To write a data analysis report, you describe your data, show how you prepared and checked it, then report the result of each planned analysis clearly and accurately. In a dissertation this is usually the results chapter, often Chapter 4.
This page focuses on quantitative reports. It covers structure, data screening, choosing and reporting tests, presenting tables and the mistakes examiners notice most. Reporting conventions follow APA 7th edition, which many social and health science departments use; check your own style guide.
What a Data Analysis Report Must Show
- Who was in the sample and how complete the data were.
- How the data were cleaned, coded and checked.
- Whether the assumptions of each test were met.
- The result of every planned analysis, organized by research question or hypothesis.
- Effect sizes alongside significance tests.
Interpretation of what results mean for theory and practice usually belongs in the discussion chapter.
Data Analysis Report Structure
| Section | Contents |
|---|---|
| Introduction | Brief restatement of purpose, research questions and hypotheses |
| Data preparation | Cleaning, missing data, outliers, recoding, scale reliability |
| Sample description | Response rate, demographics, descriptive statistics for key variables |
| Assumption checks | Normality, homogeneity of variance, linearity, multicollinearity as relevant |
| Results by question | One subsection per research question or hypothesis, with test results and tables |
| Additional analyses | Any exploratory or sensitivity analyses, clearly labeled as such |
| Summary | Short statement of which hypotheses were supported |
Step-by-Step: From Dataset to Write-up
- Restate the questions. Readers should see immediately what each analysis answers.
- Screen the data. Check for entry errors, impossible values and duplicate cases.
- Handle missing data. Report how much was missing and the method used, such as listwise deletion or multiple imputation, with a reason.
- Check scale reliability. Report Cronbach's alpha or another reliability index for multi-item scales.
- Describe the sample. Use a demographics table and means, standard deviations or frequencies for main variables.
- Check assumptions. Report the checks and what you did if an assumption was violated.
- Run and report the planned tests. Follow the analysis plan from your proposal and say if anything changed.
Choosing the Right Test
| Question type | Parametric test | Non-parametric alternative |
|---|---|---|
| Compare two independent groups | Independent-samples t-test | Mann-Whitney U |
| Compare the same group twice | Paired-samples t-test | Wilcoxon signed-rank |
| Compare three or more groups | One-way ANOVA | Kruskal-Wallis |
| Relationship between two continuous variables | Pearson correlation | Spearman correlation |
| Association between two categorical variables | Chi-square test of independence | Fisher's exact test for small expected counts |
| Predict a continuous outcome | Multiple linear regression | Depends on the problem; seek advice |
| Predict a binary outcome | Logistic regression | Not applicable |
The right test depends on your design, measurement level and assumptions, so treat this as a starting point and confirm choices with your advisor or a statistician.
Reporting Results in APA Style
APA style reports the test statistic, degrees of freedom, exact p value and an effect size. Give p values to two or three decimal places, and write p < .001 for very small values.
Sample excerpt (illustrative format only). H1 predicted that night-shift nurses would report higher burnout than day-shift nurses. An independent-samples t-test showed that night-shift nurses (M = 3.42, SD = 0.71) scored higher than day-shift nurses (M = 3.05, SD = 0.68), t(146) = 3.24, p = .001, d = 0.53. H1 was supported.
A multiple regression with workload, staffing ratio and years of experience as predictors explained 31% of the variance in burnout, F(3, 144) = 21.56, p < .001, R² = .31. Workload was the strongest predictor (β = .42, p < .001).
The numbers above only show the format. Every figure in your report must come from your own analysis.
Tables and Figures
Use tables for exact values and figures for patterns. Every table and figure needs a number and title, and must be referred to in the text.
- Do not paste raw software output into the chapter; rebuild it as a clean table.
- Do not repeat the same numbers in both a table and the text; summarize the key point in prose.
- Label axes and units, and start bar chart axes at zero.
- Choose a chart type that fits the data: bar charts for groups, scatterplots for relationships, line charts for change over time.
Common Mistakes in Data Analysis Reports
- Reporting p values without effect sizes.
- Writing "p = .000", which software shows but should be reported as p < .001.
- Saying a result "proves" a hypothesis; use "supported" or "not supported".
- Treating non-significant results as evidence of no effect.
- Running unplanned tests until something is significant without saying so.
- Interpreting causes from correlational data.
Checklist Before You Submit
- Every research question and hypothesis has a matching result.
- Missing data, outliers and assumption checks are reported.
- All statistics include df, exact p values and effect sizes.
- Tables and figures are numbered, titled and cited in the text.
- Exploratory analyses are labeled as exploratory.
- Numbers in the text match the tables exactly.
How DissertationHorse Helps with Data Analysis Reports
Send your research questions, hypotheses, analysis plan and your anonymized dataset or software output. A writer with statistical training prepares a custom data analysis report or results chapter from your own data, with clean tables and accurate reporting, for you to study and use as a reference.
We never invent, alter or "improve" data or results. Free revisions within the scope of your original request are included.
Need Your Results Chapter Written Up?
Send your research questions, hypotheses, dataset or output and guidelines. Your writer prepares a custom data analysis report from your own data, written to that brief.
Order Your Data Analysis ReportFree revisions · Full refund if late · Written from scratch for your order
Frequently Asked Questions
Usually not in the main text. Rebuild key results as clear tables, and put full output in an appendix only if your program asks for it.
Report it exactly as found. Non-significant results are valid findings, and you explore possible reasons in the discussion chapter.
In most dissertations, the results chapter reports findings and the discussion chapter interprets them. Some programs combine the two, so check your template.
APA 7 generally recommends two decimal places for most statistics, with p values reported to two or three decimals. Follow your style guide consistently.
Yes. APA style and most examiners expect effect sizes, because significance alone does not show how large or important an effect is.
In a dissertation they are usually the same thing. A standalone data analysis report, for example for a module or client, may also include brief recommendations.