Dissertation Chapter 4 in SPSS: Structuring Your Results Section

Chapter 4 is where your dissertation stops describing what you planned to do and starts showing what you actually found. For many students, this is the hardest chapter to write — not because the analysis is difficult, but because it is hard to know how much to include, in what order, and how to turn SPSS output into clear, readable prose.

This guide shows you a structure that works for most quantitative dissertations: sample description first, then descriptive statistics, then assumption checks, then the results for each research question or hypothesis in order, and finally any supplementary analyses. It also gives you the click-by-click SPSS steps, a paste-ready syntax block, and APA-style reporting templates you can adapt to your own results.

A note before you start: Chapter 4 reports results. It does not explain why you chose your methods (that was Chapter 3), and it does not discuss what the results mean for theory or practice (that is Chapter 5). Keep interpretation in Chapter 4 brief and factual: say what the test showed, then move on.

The Structure That Works for Most Chapter 4s

A clear Chapter 4 usually follows this order:

  1. Introduction and overview — a short paragraph restating the purpose of the study, the research questions, and the order in which results will be presented.
  2. Sample and data screening — who took part, how many cases were usable, and how missing data and outliers were handled.
  3. Descriptive statistics — means, standard deviations, frequencies and percentages for your key variables.
  4. Assumption checks — the checks that support the tests you ran, reported briefly.
  5. Results by research question or hypothesis — one section per question, in the same order as Chapter 1 and Chapter 3.
  6. Supplementary or exploratory analyses — only if they genuinely add something, and clearly labelled as supplementary.
  7. Summary — a short recap of what was found, without new interpretation.

This order matters. Readers should be able to follow your argument from “who was in the study” to “did each hypothesis hold” without jumping back and forth between tables. If your university or supervisor prescribes a different order, follow their template — but keep the same logic: sample, descriptives, assumptions, then findings in question order.

Step 1: Describe Your Sample First

Every reader needs to know who your results describe before they can judge them. Start with the final analytic sample: how many people started, how many were excluded and why, and how many remained for analysis. Then summarise the sample’s composition — for example, age, gender, programme of study, or any other demographic variables your study collected.

In SPSS, frequencies give you this quickly. Go to Analyze > Descriptive Statistics > Frequencies. Move your demographic variables into the Variables box, make sure “Display frequency tables” is ticked, and click OK. For a continuous variable such as age, use Analyze > Descriptive Statistics > Descriptives instead, which gives you the mean and standard deviation.

In your chapter, present this as a short paragraph plus a table. A sentence such as “The final sample consisted of [N] participants” followed by a demographics table is enough. Do not paste the full SPSS frequency output into your dissertation — rebuild it as a clean table with counts and percentages, and report the mean and standard deviation for continuous variables.

Also report data screening briefly: how much data was missing, how you handled it (for example, listwise deletion or multiple imputation, as decided in Chapter 3), and whether any cases were removed as outliers and on what rule. One short paragraph is usually enough, and it shows the reader your final sample is trustworthy.

Step 2: Present Descriptive Statistics for Your Key Variables

Before any hypothesis test, show the reader what your variables look like on their own. For each scale or key variable, report the sample size, mean, standard deviation, and range. For categorical variables, report frequencies and percentages.

In SPSS, go to Analyze > Descriptive Statistics > Descriptives. Move your scale variables into the Variables box, click Options, and tick Mean, Std. deviation, Minimum and Maximum. Click Continue, then OK.

Present the results in a single descriptives table, one row per variable, and refer to it in the text: for example, “Table 1 presents descriptive statistics for all study variables.” Then highlight only the values worth commenting on — an unusually high or low mean, a variable with very little variation, or a score range that does not cover the full possible range. Do not describe every cell of the table in words; the table does the work, and your prose should point the reader to what matters.

If your study uses questionnaire scales, this is also the place to report reliability briefly, if your supervisor expects it here rather than in Chapter 3. A single sentence per scale is enough — the coefficient, and whether it is acceptable for your purposes.

Step 3: Report Assumption Checks Briefly

Readers and examiners want to know that your tests were appropriate, but Chapter 4 is not the place for a long methods discussion. Report each assumption check in one or two sentences, plus a small table or figure only where it genuinely helps.

The checks depend on your tests. For a t-test or ANOVA, that usually means normality and homogeneity of variance. For regression, it means linearity, normality of residuals, homoscedasticity, and multicollinearity. For correlation, it means linearity and the absence of extreme outliers.

In SPSS, a useful starting point is Analyze > Descriptive Statistics > Explore. Move your dependent variable into the Dependent List and your grouping variable into the Factor List, click Plots, and tick “Normality plots with tests” and “Histogram”. This gives you the Shapiro-Wilk test and Q-Q plots. For homogeneity of variance, Levene’s test appears in the t-test and ANOVA output itself, so you do not need a separate run.

In your chapter, write something like: “Normality was assessed using the Shapiro-Wilk test and visual inspection of Q-Q plots. The assumption was reasonably met for [variable], with [any exception noted]. Levene’s test indicated equal variances could be assumed.” If an assumption was violated, say so plainly and say what you did about it — for example, that you used a non-parametric alternative or a robust approach — and keep the detail consistent with Chapter 3.

Step 4: Present Results in Research-Question Order

This is the heart of the chapter. Give each research question or hypothesis its own headed section, in exactly the same order as in Chapters 1 and 3, and follow the same pattern inside each section:

  • Restate the question or hypothesis in one sentence.
  • Name the test you ran and why it fits.
  • Present the key table.
  • State the result in words, with the test statistic, degrees of freedom, p value and effect size.
  • State whether the hypothesis was supported or not.

Work through the common tests below in the same way, adapting the steps to your own variables.

Comparing two groups: independent-samples t-test. Go to Analyze > Compare Means > Independent-Samples T Test. Move your dependent variable into Test Variable(s) and your grouping variable into Grouping Variable, click Define Groups, enter the two group codes, and click OK. Read the “Equal variances assumed” row unless Levene’s test is significant, in which case read the “Equal variances not assumed” row. Report the group means and standard deviations first, then the test result.

Comparing three or more groups: one-way ANOVA. Go to Analyze > Compare Means > One-Way ANOVA. Move your dependent variable into Dependent List and your grouping variable into Factor. Click Post Hoc and choose a test appropriate to your data (Tukey is a common default when variances are equal), then click OK. Report the overall ANOVA result first, then the post-hoc comparisons that answer your question.

Looking at relationships: correlation. Go to Analyze > Correlate > Bivariate. Move your two variables into the Variables box, keep Pearson ticked for normally distributed continuous variables (or Spearman for ranked or non-normal data), and click OK. Report the correlation coefficient, its direction and strength, and the p value.

Predicting an outcome: regression. Go to Analyze > Regression > Linear. Move your outcome into Dependent and your predictors into Independent(s), click Statistics and tick Estimates, Model fit and Descriptives, then click OK. Report the overall model first (R-squared and the F test), then the individual predictors that were significant and those that were not.

In every case, lead with the answer to the question. A reader skimming your headings and first sentences should be able to tell which hypotheses were supported and which were not, without digging through tables.

Paste-Ready SPSS Syntax

If you prefer syntax, or your supervisor wants a reproducible record, the block below covers the core Chapter 4 runs. Replace the placeholder variable names with your own, and delete the sections you do not need. Comments (lines starting with an asterisk) explain each step and do not affect the analysis.

* Sample frequencies and descriptives.
FREQUENCIES VARIABLES=gender programme
  /ORDER=ANALYSIS.
DESCRIPTIVES VARIABLES=age score1 score2
  /STATISTICS=MEAN STDDEV MIN MAX.

* Normality checks.
EXAMINE VARIABLES=score1 BY group
  /PLOT NPPLOT HISTOGRAM
  /STATISTICS DESCRIPTIVES.

* Group comparison and relationship examples — adapt to your design.
T-TEST GROUPS=group(1 2)
  /VARIABLES=score1.
ONEWAY score1 BY group
  /POSTHOC=TUKEY.
CORRELATIONS VARIABLES=score1 score2
  /PRINT=TWOTAIL NOSIG.

* Regression example.
REGRESSION
  /DEPENDENT outcome
  /METHOD=ENTER predictor1 predictor2
  /STATISTICS COEFF R ANOVA.

Run the syntax via File > New > Syntax, paste it in, select the lines you want, and click Run > Selection. Save the syntax file alongside your data file — it is the cleanest audit trail you can offer an examiner who asks how a table was produced.

How to Interpret SPSS Output Without Overstating It

SPSS output is long, and most of it does not belong in your dissertation. For each test, you need only a handful of numbers: the descriptive statistics for the groups or variables involved, the test statistic, the degrees of freedom, the p value, and an effect size where one is available.

Read the output in this order: first check the sample sizes (a group with far fewer cases than expected usually means missing data or a coding error), then check the assumption test (such as Levene’s), then read the main test result. If the p value is below your alpha level (usually .05), the result is statistically significant — but that word means only that the pattern is unlikely to be due to sampling variation alone. It does not mean the effect is large or important, which is why you should also report an effect size such as Cohen’s d, eta-squared, or R-squared, and interpret it in plain language.

Be careful with two common traps. First, do not report every row of output: non-significant predictors, non-significant post-hoc comparisons and assumption-test details belong in a brief sentence or an appendix, not in paragraph after paragraph. Second, do not write causal language unless your design supports it. A significant correlation shows an association, not that one variable caused the other, and your Chapter 4 wording should reflect that.

Tables: Build Them Clean, Not Pasted

Examiners notice messy tables immediately. Never paste raw SPSS output tables into your dissertation. Rebuild each table with a clear number and title above it (for example, “Table 2. Independent-Samples t-Test Comparing Scores by Group”), column headings that a reader can understand without the SPSS manual, and only the statistics your text refers to.

A good rule is one table per research question, plus a demographics table and a descriptives table at the start. If a table grows beyond a page, split it or move the detail to an appendix and keep a summary table in the chapter. Every table must be referred to in the text before it appears, and every number in the text should match the table exactly — a mismatch between text and table is one of the easiest errors for an examiner to spot, so check them against each other before submission.

Reporting in APA Style (7th ed.)

APA style keeps Chapter 4 results consistent and easy to scan. The general pattern is: test statistic, degrees of freedom in parentheses, p value, and effect size. Use italics for statistical symbols such as t, F, r and p in your dissertation document (plain text here). Report exact p values where SPSS gives them, and use p < .001 for very small values rather than p = .000, since SPSS rounds to three decimals.

Templates you can adapt (replace the placeholders with your own values):

  • Independent-samples t-test: “Scores were higher in Group A (M = [value], SD = [value]) than in Group B (M = [value], SD = [value]), t(df) = [value], p = [value], d = [value].”
  • One-way ANOVA: “There was a significant difference between groups, F(df1, df2) = [value], p = [value], eta-squared = [value]. Post-hoc comparisons using Tukey’s test showed [which groups differed].”
  • Correlation: “Scores on [variable 1] were positively associated with scores on [variable 2], r(df) = [value], p = [value].”
  • Regression: “The model explained [percentage] of the variance in [outcome], R-squared = [value], F(df1, df2) = [value], p = [value]. [Predictor] was a significant predictor (beta = [value], p = [value]), whereas [predictor] was not (beta = [value], p = [value]).”

Always report group means and standard deviations alongside a t-test or ANOVA, and always give the direction of an effect — “higher”, “lower”, “positively associated” — not just the statistic. A result sentence without direction forces the reader to reconstruct your finding from the table, which is exactly what Chapter 4 prose should prevent.

Key Takeaways

  • Order your chapter: sample, descriptives, assumptions, then one section per research question in question order, then a short summary.
  • Chapter 4 reports findings only — save the meaning and implications for Chapter 5.
  • Rebuild SPSS output as clean, numbered tables; never paste raw output into the chapter.
  • For each test, report the test statistic, degrees of freedom, exact p value and an effect size, and state the direction of the effect.
  • Check every number in your text against its table before submission — text-table mismatches are easy for examiners to spot.

Frequently Asked Questions

How long should Dissertation Chapter 4 be?

There is no fixed length, because it depends on how many research questions you have and how many tables each one needs. As a rough guide, a chapter that answers three or four questions with one table each is often the right scale; a chapter padded with repeated output or one that hides its findings in appendices is usually a sign the structure needs tightening. Follow your university’s word-count guidance first.

Should I include SPSS output in my dissertation?

No — include clean, rebuilt tables in the chapter and, if your university requires it, put fuller output in an appendix. Raw SPSS output is hard to read, includes statistics you never discuss, and looks unfinished to an examiner.

Where do I report non-significant results?

In the same section as the hypothesis they relate to, with the same reporting pattern. A non-significant result is still a finding: state the statistic and p value, say the hypothesis was not supported, and move on. Do not hide non-significant results or move them to a footnote.

What if an assumption is violated?

Say so briefly, say what you did (for example, a non-parametric alternative, a transformation, or a robust method, as appropriate to your design), and report the result of that alternative analysis. Trying to hide a violation is riskier than reporting it honestly with a sensible remedy.

Do I need an effect size for every test?

You should report one wherever a standard effect size exists for your test — Cohen’s d for t-tests, eta-squared for ANOVA, r for correlations, R-squared for regression. Effect sizes tell the reader whether a significant result is also a meaningful one, and examiners increasingly expect them.

Need This Done on Your Data? Get a Free Quote

If your Chapter 4 is stalling — the analysis is run but the tables will not come together, the APA reporting feels uncertain, or the deadline is closer than the chapter is to finished — you do not have to work through it alone. Our team runs SPSS analyses for dissertations every day, from data screening and assumption checks to clean, APA-style results tables and a results section your supervisor can actually follow.

We can take your dataset and research questions and return a structured Chapter 4 draft: sample description, descriptives, assumption checks, and results reported question by question, with the syntax saved so every table can be reproduced. You stay in control of the interpretation and the final wording — we handle the analysis and the presentation.

Need this done on your data? Get a free quote → /services/