How to write a thesis proposal

What is a proposal?

A proposal (research plan) is the document you write before starting your thesis or dissertation to show your supervisor and committee what problem you will study, why, with what method and in what time. An approved proposal is the basis of your work; written well, it makes the thesis chapters much easier.

A good proposal is clear (a reviewer understands the study in a few minutes), feasible (possible with your time, data and resources) and aligned (problem, objectives, questions and analysis fit together).

Proposal forms differ between universities. Always follow your own faculty's form and regulations; the sections in this guide appear in almost every form.

Before you write: choosing a topic

Most proposal problems start with an unsuitable topic. Test yours with these questions:

• Interest and expertise: are you interested in it, and does it fit your field and your supervisor's expertise?
• Data: do you have access to the population, sample, lab or data you need?
• Time and cost: can it be done within the programme's time limit and your budget?
• Originality: does it add something — a new population, variable, method or combination?
• Scope: is it too broad? “Factors affecting success” is broad; “The role of organisational support in the work commitment of nurses in teaching hospitals in city X” is specific.

Good sources of topics are the “suggestions for future research” sections of recent papers and theses, real problems in organisations and industry, and systematic reviews in your field.

Map of the proposal sections

The table shows the usual sections of a proposal form and the question each must answer:

Title and keywords

In one sentence, the title should say which variables are studied, in which population and sometimes where and when.

Questions the title should answer

• What is the main variable (or phenomenon)?
• Which relationship or effect is studied?
• Who is the study population?
• Is there a time or place limit?

Tips

• Titles are usually 10–20 words; avoid filler such as “A study and investigation of…”.
• Do not use unfamiliar abbreviations in the title.
• Example: “The role of transformational leadership in employee innovation, mediated by psychological empowerment (case study: knowledge-based companies in city X)”.
• Write 3–7 keywords that other researchers would use to search for this topic; repeating the title is not enough.

Problem statement

The problem statement is the most important part of the proposal and gets the most attention from reviewers. Its aim is to convince the reader that a real, researchable problem exists. The best structure is a funnel, from general to specific:

1. Context: introduce the general topic and its importance in one or two paragraphs (with statistics and reliable sources).
2. Current situation: show what problem appears in practice or in current knowledge.
3. Desired situation: say what things would look like if the problem were solved.
4. Research gap: referring to previous studies, say what has not been studied yet or where results conflict.
5. This study's problem: end by stating the problem as one clear main question, for example “This study therefore seeks to answer the question…”.

Questions the problem statement should answer

• What exactly is the problem and where does it appear?
• What evidence (statistics, reports, research) shows it exists?
• What are its possible causes or factors?
• What have previous studies done and what do they lack?
• Which part of the problem does this study address?

Common mistake: filling the problem statement with definitions or the history of the topic. Every paragraph should bring the reader one step closer to the main question. Always give a source for statistical claims.

Significance of the study

Do not confuse this with the problem statement. The problem statement says “what the problem is”; significance says “why solving it is worth it, and why now”.

• Theoretical significance: what does the study add to knowledge and theory in the field?
• Practical significance: which organisation, group or decision will use the results?
• Urgency: what cost or harm continues if it is not done? Why is now the right time?

Vague sentences such as “this topic is very important” add nothing; show importance with figures, specific consequences or a need stated by an organisation.

Literature review

A proposal usually has two parts: the theoretical framework (definitions and theories the study builds on) and the review of previous research (related empirical studies).

What to write for each previous study

• Author(s) and year
• Aim or topic of the study
• Population, sample and method
• The main finding relevant to your study

For example: “Smith et al. (2022), in a correlational study of 240 primary-school teachers, showed that…”. Many faculties also want the review summarised in a table.

Tips

• Organise the review by theme, not just by year; put agreeing and conflicting studies side by side.
• End with a synthesis paragraph: what is clear, what conflicts and exactly where the gap lies. This synthesis supports your originality section.
• Use recent studies (usually the last five to ten years) from reputable journals; classic foundational works still have their place.
• Find and check every source yourself. Before citing, make sure it really exists (check title, journal, year and DOI) and says what you attribute to it.

Good databases include Google Scholar, Scopus, Web of Science and PubMed (for medical sciences), plus your country's national thesis and journal databases.

Originality

Originality answers the question: “What does this study have that previous studies did not?” It does not have to be revolutionary — just precise and defensible.

• A relationship not studied before (a new mediating or moderating variable)
• A new population or context (for example the first study in your country or a particular industry)
• A new method, instrument or analytical model
• Combining two separate fields, or resolving conflicting earlier findings

Writing “no study has ever been done on this” is a strong claim that reviewers will check. Be precise: “In the author's search of databases …, no study examining … in … was found”.

Research objectives

Objectives are usually written in three groups:

• General (main) objective: one sentence that states the main problem as a goal; usually close to the title.
• Specific objectives: the general objective broken into measurable goals, one for each relationship or variable.
• Applied objective: who will use the results, and for what decision.

Use measurable verbs: “determine”, “compare”, “examine the relationship between”, “identify”, “design”. Verbs such as “better understand” or “help” cannot be measured.

Each specific objective should have exactly one matching question or hypothesis, and each question or hypothesis should have a clear analysis method.

Beneficiaries and users of the results

Name the organisations, groups or people who will use the results (for example “the regional education department”, “HR managers in … companies”, “researchers in …”). If an organisation has commissioned or supports the study, mention it here.

Research questions and hypotheses

Use questions when you have no clear prediction (descriptive, exploratory and qualitative studies). A hypothesis is a testable prediction of a relationship or difference between variables, based on theory and previous research (quantitative correlational, causal and experimental studies).

A good hypothesis

• is a statement, not a question;
• states a relationship or difference between at least two variables;
• is testable — data can support or reject it;
• is directional if the literature shows the direction: “Transformational leadership has a positive, significant effect on employee innovation”.

In a proposal, hypotheses are usually written as research hypotheses (the researcher's claim), not statistical null hypotheses. Keep the number of questions and hypotheses reasonable; each must be answered separately in the results chapter.

Conceptual and operational definitions

Give two definitions for each key variable or concept:

• Conceptual (theoretical) definition: how theorists or reliable sources define the variable, with a citation.
• Operational definition: exactly how the variable is measured in your study, for example “the score a participant obtains on the 15-item … questionnaire on a five-point Likert scale”.

Also state each variable's type: independent, dependent, mediating, moderating or control. A conceptual model (a simple diagram of how the variables connect) makes this section much easier for reviewers.

Methodology

The methodology must be detailed enough for the reviewer to be sure the study is feasible and will answer the questions. It usually has these subsections:

1. Type of research

• By purpose: basic, applied or developmental.
• By data collection: descriptive (survey, correlational, causal-comparative), experimental or quasi-experimental, qualitative (phenomenology, grounded theory, thematic analysis, case study) or mixed methods.
• By time: cross-sectional or longitudinal.

2. Population, sample and sampling method

• Describe the population precisely, with its size if known.
• State how the sample size was set (for example Cochran's formula, the Krejcie & Morgan table, power analysis, or theoretical saturation in qualitative research). You can get the number with the sample size calculator.
• State the sampling method — simple random, stratified, cluster, convenience, purposive, snowball and so on — and why it suits the study.
• Give the inclusion and exclusion criteria.

3. Data collection instruments

Questionnaires (standardised or researcher-made), interviews, observation, tests, documentary and secondary data, or laboratory instruments. For a standardised questionnaire give the author, year, number of items, subscales and response scale.

4. Validity and reliability

• Validity: does the instrument measure what it should? (face and content validity through expert judgement, CVR and CVI, construct validity through factor analysis, convergent and discriminant validity)
• Reliability: does it give stable results? (Cronbach's alpha, composite reliability, test–retest; in qualitative research: credibility and dependability)
• If you have no data yet, say how validity and reliability will be assessed (for example with a pilot study of 30 people).

5. Procedure and data collection

List the steps in order: obtaining permissions, how you will reach the sample, how questionnaires will be distributed or interviews or experiments run, the time needed and how data will be recorded.

6. Data analysis

• Descriptive statistics: frequencies, percentages, means, standard deviations, tables and charts.
• Inferential statistics: the right test for each hypothesis — for example Pearson or Spearman correlation, independent or paired t-tests, ANOVA, regression, structural equation modelling, chi-square — and each test's assumptions (such as normality).
• Software: SPSS, AMOS, SmartPLS, R, Stata, MAXQDA, NVivo and so on.
• In qualitative research, explain the coding and analysis method (open, axial and selective coding, or the stages of thematic analysis).

Common mistake: choosing a statistical test without considering the type of variable and hypothesis. For each hypothesis, say which test will be used; the alignment table below is designed for this.

Ethical considerations

• Informed consent from participants and the right to withdraw at any stage
• Confidentiality of information and anonymous questionnaires
• Approval from a research ethics committee (especially in medical and psychological research and any research on humans or animals)
• Honest use of others' work and correct citation of sources
• Transparent declaration of any use of AI tools, according to your university's rules

Timeline

List the main stages with approximate times. An example for a master's thesis:

Be realistic and leave enough time for permissions and access to the sample.

References

• List only sources cited in the proposal, and make sure every in-text citation appears in the reference list.
• Use one consistent citation style (usually APA, or whatever your faculty requires).
• Reference managers such as EndNote, Zotero or Mendeley make citing much easier and more accurate.
• Check every source before adding it: authors, year, title, journal and DOI must be correct and searchable.

The alignment table: the final test of your proposal

Before submitting, write one row for each specific objective in a table like the one below. If any cell stays empty or does not fit the rest, the proposal still has a problem.

Common proposal mistakes

• A topic too broad or too big for the time and resources of a thesis
• A problem statement that is only definitions and history and never shows the problem
• A mismatch between title, objectives, hypotheses and analysis
• An old, list-like literature review with no synthesis and no clear gap
• Vague methodology: sample size with no justification, unclear instruments, an unsuitable statistical test
• Incomplete or invented references, or citations missing from the reference list
• Copying text from other proposals or papers; plagiarism checkers will find it

Checklist before you submit

• The title is precise (and, if required, the translated title matches it).
• The problem statement is funnel-shaped and ends with the main question.
• The research gap and originality are supported by the literature.
• Each specific objective has a matching question or hypothesis and a clear analysis method.
• All variables have conceptual and operational definitions.
• Population, sample size, sampling method, instruments, validity and reliability are specified.
• Ethical considerations and, if needed, ethics approval are mentioned.
• All citations are in the reference list and all sources have been checked.
• There are no typos and the faculty's template is followed.
• Your supervisor has seen and approved the final version.

Using AI when writing a proposal

AI tools can help with tasks such as brainstorming search terms, checking grammar and wording, or summarising papers you have read yourself; but the proposal text, the argument and the choice of sources must be your own work. AI can produce invented sources or incorrect information.

Before using AI at all, read your university's rules on it and declare any use accordingly. Presenting work prepared by someone else — human or machine — as your own is academic misconduct.