Reviewed October 3, 2026. A strong thesis defense explains the problem, shows how your method addresses it, and separates evidence from claims. Your institution’s research manual and adviser determine the actual procedure.
Confirm the rules before preparing the slides
Ask your department for its current requirements: endorsement, manuscript submission, presentation time, panel procedure, permitted notes, demonstration arrangements, revision deadlines, and final approval. These vary by institution and program.
For examples of official institutional guidance, see City College of Calamba’s research toolkit and LPU Batangas’s thesis and dissertation manual. They illustrate why another university’s panel size or submission schedule should not be treated as your own rule.
Build your presentation around five questions
- What problem did you investigate? Explain the specific setting and the gap, rather than beginning with a long generic history.
- What did you aim to establish? Connect each objective to a method and a result.
- How did you investigate it? Describe the design, data, tools, and evaluation choices.
- What did the evidence show? Present the important results with units, sample information, and a clear comparison.
- What remains uncertain? State limitations and the next work needed.
Choose your slide count around the time your department allows. Rehearse aloud and leave room for transitions. Put supporting tables and details in backup slides so you can retrieve them without crowding the main story.
A worked example: defend a prototype’s result
Suppose your team built a room-temperature logger and compared it against a reference instrument in one room. A claim such as “our system is perfectly accurate” is too broad.
A more defensible answer is: “We compared the logger with the reference under the conditions shown here. The observed differences are in this table. Our tests cover this room and temperature range; they do not establish performance outdoors or across the full sensor range.” Use your actual data in place of the example.
If a panelist asks why you selected that method, connect the choice to the objective and available resources. Then acknowledge alternatives and explain what a stronger follow-up study would require.
Answer questions with claim, evidence, and limit
- Claim: answer the question directly in one sentence.
- Evidence: point to the relevant result, method, or source.
- Limit: state what the evidence cannot establish.
For example: “We used repeated measurements to examine repeatability. The spread appears in Table 4. That checks consistency under our setup; it does not independently establish accuracy.” This structure is more useful than repeating a memorized paragraph.
Run a mock defense that finds weaknesses
Ask a classmate to interrupt with questions about your assumptions, sample, measurement error, comparison method, and conclusions. Keep an issue log with four columns: question, current answer, supporting evidence, and action needed. A missing source or test is an action item, not something to hide with a more confident sentence.
If you do not know an answer, say so precisely: “We did not evaluate that condition; it is outside these tests.” Distinguish a limitation from an error that needs correction. Ask for clarification when you are unsure what the panelist means.
Prepare the demonstration and a fallback
- Test the presentation on the actual equipment when possible.
- Keep a PDF copy and locally available copies of essential figures.
- For a prototype, rehearse startup, normal operation, and a likely failure case.
- Prepare an accurately labelled recording or screenshots if your program permits them; explain when they were captured.
- Confirm who presents each part and who answers questions about each subsystem.
Turn feedback into a revision record
After the defense, organize comments by issue rather than by the order they were spoken. Record the requested change, the section or page affected, the revision made, and its status. Resolve conflicting instructions with your adviser or the designated academic authority. Complete the required approvals before treating the manuscript as final.
Next reading
If your thesis involves measurements, revisit Numerical Differentiation to understand how calculations can amplify noise. For a career portfolio after graduation, read the CpECB preparation guide and describe the work you can support with evidence.