Graduate Guide

Dissertation Data Analysis: From Output to Results Chapter

How to turn statistical output, coded transcripts or mixed data into a results chapter that answers each research question cleanly and leaves interpretation for the discussion.

Updated October 2026 · 7 min read

Dissertation data analysis means running the analysis your method chapter promised and reporting what it found, question by question. The results chapter, usually chapter four, reports findings; the discussion chapter explains them.

This guide covers how to plan the analysis around your research questions, how to report quantitative and qualitative findings, how to integrate mixed methods and the errors committees catch most often.

The Results Chapter in One Rule

Report what you found, organized by research question or hypothesis, in a form a reader could check against your tables. Leave the "why" and the links to the literature for the discussion.

For statistics, give the test, the statistic, degrees of freedom, the exact p value and an effect size. For qualitative work, give each theme a definition and evidence from several participants.

Start from the Analysis Plan in Your Method Chapter

Your committee approved a plan at the proposal stage. The results chapter should follow it, and any change needs a reason stated openly.

Research question typeTypical analysisWhat the results section reports
Difference between two groupsIndependent t test, or Mann-Whitney U if assumptions failGroup means or medians, test statistic, p, effect size
Difference across three or more groupsOne-way ANOVA with post hoc tests, or Kruskal-WallisF or H statistic, p, effect size, which pairs differ
Change over time in one groupPaired t test or repeated-measures ANOVAChange scores, statistic, p, effect size
Relationship between variablesPearson or Spearman correlationCoefficient, p, confidence interval
Prediction of an outcomeMultiple linear or logistic regressionModel fit, coefficients or odds ratios, confidence intervals
Lived experience or meaningThematic, phenomenological or other qualitative analysisThemes, definitions, supporting quotations

If your design is still unsettled, sort that out before analysis. Our guide to dissertation methodology help covers matching design and analysis to your questions.

Prepare the Data Before You Test Anything

Errors made during data preparation surface later as strange results and awkward defense questions. Document each step so you can describe it in a few sentences.

  1. Screen for entry errors and impossible values.
  2. Describe missing data and how you handled it, for example listwise deletion or multiple imputation.
  3. Score scales as their authors instruct, including reverse-coded items.
  4. Report internal consistency for multi-item scales, such as Cronbach's alpha, for your sample.
  5. Check the assumptions of each planned test and note any alternative used.

Reporting Quantitative Results

Open with the sample: how many responded, how many were included and their key characteristics, usually in a table. Then give one section per research question or hypothesis.

Each section names the test, reports the result in standard notation, points to the table and states plainly whether the hypothesis was supported. In APA 7 style, give exact p values to two or three decimals, write p < .001 for very small values, and drop the leading zero for values that cannot exceed one, such as p and correlations.

Example results paragraph (APA 7). Nurses on units with protected break time reported lower emotional exhaustion (M = 21.4, SD = 7.9) than nurses on units without it (M = 26.1, SD = 8.6), t(182) = 3.84, p < .001, d = 0.57. Hypothesis 2 was supported. Full descriptive statistics appear in Table 3.

Notice what is missing: no explanation of why breaks matter and no comparison with earlier studies. Those belong in the discussion.

Why Effect Sizes Matter

A p value tells you how surprising a result would be if there were no effect. It does not tell you how large the effect is. With a big sample, trivial differences become significant; with a small one, meaningful differences may not.

Report effect sizes such as Cohen's d, eta squared or odds ratios, and confidence intervals where your software provides them. APA 7 recommends this, and many committees now expect it.

Presenting Qualitative Themes

For qualitative studies, the results chapter presents the themes your analysis produced, usually organized by research question. Each theme needs a short definition, evidence and a link back to the question.

  • Define the theme in one or two sentences, so the reader knows its boundaries.
  • Show breadth with quotations from several participants, not one articulate voice.
  • Note variation where participants differed, which strengthens credibility.
  • Use pseudonyms or codes consistently, as your ethics approval requires.

Keep your commentary descriptive. "Participants described the night shift as isolating" is a findings sentence. "This reflects wider organizational neglect" is discussion.

Integrating Mixed Methods Results

A mixed methods chapter that reports two strands separately and stops there has not shown integration. Committees look for the point where the strands meet.

A joint display is the common tool: a table that sets quantitative results beside related qualitative themes, with a column stating whether they converge, diverge or expand on each other. Whether integration sits in the results or the discussion chapter depends on your design and your program's template, so check with your chair.

Tables and Figures

Good tables carry the detail so your prose can stay readable. In APA 7, tables and figures are numbered in bold, the title appears on the next line in italic title case, and notes sit below.

DoAvoid
Refer to every table and figure in the textTables that appear with no mention
Round consistently, usually to two decimalsPasting raw software output
Report the same statistics across similar tablesSwitching formats between sections
Explain abbreviations in a table noteUndefined codes and variable names

When Results Do Not Support Your Hypotheses

Non-significant and unexpected results are reported exactly like significant ones: factually and in full. They are a normal part of research, not a failure.

Never drop a planned analysis because the result was disappointing, and never add tests after the fact without labeling them as exploratory. Changing or selectively reporting results is research misconduct. The discussion chapter is where you consider explanations such as power, measurement or context.

Closing the Chapter with a Summary

End the results chapter with a short summary that restates the answer to each research question in a sentence or two. It gives your committee a clear checklist and gives the discussion chapter its agenda.

Keep the summary factual. A simple pattern works well: "Research question 1 asked whether... The analysis showed... Research question 2 asked..." Then close with one sentence noting that the next chapter interprets these findings in light of the literature and the study's framework.

Common Errors Committees Catch

ErrorFix
Interpreting results in the results chapterMove explanation and literature links to the discussion
Reporting every test you ranReport planned analyses; label any extras as exploratory
Significance with no effect sizeAdd effect sizes and confidence intervals
Themes supported by one participantShow evidence across participants or reconsider the theme
No summary at the endClose with a short summary that sets up the discussion

How IvyDrafts Helps with Dissertation Data Analysis

Send your data file or analysis output (SPSS, R, Stata, Excel, NVivo or ATLAS.ti), your research questions, your method chapter and your style guide. A writer experienced in your type of analysis can run the planned tests on your real data, or work from completed output, and deliver a custom results chapter with formatted tables.

We use only your genuine data and never invent, alter or selectively report results. The work is for study and reference while you prepare your own chapter, and free revisions within the scope of your original request are included.

Turn Your Analysis into a Results Chapter

Send your data or output, research questions, method chapter and style guide. Your writer organizes the findings by question, in correct reporting format, using only your real results.

Order Data Analysis Help

Free revisions · Full refund if late · Written from scratch for your order

Frequently Asked Questions

Should the results chapter explain what the findings mean?

No. It reports findings by research question. Meaning, implications and links to earlier studies belong in the discussion chapter.

What should I report for each statistical test?

The test used, the test statistic, degrees of freedom where relevant, the exact p value, an effect size and, ideally, a confidence interval, with a pointer to the relevant table.

How do I write p values in APA 7?

Give exact values to two or three decimals, such as p = .032, and write p < .001 for very small values. Omit the leading zero because p cannot exceed one.

Can you run the analysis or only write it up?

Either. Some students send completed output for write-up; others send their data for the planned analysis to be run first. Both use only your real data.

How many quotations should support a qualitative theme?

There is no fixed number, but evidence from several participants shows a theme is a pattern rather than one person's view. Follow your method's conventions and your chair's guidance.

What is a joint display?

A table or figure that places quantitative results and qualitative themes side by side and states how they relate, used to show integration in mixed methods research.

What if my results did not support my hypotheses?

Report them factually like any other result. Discuss possible reasons in the discussion chapter, and never alter or drop planned analyses.

Which software should I use?

Use whatever your program supports and your method suits, commonly SPSS, R or Stata for statistics and NVivo or ATLAS.ti for qualitative coding. The reporting principles are the same.