How do I write the results section of a research paper?

The first time I wrote a results section, I made it much harder than it needed to be. I had spent weeks collecting and analyzing data, so when I reached the results, I felt I had to explain everything I had discovered. I ended up mixing observations with interpretations, repeating information from my methods section, and starting to discuss why the findings mattered before I had properly shown what they were.

That experience changed how I approach the section. When I write one now, I think of the results as the place where I give the reader a clean account of what the research produced. The methods explain how I got there, and the discussion gives me room to interpret the meaning. Keeping those jobs separate makes the whole research paper structure easier to manage.

So, how do I write the results section of a research paper? I start by returning to my research questions and identifying the findings that directly answer them. I then present those findings in a logical order, use tables or figures when they communicate information more efficiently than prose, and describe patterns without turning the section into a discussion of their implications. That sounds simple, but deciding what belongs in the section requires more judgment than the basic formula suggests.

I start with the research question, not the data

When I have a large dataset in front of me, it is tempting to report everything that looks interesting. I have caught myself doing exactly that. A result can be statistically significant, visually striking, or simply unexpected and still be irrelevant to the question the study was designed to answer.

Before drafting, I go back to the research questions and hypotheses. I write down what each one was actually asking and then identify the results that speak directly to it. That gives me an organizing principle.

For example, imagine I conducted a study examining whether sleep duration is associated with test performance among university students. Suppose the analysis shows that students who reported longer sleep also had higher average test scores. I would report the relevant relationship and the statistical results supporting it. If I also discovered that students who preferred studying in the morning drank slightly more coffee, that might be interesting, but it does not automatically belong in the results section.

This filtering step is easy to skip because writing about every discovery feels thorough. In practice, it can make the section difficult to follow.

Purdue University’s guidance on research writing describes the results section as the place where researchers report their findings, with each result connected clearly to the corresponding method. Interpretation belongs primarily in the discussion.

That distinction gives me a useful test: if I remove a sentence, does the reader lose information about what the analysis actually found? If not, I question whether the sentence needs to be there.

I decide on the order before writing paragraphs

The order of results deserves more attention than I used to give it.

I generally follow the order established by my research questions or hypotheses. If my first question concerns the characteristics of the sample, I may establish those first. Then I move through the main analyses in an order that allows each finding to make sense in relation to the previous one.

This does not mean I mechanically reproduce the order in which I ran statistical tests. The computer might have produced twenty outputs in an order that made sense while I was analyzing the dataset. That does not mean the reader should encounter them in that same sequence.

I think of the results section as a guided tour through the evidence. The reader should understand why one finding appears before the next.

For a quantitative study, that might mean describing the sample before presenting the primary analysis, followed by secondary analyses that answer additional questions. For a qualitative project, the organization may instead follow themes, categories, or research questions. The exact convention varies by discipline, which is one reason I always check the requirements of the journal, department, or instructor rather than assuming that one universal format applies.

Purdue also notes that research-document conventions vary substantially by discipline, even though many scientific papers use some version of the Introduction, Methods, Results, and Discussion model.

I write the main finding before worrying about elegant prose

A results section is one place where I find plain language especially useful.

I first write a rough version of each finding without worrying about transitions or style. For example:

“The intervention group had a higher mean score than the control group.”

Then I add the information necessary for the reader to understand the result accurately, such as the relevant sample sizes, means, measures of variability, statistical test, effect size, confidence interval, or p-value when those are appropriate for the study.

The exact reporting requirements depend on the discipline and statistical method. I would never add statistical measures simply because I have seen them in other papers. They should correspond to the analysis I actually performed and to the reporting conventions of the field.

The rough sentence also helps me notice when I am trying to interpret the result too early.

Suppose the intervention group scored higher. Writing “This demonstrates that the intervention is an effective way to improve academic performance” goes beyond simply reporting the finding. The data might support that interpretation, but whether the evidence justifies such a conclusion depends on the study design, comparison group, effect size, limitations, and other factors.

I would save that reasoning for the discussion.

Tables and figures should reduce the reader’s workload

I used to think that including a table made a research paper look more sophisticated. Now I treat tables and figures much more practically.

If a table allows the reader to compare several related values quickly, I use one. If a finding can be understood more clearly from a graph, I consider a figure. But I do not create a visual simply because the assignment allows one.

The prose still needs to tell the reader what matters.

For instance, imagine a study comparing average stress scores across four academic years. A table could show the exact means and standard deviations for each group. The accompanying text might identify the main pattern and point the reader toward the table for the full values.

I would not rewrite every number from the table in the paragraph. That creates duplication without giving the reader additional understanding.

At the same time, I would not simply write “The results are shown in Table 1.” A useful results section tells the reader what to notice.

This is one of the places where editing becomes more valuable than adding words. I ask myself whether each sentence is doing something the table or figure cannot do as efficiently.

I keep interpretation under control

This is probably the boundary I have to watch most carefully.

Results can contain descriptive language. I can say that one group scored higher than another, that a particular variable increased over time, or that a relationship was statistically significant when the analysis supports that statement.

What I try to avoid is turning those observations into explanations before the discussion.

For example, if participants in one group reported lower stress, I can report that difference. Saying that the intervention reduced stress because participants felt more supported is a different claim. Unless I actually measured perceived support and the study design supports that explanation, I am moving beyond the evidence.

The same applies to causal language. If my study is observational, I need to be careful about writing as though one variable caused another. The results may show an association without establishing causation.

That distinction may sound obvious when stated explicitly, but it becomes easy to blur while drafting because I already know what I think the findings mean.

One trick that helps me is to imagine that another researcher is reading the section without knowing my interpretation. I ask whether the reader can see exactly what I observed and how I calculated it without being pushed toward a conclusion that the evidence has not established.

Unexpected findings deserve space, but not drama

Research rarely produces a perfectly tidy set of results.

Sometimes a hypothesis is not supported. Sometimes an expected relationship is weaker than anticipated. Sometimes a secondary analysis produces a result that I did not predict.

I do not think an unexpected result should be hidden simply because it makes the paper less neat. If it is relevant to the research question and supported by the analysis, it belongs in the report.

The important part is describing it accurately.

For example, if I expected two groups to differ significantly but the analysis did not show a statistically significant difference, I would report that outcome rather than trying to rescue the original hypothesis through selective wording.

I also would not immediately explain why the expected difference failed to appear. That interpretation can be developed later, where I can consider alternative explanations and limitations.

A result that contradicts my expectation is still a result.

I check that every result can be traced back to the method

Before I consider the section finished, I perform a kind of reverse audit.

I look at each major result and ask: where did this number, comparison, category, or observation come from?

The answer should lead back to something described in the methods.

This matters because a results section can accidentally introduce analyses that were never explained earlier. Perhaps I ran an additional test after noticing an unexpected pattern. That analysis may be legitimate, but if I intend to include it, I need to make sure the methodology and reporting are consistent with it.

Purdue’s research-writing guidance makes a similar connection between methods and results: readers should be able to understand how the reported findings correspond to the procedures used to obtain them.

This check has caught mistakes in my own drafts, especially when I have revised the analysis after writing an early version of the methods.

My final editing pass is surprisingly mechanical

Once the content is correct, I stop treating the results section as a piece of creative writing.

I check that terminology stays consistent. If I call something the “treatment group” in one paragraph and the “intervention group” in another, I make sure I am not accidentally referring to different populations. I verify numbers against my statistical output. I check table and figure references. I make sure reported percentages add up where they should and that rounding has not introduced an obvious inconsistency.

I also read the section while deliberately ignoring what I think the findings mean.

That last step is useful because interpretation has a tendency to creep back in during revision. A sentence that seemed harmless in the first draft can quietly become a claim about why the result occurred.

If I need help identifying those problems, I can also use writing resources or editing tools during revision. EssayPay, for example, provides academic writing guides and an online essay checker that can flag issues involving readability, grammar, structure, and other aspects of a draft. I would use something like that as a second set of eyes, though I would still verify statistical reporting and scientific claims against my own data and the requirements of the discipline.

The question I use before submitting

The most useful test I have found is simple: could another researcher read my results section and understand what my study found without needing me to explain my interpretation aloud?

If the answer is yes, the section is probably doing its job.

The reader should be able to follow the findings in a deliberate order, see the evidence supporting the research questions, understand the important numbers or themes, and distinguish reported observations from later interpretation. The section does not have to sound dramatic. In fact, I usually take that as a good sign. Clear reporting lets the evidence carry the weight.

When I finish a results section now, I resist the urge to make it sound more impressive by adding conclusions that belong somewhere else. I concentrate on being exact about what the research produced. That discipline makes the discussion easier to write afterward because I have already established a clean record of the evidence. I know what I found, I know where each finding came from, and I can then spend the next section asking the more interesting question: what do those findings actually mean?

Explore More Guides

AI Writers Are Getting Smarter – But Can They Replace Human Creativity?
How to choose the best admission essay writing service | Pulse Kenya
What Do We Know About Tourism Education?
What professors really think about essay writing services - AZ Big Media

![img](https://plus.unsplash.com/premium_photo-1685136479835-1dfa5dbc5d44?q=80&w=1170&auto=format&fit=crop&ixlib=rb-4.1.0&ixid=M3wxMjA3fDB8MHxwaG90by1wYWdlfHx8fGVufDB8fHx8fA%3D%3D) The first time I wrote a results section, I made it much harder than it needed to be. I had spent weeks collecting and analyzing data, so when I reached the results, I felt I had to explain everything I had discovered. I ended up mixing observations with interpretations, repeating information from my methods section, and starting to discuss why the findings mattered before I had properly shown what they were. That experience changed how I approach the section. When I write one now, I think of the results as the place where I give the reader a clean account of what the research produced. The methods explain how I got there, and the discussion gives me room to interpret the meaning. Keeping those jobs separate makes the whole [research paper structure](https://essaypay.com/research-paper-writing-service/) easier to manage. So, how do I write the results section of a research paper? I start by returning to my research questions and identifying the findings that directly answer them. I then present those findings in a logical order, use tables or figures when they communicate information more efficiently than prose, and describe patterns without turning the section into a discussion of their implications. That sounds simple, but deciding what belongs in the section requires more judgment than the basic formula suggests. ## I start with the research question, not the data When I have a large dataset in front of me, it is tempting to report everything that looks interesting. I have caught myself doing exactly that. A result can be statistically significant, visually striking, or simply unexpected and still be irrelevant to the question the study was designed to answer. Before drafting, I go back to the research questions and hypotheses. I write down what each one was actually asking and then identify the results that speak directly to it. That gives me an organizing principle. For example, imagine I conducted a study examining whether sleep duration is associated with test performance among university students. Suppose the analysis shows that students who reported longer sleep also had higher average test scores. I would report the relevant relationship and the statistical results supporting it. If I also discovered that students who preferred studying in the morning drank slightly more coffee, that might be interesting, but it does not automatically belong in the results section. This filtering step is easy to skip because writing about every discovery feels thorough. In practice, it can make the section difficult to follow. Purdue University's guidance on research writing describes the results section as the place where researchers report their findings, with each result connected clearly to the corresponding method. Interpretation belongs primarily in the discussion. That distinction gives me a useful test: if I remove a sentence, does the reader lose information about what the analysis actually found? If not, I question whether the sentence needs to be there. ## I decide on the order before writing paragraphs The order of results deserves more attention than I used to give it. I generally follow the order established by my research questions or hypotheses. If my first question concerns the characteristics of the sample, I may establish those first. Then I move through the main analyses in an order that allows each finding to make sense in relation to the previous one. This does not mean I mechanically reproduce the order in which I ran statistical tests. The computer might have produced twenty outputs in an order that made sense while I was analyzing the dataset. That does not mean the reader should encounter them in that same sequence. I think of the results section as a guided tour through the evidence. The reader should understand why one finding appears before the next. For a quantitative study, that might mean describing the sample before presenting the primary analysis, followed by secondary analyses that answer additional questions. For a qualitative project, the organization may instead follow themes, categories, or research questions. The exact convention varies by discipline, which is one reason I always check the requirements of the journal, department, or instructor rather than assuming that one universal format applies. Purdue also notes that research-document conventions vary substantially by discipline, even though many scientific papers use some version of the Introduction, Methods, Results, and Discussion model. ## I write the main finding before worrying about elegant prose A results section is one place where I find plain language especially useful. I first write a rough version of each finding without worrying about transitions or style. For example: “The intervention group had a higher mean score than the control group.” Then I add the information necessary for the reader to understand the result accurately, such as the relevant sample sizes, means, measures of variability, statistical test, effect size, confidence interval, or p-value when those are appropriate for the study. The exact reporting requirements depend on the discipline and statistical method. I would never add statistical measures simply because I have seen them in other papers. They should correspond to the analysis I actually performed and to the reporting conventions of the field. The rough sentence also helps me notice when I am trying to interpret the result too early. Suppose the intervention group scored higher. Writing “This demonstrates that the intervention is an effective way to improve academic performance” goes beyond simply reporting the finding. The data might support that interpretation, but whether the evidence justifies such a conclusion depends on the study design, comparison group, effect size, limitations, and other factors. I would save that reasoning for the discussion. ## Tables and figures should reduce the reader's workload I used to think that including a table made a research paper look more sophisticated. Now I treat tables and figures much more practically. If a table allows the reader to compare several related values quickly, I use one. If a finding can be understood more clearly from a graph, I consider a figure. But I do not create a visual simply because the assignment allows one. The prose still needs to tell the reader what matters. For instance, imagine a study comparing average stress scores across four academic years. A table could show the exact means and standard deviations for each group. The accompanying text might identify the main pattern and point the reader toward the table for the full values. I would not rewrite every number from the table in the paragraph. That creates duplication without giving the reader additional understanding. At the same time, I would not simply write “The results are shown in Table 1.” A useful results section tells the reader what to notice. This is one of the places where editing becomes more valuable than adding words. I ask myself whether each sentence is doing something the table or figure cannot do as efficiently. ## I keep interpretation under control This is probably the boundary I have to watch most carefully. Results can contain descriptive language. I can say that one group scored higher than another, that a particular variable increased over time, or that a relationship was statistically significant when the analysis supports that statement. What I try to avoid is turning those observations into explanations before the discussion. For example, if participants in one group reported lower stress, I can report that difference. Saying that the intervention reduced stress because participants felt more supported is a different claim. Unless I actually measured perceived support and the study design supports that explanation, I am moving beyond the evidence. The same applies to causal language. If my study is observational, I need to be careful about writing as though one variable caused another. The results may show an association without establishing causation. That distinction may sound obvious when stated explicitly, but it becomes easy to blur while drafting because I already know what I think the findings mean. One trick that helps me is to imagine that another researcher is reading the section without knowing my interpretation. I ask whether the reader can see exactly what I observed and how I calculated it without being pushed toward a conclusion that the evidence has not established. ## Unexpected findings deserve space, but not drama Research rarely produces a perfectly tidy set of results. Sometimes a hypothesis is not supported. Sometimes an expected relationship is weaker than anticipated. Sometimes a secondary analysis produces a result that I did not predict. I do not think an unexpected result should be hidden simply because it makes the paper less neat. If it is relevant to the research question and supported by the analysis, it belongs in the report. The important part is describing it accurately. For example, if I expected two groups to differ significantly but the analysis did not show a statistically significant difference, I would report that outcome rather than trying to rescue the original hypothesis through selective wording. I also would not immediately explain why the expected difference failed to appear. That interpretation can be developed later, where I can consider alternative explanations and limitations. A result that contradicts my expectation is still a result. ## I check that every result can be traced back to the method Before I consider the section finished, I perform a kind of reverse audit. I look at each major result and ask: where did this number, comparison, category, or observation come from? The answer should lead back to something described in the methods. This matters because a results section can accidentally introduce analyses that were never explained earlier. Perhaps I ran an additional test after noticing an unexpected pattern. That analysis may be legitimate, but if I intend to include it, I need to make sure the methodology and reporting are consistent with it. Purdue's research-writing guidance makes a similar connection between methods and results: readers should be able to understand how the reported findings correspond to the procedures used to obtain them. This check has caught mistakes in my own drafts, especially when I have revised the analysis after writing an early version of the methods. ## My final editing pass is surprisingly mechanical Once the content is correct, I stop treating the results section as a piece of creative writing. I check that terminology stays consistent. If I call something the “treatment group” in one paragraph and the “intervention group” in another, I make sure I am not accidentally referring to different populations. I verify numbers against my statistical output. I check table and figure references. I make sure reported percentages add up where they should and that rounding has not introduced an obvious inconsistency. I also read the section while deliberately ignoring what I think the findings mean. That last step is useful because interpretation has a tendency to creep back in during revision. A sentence that seemed harmless in the first draft can quietly become a claim about why the result occurred. If I need help identifying those problems, I can also use writing resources or editing tools during revision. EssayPay, for example, provides academic writing guides and an online essay checker that can flag issues involving readability, grammar, structure, and other aspects of a draft. I would use something like that as a second set of eyes, though I would still verify statistical reporting and scientific claims against my own data and the requirements of the discipline. ## The question I use before submitting The most useful test I have found is simple: could another researcher read my results section and understand what my study found without needing me to explain my interpretation aloud? If the answer is yes, the section is probably doing its job. The reader should be able to follow the findings in a deliberate order, see the evidence supporting the research questions, understand the important numbers or themes, and distinguish reported observations from later interpretation. The section does not have to sound dramatic. In fact, I usually take that as a good sign. Clear reporting lets the evidence carry the weight. When I finish a results section now, I resist the urge to make it sound more impressive by adding conclusions that belong somewhere else. I concentrate on being exact about what the research produced. That discipline makes the discussion easier to write afterward because I have already established a clean record of the evidence. I know what I found, I know where each finding came from, and I can then spend the next section asking the more interesting question: what do those findings actually mean? ## Explore More Guides [AI Writers Are Getting Smarter – But Can They Replace Human Creativity?](https://www.saasgenius.com/blog/ai-writers-are-getting-smarter/) [How to choose the best admission essay writing service | Pulse Kenya](https://www.pulse.co.ke/story/how-to-choose-the-best-admission-essay-writing-service-2025120913151941935) [What Do We Know About Tourism Education?](https://elsalvadorinfo.net/what-do-we-know-about-tourism-education/) [What professors really think about essay writing services - AZ Big Media](https://azbigmedia.com/business/education-news/what-professors-really-think-about-essay-writing-services/)

#7 How do I write the results section of a research paper?