page design rationale, operational matrix and risk log for supervisor review.
Definition and purpose: A research proposal is a structured document that describes the overall design of the proposed study. Its purpose is to convince reviewers/supervisors of the importance, methodological soundness and feasibility of the study.
Research title — how to write an effective title
The title should be concise, clear and include the main variables and sample. Avoid vague expressions. Suggested format and examples:
- Format: "[X] and [Y] in [Z]: [study design]"; use "effect" only when the design supports causal estimation.
- Observational example: "Active learning and academic motivation among engineering students: a cross-sectional study".
- Tip: Avoid excessively long titles and long subtitles.
Abstract — writing an effective abstract
Follow the institution’s abstract limit; 200–300 words is only an illustrative range and include: background, problem statement, research aim, general methods, planned contribution and the significance of the study; do not invent future findings. For quantitative projects briefly mention sampling method and main analysis techniques.
Problem statement and research significance
The problem statement must describe what the problem is, why it is important, and how your research can help address it. Use up-to-date facts and statistics when applicable to demonstrate relevance.
- Description of the current situation or gap
- Link the problem to scientific or social needs
- Sub-questions that derive from observed indicators
Literature review and gap identification
The literature review must be organized, topic-focused and critical. Practical steps:
- Define keywords and search strategy (databases: Scopus, Web of Science, PubMed, Google Scholar)
- Create a summary table (Author, Year, Method, Sample, Key Findings, Limitations)
- Critical appraisal and extraction of theoretical, methodological and practical gaps
- State explicitly what is unknown and how your study will address it
Objectives and SMART framework
Objectives should be clear, measurable and aligned with the research question. Example general and specific objectives:
- General objective: "Assess the effect of method X on variable Y in group Z over 12 months"
- Specific objective 1: "Compare mean Y between control and intervention groups at month 3"
- Specific objective 2: "Identify predictors of Y using multiple regression"
Check each objective against SMART criteria: Specific, Measurable, Achievable, Relevant, Time-bound.
Research questions and hypotheses
Research questions need to be clear and testable. Where the study goal calls for hypotheses, specify testable claims; not every quantitative study requires a directional hypothesis. Specify the estimand, population and outcome; state a testable hypothesis only where the methodological approach requires one.
Methodology: design, sampling, instruments
The methodology section must clearly describe the study design (quantitative/qualitative/mixed), population and sampling, data collection instruments, data collection process and analysis procedures.
Sampling
Describe sampling methods (probability and non-probability), sample size calculation (for example, a proportion-estimation formula only where its sampling and precision assumptions apply) and justification for choices.
Instruments and scales
Discuss instrument validity and reliability: content validity, convergent/divergent validity, and Cronbach's alpha for internal consistency under its assumptions; a high value alone establishes neither validity nor every form of reliability.
Data analysis — statistical plan
The analysis plan should include step-by-step procedures: descriptive analysis (mean, median, SD), assumption testing, selecting appropriate regression models, multiple testing corrections, and model validation. Record the software and exact version actually used in the analysis.
Ethical issues and approvals
Explain need for ethics committee approval, informed consent forms, data privacy and secure storage.
Timeline and budget estimation
Provide a sample phased timeline (Gantt-like) with main phases: design, data collection, analysis, writing. Provide a simple budget template covering personnel, travel, software, equipment, printing and publication costs.
Appendices, questionnaires and supporting documents
Include questionnaires, consent forms, codebooks and sample software outputs in appendices. Organize these so a third party could reproduce the study.
Quick proposal checklist
- Title, abstract, problem statement
- Literature review and gap extraction
- SMART objectives, questions and hypotheses
- Full methodology and statistical plan
- Key references and appendices
From an idea to a feasible design
A compelling title is only a starting point. Narrow the question, explain access to data and justify why its answer matters. Sample size must reflect the analysis goal, anticipated effect, precision and study design. One general formula does not suit every study.
- Literature matrix: methods, samples, findings and limitations
- Operational definitions and inclusion/exclusion criteria
- Measurement instruments, analysis plan and necessary approvals
- Milestones, data access risks and contingency plan
A decision framework for a coherent proposal
Suppose you want to examine classroom participation and academic motivation. First decide whether you intend to describe an association or evaluate an intervention. Calling a cross-sectional survey an “effect” study may overstate its design. Define suitable instruments, recruitment and potential confounders. Question, measurement and analysis should follow the same logic.
| Component | Review question |
|---|---|
| Problem | What remains uncertain or contradictory? |
| Aim | Which concrete output addresses the problem? |
| Methods | Can the design and data support the answer? |
| Feasibility | Are approvals, participants and resources accessible? |
Three costly mistakes
Collecting data before defining analysis, choosing a topic only because it is fashionable and using an instrument without checking population suitability can create substantial rework. Review a short statement of problem, design, outputs and risks with your supervisor. Changes after seeing results require transparent justification.
Worked case and implementation decisions
The following is a fictional teaching case. Do not use its numbers or wording as actual study findings.
Turn “Why do students participate less?” into an executable question. In this fictional case the aim is to estimate the association between use of academic support and a participation score within one semester. The design is observational, so the title uses association. Establish access to baseline measures, sampling and permissions before collection. Define a directional hypothesis only with a theoretical rationale and label exploratory analyses separately. The matrix aligns question, measurement and analysis; it does not supply a universal sample size or confounder set.
| Stage | Teaching example | Verification question |
|---|---|---|
| Question | Support–participation association in one semester | Are population, timing and variables explicit? |
| Measure | Permitted instrument with relevant evidence | Are definitions and measurement quality documented? |
| Design | Observational, without a causal-effect claim | Do title and aim match the design? |
| Analysis | Model suited to score and class structure | Have assumptions, missingness and clustering been assessed? |
| Feasibility | Permissions, access and fallback | Is the project executable before collection? |
Exercise output: Prepare a one-page design rationale, operational matrix and risk log for supervisor review.
Sources and further reading
Official sources for verification and further reading