How to Relate Your Results to Your Hypothesis in the Discussion

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-27

Quick Answer

• State plainly whether your data support or reject your hypothesis, then explain the specific evidence that led to that judgment.
• Use a structured decision process that separates statistical outcomes, effect sizes, and biological or practical relevance before you write your conclusion.
• A rejected hypothesis is a valid scientific outcome when your methods were sound and your interpretation is honest about limitations.

At a Glance

Understanding the Purpose of the Discussion Section

The discussion section is where you interpret what your results mean in the context of your original research question. Your hypothesis is the prediction you made before collecting data, and the discussion is where you explicitly state whether your observations align with that prediction. This is not a place for vague language or hedging. Readers expect a direct statement about whether your hypothesis was supported, rejected, or left unresolved.

Your job in the discussion is to connect three elements: the hypothesis you stated in your introduction, the results you reported in your results section, and the broader scientific context that helps explain why you observed what you did. When these three elements are clearly linked, your reader can follow your reasoning from question to answer. When they are disconnected, even strong data can appear confusing or unconvincing.

The discussion also serves an honesty function. Science advances when researchers report what they actually found, including results that contradict their predictions. A well written discussion acknowledges when results are ambiguous, when methods limited interpretation, and when alternative explanations remain plausible. This transparency is a core expectation in ethical research reporting, as outlined in the Committee on Publication Ethics Core Practices, which emphasize honest and complete reporting of research findings.

Distinguishing Hypothesis, Prediction, and Result

Before you can relate your results to your hypothesis, you need to be clear about what each term means. A hypothesis is a broad explanatory statement about how something works. A prediction is a specific, testable outcome that follows logically from the hypothesis. A result is what you actually observed after running your experiment or analysis.

For example, a hypothesis might state that increased soil nitrogen affects plant growth. The prediction would specify that plants receiving supplemental nitrogen will produce more biomass than control plants. The result would be the measured biomass difference between the two groups. When you write your discussion, you are comparing the prediction to the result and then deciding what that comparison says about the hypothesis.

This distinction matters because students often confuse them. They might write that their results support their hypothesis when they actually only support a specific prediction. Or they might reject an entire hypothesis because one prediction failed, even when other predictions held. Being precise about the level at which you are making your claim strengthens your discussion and makes your reasoning easier for readers to follow.

The Decision Tree for Interpreting Hypothesis Outcomes

A structured approach helps you avoid the common trap of deciding what you want to conclude and then searching for evidence to support it. Instead, work through your results systematically and let the evidence guide your conclusion.

Step 1: Check Whether Your Data Are Complete and Reliable

Before interpreting anything, confirm that your dataset is complete and that your measurements are trustworthy. Review your lab notebook or data files for missing values, recording errors, or anomalies that could distort your analysis. If you find problems, address them before drawing conclusions. A conclusion based on faulty data is worse than no conclusion at all.

Consider whether your sample size was adequate for the type of comparison you are making. Small samples produce variable estimates, and that variability may explain why you did or did not see a difference. Your discussion should acknowledge the role of sample size in shaping your conclusions.

Step 2: Determine Whether Your Results Match Your Predictions

Compare each prediction to the corresponding result. For each one, ask whether the observed outcome matches what you expected. Record this comparison for every prediction, beyond the ones that worked out. This prevents you from selectively reporting only favorable outcomes, which is a form of reporting bias that undermines scientific integrity.

Step 3: Assess the Magnitude and Consistency of Effects

Statistical significance tells you whether an effect is likely real, but it does not tell you whether the effect matters. Consider the size of the effect you observed and whether it is large enough to be practically or biologically meaningful. Also consider consistency. Did the effect appear across multiple measurements, time points, or replicates, or did it depend on a single observation?

Step 4: Consider Alternative Explanations

Before concluding that your hypothesis is supported or rejected, ask whether something other than your hypothesis could explain your results. Confounding variables, measurement error, procedural differences between groups, and chance can all produce patterns that look meaningful. Your discussion should address the most plausible alternative explanations and explain why you think your interpretation is the best fit for the data.

Step 5: State Your Conclusion Explicitly

Once you have worked through the first four steps, write a direct statement about your hypothesis. Use clear language such as "the data support the hypothesis" or "the results do not support the hypothesis." Avoid phrases like "the data seem to suggest" or "it appears that," which weaken your conclusion and leave readers uncertain about what you actually found.

flowchart TD
    A[Review data completeness and reliability] --> B{Do results match predictions?}
    B -->|Yes for all predictions| C[Assess effect size and consistency]
    B -->|Yes for some predictions| D[Identify which predictions held and which did not]
    B -->|No for any prediction| E[Check for procedural errors or confounders]
    C --> F{Are effects meaningful and consistent?}
    F -->|Yes| G[State hypothesis is supported]
    F -->|No| H[State hypothesis is not clearly supported]
    D --> I[State partial support with specific boundaries]
    E --> J{Are alternative explanations plausible?}
    J -->|Yes| K[State hypothesis is not supported or inconclusive]
    J -->|No| L[State hypothesis is not supported]

Writing the Direct Statement of Support or Rejection

The most important sentence in your discussion is the one where you state whether your hypothesis was supported. This sentence should appear early in the discussion, usually in the first paragraph, so readers immediately know where you stand. It should be specific enough that someone could quote it and understand your conclusion without reading the rest of the section.

A supported hypothesis statement might read: "The results support the hypothesis that light intensity affects photosynthetic rate, as plants exposed to higher light levels produced significantly more oxygen than plants in low light conditions." This sentence names the hypothesis, states the outcome, and points to the specific evidence that supports the claim.

A rejected hypothesis statement might read: "The data do not support the hypothesis that temperature influences germination rate, because germination percentages were similar across all temperature treatments." Again, the sentence is direct, names the hypothesis, and references the specific evidence.

Avoid the word "prove" in either direction. Science does not prove hypotheses in a single experiment. It accumulates evidence that supports or fails to support explanations. Using "prove" overstates what your data can show and invites criticism from reviewers or instructors who expect appropriately cautious language.

Handling Partial Support

Many experiments produce results that support some predictions but not others. This is a normal outcome and does not mean your experiment failed. It means your hypothesis was partially correct, or that your hypothesis was correct in some conditions but not others.

When you have partial support, say so explicitly. Identify which predictions were supported and which were not, and offer an explanation for the pattern. For example, you might write: "The hypothesis that fertilizer increases yield was supported for nitrogen treatments but not for phosphorus treatments, suggesting that nitrogen was the limiting nutrient in this soil system."

Partial support often points to a more interesting scientific story than full support. It suggests that your hypothesis needs refinement, and your discussion is the place to propose how. You might suggest that the hypothesis applies under certain conditions, or that additional variables need to be incorporated into the model.

Explaining Why a Hypothesis Was Not Supported

When your results reject your hypothesis, you have several legitimate explanations to consider. The hypothesis itself may be wrong. Your methods may have been inadequate to test the hypothesis properly. Your sample may have been too small or too variable to detect an effect. Or the experimental conditions may not have been appropriate for the phenomenon you were studying.

Your discussion should address these possibilities honestly. Start by stating that the hypothesis was not supported. Then walk through the most plausible explanations, using your observations and knowledge of the system to evaluate each one. You do not need to resolve every uncertainty, but you should show that you have thought carefully about why you got the result you did.

A common student error is to assume that a rejected hypothesis means the experiment was a failure. This is incorrect. A well designed experiment that produces a clear negative result is scientifically valuable because it rules out an explanation and helps narrow the range of plausible hypotheses. The National Library of Medicine Research Methods Resources provide extensive guidance on how to think about hypothesis testing and the interpretation of negative results in biomedical research contexts.

The Role of Statistical Significance in Your Conclusion

Statistical significance is a tool for deciding whether an observed effect is likely real or likely due to chance. It is not the same as scientific importance, and it should not be the only factor in your conclusion. A statistically significant result with a tiny effect size may have no practical relevance. A non significant result with a large effect size may indicate that your sample was too small to detect a real difference.

When you write your discussion, report what your statistical tests showed, but interpret them in context. State the p value or confidence interval if you used inferential statistics, then explain what that result means for your hypothesis. If your test was not significant, say so directly and discuss whether the result reflects a true absence of effect or insufficient statistical power.

Your discussion should also acknowledge the limitations of your statistical approach. Different tests make different assumptions about your data, and violations of those assumptions can affect your conclusions. If you are unsure whether your data met the assumptions of your chosen test, say so and explain how that uncertainty affects your interpretation.

Connecting Results to the Broader Scientific Context

Your discussion should place your findings in the context of existing knowledge. This does not mean writing a full literature review. It means identifying the key studies or concepts that relate to your hypothesis and explaining how your results fit with what is already known.

If your results support your hypothesis, explain how they extend or confirm previous work. If they contradict previous work, consider why the discrepancy exists. Differences in methods, study populations, or experimental conditions can all explain why two studies reach different conclusions. Your discussion should address these possibilities instead of simply asserting that your results are correct and the other study is wrong.

The EQUATOR Network provides reporting guidelines that help researchers describe their methods and results transparently. Following these guidelines improves the quality of your reporting and makes it easier for readers to evaluate how your results relate to your hypothesis and to the broader literature.

Addressing Limitations Honestly

Every study has limitations, and your discussion should acknowledge the ones that affect your ability to draw conclusions. Common limitations include small sample sizes, short study durations, limited geographic scope, measurement error, and the use of simplified experimental systems that do not fully represent real world conditions.

For each limitation you identify, explain how it affects your conclusions. A limitation that weakens your confidence in a supported hypothesis should be stated clearly. A limitation that explains why you could not detect an effect should also be stated clearly. The goal is to give readers an accurate sense of how much weight to place on your conclusions.

Avoid the temptation to list every possible limitation without explaining its relevance. A long list of generic limitations does not strengthen your discussion. Instead, focus on the limitations that genuinely matter for interpreting your specific results.

Common Failure Patterns in Student Discussions

Students tend to repeat certain patterns when writing discussions. Recognizing these patterns can help you avoid them.

Overstating the Conclusion

Some students write that their hypothesis was "proven" or "confirmed" when their data only provide support. This overstatement invites criticism and misrepresents the strength of the evidence. Use language that matches what your data can actually show.

Understating the Conclusion

The opposite problem is hedging so heavily that the reader cannot tell what you concluded. Phrases like "it seems possible that" or "the data might suggest" obscure your findings. State your conclusion directly, then explain the reasoning that supports it.

Ignoring Results That Do Not Fit

Students sometimes focus only on results that support their hypothesis and ignore or downplay results that contradict it. This selective reporting is a form of bias that undermines the credibility of your work. Address all of your results, including the ones that complicate your story.

Confusing Correlation with Causation

If your study was observational, your results can show associations but cannot establish causation. Your discussion should respect this distinction. Do not claim that your results prove a causal relationship when your design does not support that claim.

Writing the Discussion as a Results Summary

Some students simply restate their results in paragraph form instead of interpreting them. The discussion should explain what the results mean, not repeat them. If you find yourself describing your findings again, stop and ask what those findings imply for your hypothesis.

Records and Measurements That Support Your Conclusion

Your ability to relate results to your hypothesis depends on the quality of your records. Keep detailed notes about your experimental design, your procedures, and any deviations from your plan. Record your raw data before performing any calculations or transformations. Document the statistical tests you used and the assumptions you checked.

Good record keeping also means documenting decisions you made during the experiment. If you excluded data points, record why. If you changed your procedure partway through, record when and why. These details matter because they affect how readers interpret your results and because they demonstrate that your work meets ethical standards for research conduct.

The NIH Data Management and Sharing Policy describes expectations for managing and sharing research data. While this policy applies to NIH funded research, the underlying principles of careful data management and transparent sharing are relevant to any research project.

Practical Steps for Writing Your Discussion

Step 1: Draft Your Hypothesis Statement

Write your hypothesis in one sentence at the top of your discussion draft. This keeps you focused on what you are actually testing.

Step 2: List Your Predictions and Results Side by Side

Create a table or list that pairs each prediction with the corresponding result. This makes it easy to see which predictions were supported and which were not.

Step 3: Write Your Direct Conclusion First

Draft the sentence that states whether your hypothesis was supported, partially supported, or not supported. Write this before you write anything else in the discussion.

Step 4: Explain Your Reasoning

For each prediction, explain why the result supports or does not support it. Reference your statistical tests, effect sizes, and any observations that shaped your interpretation.

Step 5: Address Alternative Explanations

Identify the most plausible alternative explanations for your results and explain why you favor your interpretation.

Step 6: Connect to the Broader Context

Explain how your findings relate to previous work and what they add to the understanding of your research question.

Step 7: State Your Limitations

Identify the limitations that affect your conclusions and explain their impact.

Step 8: Revise for Clarity and Directness

Read your draft and check that every paragraph serves a clear purpose. Remove hedging language, clarify vague statements, and make sure your conclusion is stated explicitly.

Using Reporting Guidelines to Strengthen Your Discussion

Reporting guidelines are checklists that help researchers write complete and transparent reports of their studies. The EQUATOR Network maintains a comprehensive library of these guidelines for different study types. While these guidelines are primarily designed for journal articles, the principles they embody are useful for any research report, including lab reports and theses.

Using a reporting guideline helps you remember to include important elements that are easy to overlook, such as describing how sample size was determined, how participants or samples were allocated to groups, and how missing data were handled. Including these details strengthens your discussion because it gives readers the information they need to evaluate your conclusions.

Ethical Considerations in Reporting Results

Honest reporting is a fundamental expectation in research. The Committee on Publication Ethics Core Practices describe the responsibilities of researchers to report their work accurately and completely. These responsibilities include not fabricating data, not falsifying results, and not selectively reporting outcomes that support a preferred conclusion.

When you write your discussion, you are making a public claim about what your data show. That claim should be accurate, complete, and appropriately cautious. If your results are ambiguous, say so. If your methods had weaknesses, acknowledge them. If your conclusions depend on assumptions, state those assumptions.

These ethical obligations apply to student lab reports as well as to published research. Developing honest reporting habits early in your training will serve you throughout your career.

The Role of Peer Review and Feedback

Before finalizing your discussion, seek feedback from others. A classmate, teaching assistant, or instructor can read your draft and tell you whether your conclusion is clear and whether your reasoning is convincing. They can also catch places where you have overclaimed or underclaimed.

Peer review serves a similar function in professional research. The Committee on Publication Ethics Core Practices describe the expectations for peer review, including the importance of fair and constructive evaluation. When you receive feedback on your writing, use it to strengthen your discussion instead of treating it as a personal criticism.

When to Seek Professional Guidance

If you are writing a discussion for a thesis, dissertation, or journal submission, you may benefit from guidance beyond what this article provides. Your advisor, committee members, or institutional writing center can offer feedback specific to your field and your project.

If you are preparing a manuscript for publication, consult the reporting guidelines relevant to your study type through the EQUATOR Network. These guidelines will help you structure your report and ensure you have included all necessary elements.

If you are conducting research funded by the NIH, review the NIH Grants and Funding resources for information about application requirements and award management. The NIH Data Management and Sharing Policy describes expectations for data management plans and data sharing that may affect how you report your results.

If you are establishing your identity as a researcher, consider creating an ORCID record. An ORCID iD provides a persistent identifier that links your research outputs to you, which is useful when you begin publishing and sharing your work.

Frequently Asked Questions

How do I know if my hypothesis is supported or rejected?

Compare each prediction to the corresponding result. If the results match your predictions and the effects are consistent and meaningful, your hypothesis is supported. If the results contradict your predictions, your hypothesis is not supported. If the results are mixed, your hypothesis is partially supported.

What should I write first in my discussion section?

Write the direct statement of whether your hypothesis was supported, partially supported, or not supported. Put this statement early in the discussion so readers immediately know your conclusion. Then explain the reasoning that led to that conclusion.

Can I say my hypothesis was proven?

No. A single experiment does not prove a hypothesis. It provides evidence that supports or fails to support the hypothesis. Use language like "the data support" or "the results are consistent with" instead of "prove."

What if my results do not support my hypothesis?

State clearly that the hypothesis was not supported, then explain why. Consider whether the hypothesis was wrong, whether your methods were adequate, whether your sample was sufficient, and whether alternative explanations exist. A rejected hypothesis is a valid outcome when your reasoning is sound.

How do I handle results that support some predictions but not others?

Describe the pattern of partial support explicitly. Identify which predictions were supported and which were not, then offer an explanation for the difference. Partial support often suggests that your hypothesis needs refinement.

Should I discuss limitations in my discussion section?

Yes. Identify the limitations that affect your ability to draw conclusions and explain how they influence your interpretation. Focus on limitations that genuinely matter for your specific study instead of listing generic weaknesses.

How much detail should I include about my statistical analysis?

Report what your statistical tests showed, including test statistics and p values if you used inferential statistics. Then interpret those results in the context of your hypothesis. Explain what the statistical results mean instead of simply reporting numbers.

What is the difference between a hypothesis and a prediction?

A hypothesis is a broad explanatory statement about how something works. A prediction is a specific, testable outcome that follows from the hypothesis. Your results are compared to your predictions, and that comparison tells you whether the hypothesis is supported.

Using the Evidence

Related Articles

• Western Blot Results
• Rna World Hypothesis
• Veterinary Terminology: A Guide for Students and Professionals
• How to Write Expected Results in a Research Proposal: A Practical Guide
• Careers in Agriculture: A Decision Framework for Choosing Your Path

References and Further Reading

• Research Methods Resources. National Library of Medicine.
• EQUATOR Network. EQUATOR Network.
• Core Practices. Committee on Publication Ethics.
• NIH Grants and Funding. National Institutes of Health.
• ORCID for Researchers. ORCID.
• Data Management and Sharing Policy. National Institutes of Health.
• NCBI Literature Resources. National Center for Biotechnology Information.
• PubMed. National Library of Medicine.

This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.