What to Do When Your Thesis Topic Is Too Similar

The situation you are facing

You have found a thesis, research article, or earlier project with a title that is almost the same as yours. The model may use the same independent and dependent variables, target the same survey population, and even rely on familiar scale items. The question of what to do when your thesis topic is similar often appears after you have written the proposal or collected part of the data.

Start by checking the design rather than judging only from the title. Two studies may both examine purchase intention, satisfaction, or intention to use while differing in context, timing, location, sample group, mediator, moderator, or measurement approach. The opposite can also happen. Two titles may look slightly different while sharing almost the entire model and data-collection design.

Open a notes file and compare six components: the research problem, the variables and relationships between them, the target population, geographical scope, data-collection period, and analysis method. This separates a similar title from a similar research design. If you need to deal with a situation where the entire research direction has to change, see what to do when your supervisor asks you to change the model.

Common causes

A thesis topic is often too similar because the topic was chosen too broadly. Phrases such as “factors affecting customer satisfaction” or “factors affecting purchase intention” have been used extensively. If you only change the company name while keeping the same model, the committee may ask what the new study contributes.

Another cause is searching the literature too late. Students often settle on a topic based on an initial suggestion, design the questionnaire, and only then review theses from their university and recent research articles. When the similarity is discovered after data collection, changing the design becomes much more expensive.

Save the URL, year, author, sample, and conclusion for every source. A comparison table with an update date gives you stronger evidence than saying, “I searched Google but could not find anything.” The Scientific research topics page can be a starting point for reviewing related posts in the same topic group.

How to handle the problem in priority order

Do not immediately delete the questionnaire or discard all your data. Work through the least costly option first, and keep evidence for each decision. What to do if your thesis topic is taken depends on whether you are still choosing an idea, writing the proposal, or have already run SPSS.

Check the level of similarity with a comparison matrix

Create a table in which each column represents a study you have found and each row represents the problem, variables, sample, location, period, scales, and method. Mark each cell as “same,” “different,” or “unclear.” This forces you to identify the difference through specific components.

If only the title is similar but there is at least one meaningful difference in the context, target population, or research question, you may not need to change the entire direction. If nearly every row is the same, move to narrowing the scope or changing the model before collecting more data.

Narrow the target population and research context

You can narrow “e-commerce customers” to a specific user group, “Vietnamese consumers” to students in one location, or an entire industry to one type of service. The new location has value only when it creates a practical question that needs to be examined, such as different behavior, access levels, or user characteristics from the earlier study.

In the proposal, explain why this group needs to be studied separately. Avoid writing that the topic is new simply because “nobody has studied it at university X.” The survey location is part of the context. It does not automatically become an academic gap.

Adjust the model by adding a justified variable

If the old model has been used repeatedly, you can consider a mediator, moderator, or new contextual factor. Adding a variable must come from the research question and literature, not from the desire to put more boxes in the diagram.

For example, instead of testing only the direct factors affecting intention to use, you could ask whether perceived usefulness transmits an effect to intention or whether user experience changes the strength of that effect. You then need the corresponding hypotheses, scales, and analysis plan. If you choose PLS-SEM, prepare your explanation for choosing the tool in advance, because the committee often asks about it. You can also read what to do when asked why you chose PLS-SEM.

Change the research question instead of the entire topic

A broad topic can produce several different research questions. The earlier study may focus on the level of influence, while you focus on differences between groups, a mediating role, or predictive ability. Write the new question as one sentence with clear variables, population, and context.

If you already have data, change the question only when the questionnaire actually measures the concept you need to analyze. Do not take a variable that is absent from the questionnaire and declare it to be the new direction. With quantitative data, the questionnaire sets the practical boundary of the research question.

Confirm the decision with your supervisor before collecting or running the data

Send your supervisor one page containing the old topic, the study you suspect is similar, the comparison table, and your proposed revision. State clearly what stays the same, what changes, and how far the existing data remain suitable. This format keeps the discussion focused on a decision rather than on anxiety.

Illustrative decision table, illustrative output:

The table above is illustrative output, not the result of a real study. Replace each cell with information from the sources you found. If the plan changes the sample size or model, review the frequently asked questions about sample size in what to do when the committee asks about sample size.

When you need to collect more data

Consider collecting more data when the revised model contains a variable that the old questionnaire did not measure, the old target population does not fit the new question, or the number of responses in a comparison group is too small. Running the old file again will not solve a design problem.

You do not need to recollect data simply because you found a topic with a similar title. If the variables, context, and research questions are clearly differentiated, you can continue with the existing data after receiving approval. You also should not collect more responses merely to make the topic feel new. Sample size addresses the stability of estimates. It does not create a research gap.

If you discover the similarity after collecting 100% of the data, stop deleting or editing the original file. Save an unchanged copy, create a new working copy, and record the date of every change. Compare the actual questionnaire with the revised model. If the new variable was not measured, the honest options are to return to the design and collect additional data or keep the original question and report the limitation.

How to explain the issue to your supervisor

You can use the paragraph below and replace the content in brackets:

How to write this in your thesis: “Study [author, year] examined the relationship between [variables] in the context of [old context]. This study builds on that foundation but focuses on [new population or context] and additionally tests [new variable or relationship]. This adjustment aims to clarify [specific research problem] rather than reproduce the earlier model unchanged.”**

In a direct meeting, explain the situation in this order: which sources you found, what overlaps, what differs, why the difference matters, how far the existing data meet the requirements, and which option you propose. Do not begin by asking to keep the topic. Begin with the evidence.

If your supervisor asks you to change the model, ask three follow-up questions: which variables must be removed or added, whether the current questionnaire can still be used, and whether the hypotheses or sample scope must change. Record the answers immediately after the meeting so that you do not revise the model while leaving the old method section unchanged.

How to prevent the problem during design

Before finalizing the topic, search three groups of sources: international studies for each variable, Vietnamese theses or articles in the same context, and recent studies with similar models. You do not need to prove that the topic has never appeared anywhere in the world. You need to state what you are building on and which point remains insufficiently clear.

Write a paragraph called “expected research gap” before drawing the model. It should answer four questions: what earlier studies have established, what limitations remain, which component your topic changes, and who will benefit from the result. If you cannot answer the third question, the topic is still only a broad subject.

Name files by version, such as de_cuong_v1, de_cuong_v2_da_duyet, bang_hoi_pilot, and du_lieu_goc. Save the scale source table with the access date and each adjustment decision. If someone asks why you used or removed a variable, you can open the relevant evidence instead of relying on memory.

Finally, decide on the analysis tool in advance. SPSS works well for many regression analyses, tests, and EFA. SmartPLS is suitable when you are building a PLS-SEM model and need to assess both the measurement model and structural model. AMOS follows the CB-SEM approach, while JASP or R can be useful when you want a more flexible workflow but must control the code and reporting yourself. Changing software does not make the topic different. It changes how you test it.

Common mistakes when revising the topic

The first mistake is changing the title while keeping everything inside the study unchanged. Replacing “customers in city A” with “customers in city B” is not persuasive if you cannot explain the contextual or sample-based reason.

The second mistake is adding a variable based on intuition. A new variable needs a definition, scale, hypothesis, and corresponding data. If you add it after collecting the data, check the questionnaire before declaring a new model.

The third mistake is deleting the similar study from the literature review. You still need to mention it to show that you checked it and understand how your topic differs. Hiding the similarity creates more suspicion when the committee finds the article.

The fourth mistake is using trial-run results to prove that the topic is novel. A favorable p-value or high R² does not turn a repeated model into a new contribution. Novelty must be explained through the question, context, model, and data.

The final mistake is making changes after supervisor approval without updating the proposal, questionnaire, and method chapter. If the later results contradict the hypotheses, you still need to report them honestly, and you can refer to what to do when the results contradict the hypotheses.

Frequently asked questions

What should I do if my thesis topic is similar after I have submitted the proposal?

Keep the submitted proposal, mark the overlapping parts, and prepare a separate revision plan. Send your supervisor a comparison table showing the earlier study and your topic, then make changes only after receiving clear confirmation.

Does a similar thesis title mean that I must change the topic?

No. A similar title is a sign that you need to investigate further. The decision depends on how similar the research question, model, sample, context, period, and data are.

Can I still recover the project if I discover the similarity after collecting data?

Possibly, if the existing data measure the question correctly and you can create a well-supported difference. Keep the original file, do not add variables that were not measured, and ask your supervisor before changing the model or collecting additional data.

Does changing the survey location make the topic new?

Changing the location creates a difference only when that context has a specific research reason. If you change only the place name while keeping the model, questionnaire, and argument unchanged, the committee may still regard the study as repetitive.

Should I add many variables to avoid having a similar topic?

No. Every added variable increases the requirements for theoretical support, scales, sample size, and analysis. A concise model with a clear rationale is usually easier to defend than a model with many variables and weak reasoning.

How should I answer when the committee asks what is new about the topic?

Use this structure: what earlier studies did, where your topic differs, which problem creates that difference, and what your data allow you to test. Avoid saying only that “nobody has done it at my university” unless you can explain the value of that context.

Open your proposal file now, create the six-component comparison table, identify the study you suspect is similar, and send your proposed revision to your supervisor before changing the questionnaire or rerunning the data. If you need the full analysis workflow checked on your .sav or .csv file, see module M4 at DoThesis.