The data analysis chapter (typically Chapter 4) is where your research transforms from theory into findings. It's often the chapter that makes or breaks a thesis — yet it's the one most students struggle with the most.
This guide gives you a proven framework for writing a clear, well-structured data analysis chapter that your supervisor will approve.
The Purpose of Chapter 4
Your data analysis chapter has one job: present your findings in a way that directly answers your research questions.
Everything in this chapter should connect back to:
- Your research questions (from Chapter 1)
- Your hypotheses (if applicable)
- Your methodology (from Chapter 3)
If a table, chart, or paragraph doesn't help answer a research question, it doesn't belong in this chapter.
Recommended Structure
Here's the framework that works for most health science theses:
4.1 Introduction
A brief paragraph (3-5 sentences) stating:
- What this chapter covers
- How many respondents/samples were analyzed
- What tools were used for analysis (SPSS, Excel, etc.)
4.2 Socio-Demographic Characteristics
Present your respondents' profile:
- Age distribution
- Gender breakdown
- Educational level
- Professional experience
- Other relevant demographics
Best format: A single summary table with frequencies and percentages.
| Variable | Category | Frequency (n) | Percentage (%) |
|---|---|---|---|
| Gender | Male | 45 | 37.5 |
| Female | 75 | 62.5 | |
| Age Group | 20-29 | 32 | 26.7 |
| 30-39 | 58 | 48.3 | |
| 40-49 | 22 | 18.3 | |
| 50+ | 8 | 6.7 |
4.3 Analysis by Research Question
This is the core of your chapter. For each research question, present:
- Restate the research question — word it exactly as in Chapter 1
- Present the relevant data — tables, charts, or both
- Describe the findings — what the numbers show (not what they mean — that's Chapter 5)
- Statistical test results (if applicable) — chi-square, t-test, ANOVA, correlation, etc.
4.4 Hypothesis Testing (If Applicable)
For each hypothesis:
- State the null and alternative hypotheses
- Present the test used and why
- Show the test statistic, degrees of freedom, and p-value
- State whether the hypothesis is accepted or rejected based on your significance level (usually p < 0.05)
Example:
H1: There is a significant relationship between years of experience and hand hygiene compliance.
A chi-square test of independence was performed. The results showed a statistically significant association between years of experience and hand hygiene compliance, X2(3) = 12.45, p = 0.006. Therefore, the null hypothesis is rejected.
Presenting Data Effectively
When to Use Tables vs. Charts
- Tables — when exact numbers matter (frequencies, percentages, test statistics)
- Bar charts — for comparing categories
- Pie charts — for showing proportions of a whole (use sparingly — only when you have 2-5 categories)
- Line charts — for trends over time
- Scatter plots — for showing relationships between two variables
Table Formatting Rules
- Every table must have a number and title above it (Table 4.1: Distribution of Respondents by Gender)
- Use consistent decimal places throughout (1 or 2, not a mix)
- Include both frequency (n) and percentage (%) for categorical data
- Bold or highlight totals and significant values
- Keep tables simple — if it has more than 6 columns, split it
Common Statistical Tests
| Research Goal | Data Type | Recommended Test |
|---|---|---|
| Compare two group means | Continuous, normal | Independent t-test |
| Compare two group means | Continuous, not normal | Mann-Whitney U |
| Compare 3+ group means | Continuous, normal | One-way ANOVA |
| Test relationship between categories | Categorical | Chi-square |
| Measure correlation | Continuous | Pearson's r |
| Measure correlation | Ordinal | Spearman's rho |
| Predict an outcome | Mixed | Regression |
Common Mistakes to Avoid
- Interpreting results in Chapter 4 — Only describe findings here. Interpretation and discussion belong in Chapter 5.
- Presenting raw data — Show summaries, not individual responses.
- Skipping research questions — Every research question from Chapter 1 must be addressed.
- Using the wrong chart type — A pie chart with 15 slices helps no one.
- Ignoring missing data — State how many responses were incomplete and how you handled them.
- p-value misinterpretation — p < 0.05 means the result is statistically significant, not that it's important or large.
Tools for Your Analysis
- SPSS — Most common for health science research. Great for descriptive stats, chi-square, t-tests, ANOVA, and regression.
- Excel — Fine for descriptive statistics, frequency tables, and charts. Limited for advanced inferential statistics.
- Stata — Preferred in epidemiology and public health research.
- Python/R — For advanced or custom analyses (not typically required for undergraduate or master's theses in Nigeria).
Before You Submit: Checklist
- Every research question from Chapter 1 is addressed
- Tables are numbered and titled correctly
- Charts are clear, labeled, and referenced in the text
- Statistical tests are appropriate for the data type
- P-values are reported with test statistics
- Hypotheses are clearly accepted or rejected
- No interpretation or discussion (save that for Chapter 5)
- The chapter has a brief introduction and summary
Need Help With Your Analysis?
Writing the data analysis chapter doesn't have to be a solo struggle. Our Research Support service helps graduate students with:
- Choosing the right statistical tests
- Running analysis in SPSS or Excel
- Structuring and writing Chapter 4
- Reviewing your tables and charts for accuracy
We also have ready-made reference materials in our Research Store — complete dissertations and theses that show you exactly how a well-written analysis chapter looks.
Questions about your specific thesis? Send us a message and we'll point you in the right direction.
