What belongs in the results chapter, and what should wait for the discussion? The results chapter answers one question for each research question or hypothesis: what did the data show? It does not say what the finding means, whether it agrees with earlier studies or what should be done about it; those are the discussion chapter’s jobs. Keeping that boundary, presenting results in a logical order and following APA 7th edition conventions is most of what examiners look for. This guide covers the order, the conventions for descriptive and inferential statistics, the rules for tables and figures, a worked example and the errors that recur in marked theses.
This article is part of our guide How to Write a Master’s Thesis by the Mastermind PhD team and relates to our Statistical Analysis service. Sources last checked: November 2026.
What is the results chapter? The results chapter (usually chapter 4, about 20% of the thesis) presents the findings of the data analysis in the order of the research questions or hypotheses, using text, tables and figures. In a quantitative thesis it reports the sample, data screening, descriptive statistics, reliability, assumption checks and the inferential tests with exact p values, effect sizes and confidence intervals, and states whether each hypothesis was supported.
What goes in the results chapter, and what stays out?
In: the final sample and how it was reached, data cleaning, descriptive statistics for every variable used, scale reliability, assumption checks, the inferential results for each question or hypothesis in order, and whether each hypothesis was supported.
Out: explanations of why a result occurred, comparisons with earlier studies, implications and recommendations, raw software output, and any analysis not planned in the methodology chapter unless it is labelled as exploratory. The results chapter may say “contrary to H2, the interaction was not significant” and stop there; the paragraph of interpretation belongs in the discussion.
In what order should the results be presented?
A fixed order in the results chapter helps the examiner find what the rubric asks for:
- Response and sample: invitations, responses, exclusions, final n.
- Sample characteristics: a table of demographic and background variables.
- Data screening: missing data and how it was handled, outliers, normality.
- Descriptive statistics for the study variables.
- Reliability of the scales (Cronbach’s alpha or McDonald’s omega).
- Assumption checks for the planned tests.
- Inferential results, one subsection per research question or hypothesis, in the order of chapter 1.
- Summary table of hypotheses and whether each was supported.
📱 Ask a statistician about your data — Send a short description of your instrument and sample size on WhatsApp and get a suitable analysis plan. Every analysis comes with the raw SPSS file and a results report.
How do you describe the sample and the data screening?
State the numbers plainly: “Of 410 developers invited, 251 opened the survey and 221 completed it (response rate 54%). Nine cases with more than 10% missing items were removed, leaving 212 for analysis.” Then give a sample characteristics table with counts and percentages for categorical variables and means and standard deviations for continuous ones. Report how missing values were treated (listwise deletion, multiple imputation) and the proportion missing, outlier checks (z scores beyond 3.29, boxplots) and normality (skewness and kurtosis, Shapiro-Wilk where appropriate), each with the decision you took. These paragraphs are short, but their absence makes examiners doubt the later tests.
How do you report descriptive statistics in APA 7?
The APA Numbers and Statistics Guide gives rules that most departments apply regardless of discipline:
- Report means and standard deviations for integer scales (such as Likert-type items) to one decimal, and other statistics to two decimals.
- Use the symbol with a mathematical operator (M = 2.8) and the word in running text (“the means were higher”).
- Italicize letters used as statistical symbols (M, SD, n, t, F, p, r, d); do not italicize Greek letters.
- Put a zero before the decimal point when the statistic can exceed 1 (0.84 for d); omit it when it cannot (p = .03, r = .42).
- Do not repeat in the text every number that is already in a table; the text points to the pattern, the table holds the detail.
A descriptive table typically has columns for variable, n, M, SD, minimum, maximum, skewness and kurtosis, with scale ranges in a note.
How do you report inferential tests?
Each inferential result in the results chapter needs the test statistic with degrees of freedom, the exact p value (two or three decimals, or p < .001), an effect size and, where available, a 95% confidence interval. The templates below use invented data.
| Test | Template sentence (fictional figures) |
|---|---|
| Independent samples t test | Developers with structured onboarding reported lower turnover intention (M = 2.4, SD = 0.8) than those without (M = 3.1, SD = 0.9), t(210) = 5.92, p < .001, d = 0.82, 95% CI of the difference [0.47, 0.93]. |
| One-way ANOVA | Turnover intention differed by company size, F(2, 209) = 4.31, p = .015, η² = .04; Tukey tests showed the difference lay between the smallest and largest firms (p = .011). |
| Chi-square test of independence | Developers without a named mentor were more likely to report job search activity, χ²(1, N = 212) = 6.84, p = .009, φ = .18. |
| Pearson correlation | Onboarding structure was negatively correlated with turnover intention, r(210) = -.42, p < .001, 95% CI [-.52, -.30]. |
| Multiple regression | The model explained 31% of the variance in turnover intention, F(3, 208) = 31.2, p < .001, adjusted R² = .30. Onboarding structure was a negative predictor, B = -0.36, SE = 0.07, β = -.38, t = -5.14, p < .001, 95% CI [-0.50, -0.22]. |
Three conventions deserve emphasis. First, write p < .001 rather than p = .000; a probability is never zero, and software prints zeros only because it rounds. Second, report an effect size for every test and interpret it against the conventional benchmarks (Cohen, 1988: d of 0.2, 0.5 and 0.8; η² of .01, .06 and .14; r of .1, .3 and .5), noting that benchmarks are rough. Third, report exact p values for non-significant results too (p = .27, not “ns” or “p > .05”).
What are the rules for tables and figures?
APA 7 rules that examiners check:
- Number tables and figures consecutively in the order they are mentioned, and mention each in the text before it appears.
- Give each a brief, descriptive title in italic title case above it, on the line below the number.
- Use horizontal lines only (top, below the headings, bottom); no vertical lines and no decorative shading.
- Define abbreviations in a note below the table (Note. CI = confidence interval) and mark significance levels with asterisks explained in the same note (p < .05, p < .01, p < .001).
- Keep the same number of decimals within a column.
- Do not duplicate: a result in a table is summarized, not repeated, in the text.
- Report correlations in one triangular matrix with means and standard deviations in the first columns.
How do you handle non-significant and unexpected results?
Report them in the results chapter with the same completeness as significant ones: test statistic, exact p value, effect size and confidence interval. Do not describe a p of .07 as “approaching significance” or “a trend”; it is a non-significant result at the alpha you set, and the confidence interval shows how large the effect could still be. State plainly which hypothesis was not supported. If an unplanned analysis helps explain the pattern, label it as exploratory and keep it separate from the planned tests. Exploring the reasons belongs in the discussion.
A worked example paragraph
Hypothesis 1 predicted that onboarding structure would be negatively associated with turnover intention. The assumptions of linearity and homoscedasticity were met on inspection of the scatterplot and residual plots, and no bivariate outliers were identified. Onboarding structure was negatively correlated with turnover intention, r(210) = -.42, p < .001, 95% CI [-.52, -.30], a medium-to-large effect by Cohen’s benchmarks. In the multiple regression controlling for tenure and company size (Table 7), onboarding structure remained a significant negative predictor, β = -.38, p < .001, and the full model explained 31% of the variance in turnover intention. Hypothesis 1 was therefore supported.
Note what the paragraph does not do: it does not say why structure matters, cite earlier studies or recommend anything. Those sentences belong to the next chapter.
Which mistakes do examiners flag most often?
- p = .000, or p values with no test statistic and degrees of freedom.
- “Highly significant” or “very significant”; significance is a threshold, not a scale.
- No effect sizes, or effect sizes reported without interpretation.
- Percentages without counts, or counts without the n they refer to.
- Interpreting or comparing with the literature inside the results chapter.
- Tables that duplicate the text, or software output pasted as images.
- Assumption checks absent, so the choice of a parametric test is unexplained.
- Results in the order the software produced them rather than the order of the questions.
- A hypothesis from chapter 1 that silently disappears in chapter 4.
A statistician can check the analysis and the reporting against these conventions before submission; the hypotheses, the interpretation and the decisions about what to include stay with you, and you should be able to explain every number in your viva or defense.
📱 Ask on WhatsApp now — 5 years of experience across all disciplines and 14 academic services. Two-stage payment (80/20); revisions are free until supervisor approval, within 30% of the chapter.
Frequently asked questions
Should I interpret the results in the results chapter?
No, beyond stating whether each hypothesis was supported and describing the direction and size of effects. Explanations, comparisons and implications belong in the discussion. If your program allows a combined results and discussion chapter, keep the two functions in separate, labelled paragraphs.
How many decimal places should I use?
Per APA 7: means and standard deviations of integer scales to one decimal; other statistics to two; p values to two or three decimals, with p < .001 for smaller values. Omit the leading zero for statistics that cannot exceed 1, such as p and r.
Where do the assumption tests go?
In the results chapter, just before the test they concern, or in a single data screening section at the start if the same checks apply to all tests. State the check, the value and the decision in one or two sentences; full output can go in an appendix.
Should I include SPSS output in the thesis?
Not in the chapter. Rebuild the relevant numbers as APA-style tables and place complete output in an appendix only if your program requires it. Keep the data file and syntax in case the examiners ask.
How do I report a result that was not significant?
Exactly as you report a significant one: test statistic, degrees of freedom, exact p value, effect size and confidence interval, followed by the statement that the hypothesis was not supported. Avoid “approached significance” and avoid dropping the result from the chapter.
Sources
- American Psychological Association. (2020). Publication Manual of the American Psychological Association (7th ed.). https://apastyle.apa.org/
- American Psychological Association. Numbers and Statistics Guide, APA Style 7th Edition. https://apastyle.apa.org/instructional-aids/numbers-statistics-guide.pdf (accessed October 2026)
- Appelbaum, M., Cooper, H., Kline, R. B., Mayo-Wilson, E., Nezu, A. M., & Rao, S. M. (2018). Journal article reporting standards for quantitative research in psychology: The APA Publications and Communications Board task force report. American Psychologist, 73(1), 3–25. https://doi.org/10.1037/amp0000191
- Field, A. (2018). Discovering Statistics Using IBM SPSS Statistics (5th ed.). SAGE.
- Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences (2nd ed.). Lawrence Erlbaum.
- Lakens, D. (2013). Calculating and reporting effect sizes to facilitate cumulative science: A practical primer for t-tests and ANOVAs. Frontiers in Psychology, 4, 863. https://doi.org/10.3389/fpsyg.2013.00863
- Wasserstein, R. L., & Lazar, N. A. (2016). The ASA statement on p-values: Context, process, and purpose. The American Statistician, 70(2), 129–133. https://doi.org/10.1080/00031305.2016.1154108
Sources last checked: November 2026. This article is general educational guidance and does not replace your university regulations or your supervisor’s instructions. How we prepare our content.