Key takeaways
- Results chapters fail more often from presentation and reporting errors than from wrong analyses.
- Copying SPSS output without narrative is the most frequent examiner complaint.
- Hiding non-significant results damages trust more than reporting them honestly.
Examiners read hundreds of dissertation data analysis chapters. The same mistakes appear repeatedly—mistakes that have nothing to do with intellectual ability and everything to do with rushing the write-up. Avoiding these errors can lift your results chapter from acceptable to impressive without changing a single analysis. This guide catalogues the most costly mistakes and shows how to fix them before submission.
Mistake 1: SPSS output dump
Pasting entire SPSS output windows into your chapter forces examiners to hunt for relevant numbers. They will not. Extract only the tables and statistics needed, format them in APA style, and narrate every value you report.
Mistake 2: No connection to research questions
Running tests without mapping them to hypotheses leaves examiners wondering why each analysis exists. Every test section opens with the hypothesis or question it addresses.
Mistake 3: Wrong or inconsistent APA formatting
- Inconsistent decimal places across tables.
- Missing italics on statistical symbols.
- Reporting p = .000 instead of p < .001.
- Tables without titles or with vertical gridlines.
- Figures missing axis labels or sample sizes.
Mistake 4: Ignoring assumption violations
Running parametric tests on clearly non-normal data without comment undermines credibility. Report assumption checks. If violated, describe remedial actions—non-parametric alternatives, transformations, or robust methods.
Mistake 5: Hiding non-significant results
Selective reporting is detectable when your methodology lists five hypotheses and results discuss three. Report all pre-specified hypotheses with equal structural treatment.
Mistake 6: Overinterpretation in results
Results chapter presents findings; discussion interprets them. Causal claims from correlational data, policy recommendations, and theoretical leaps belong in discussion—not results.
Mistake 7: Reporting the wrong table row
SPSS prints multiple tables per procedure. Students report Levene's test p-value as the t-test result, or model summary R² as the coefficient p-value. Double-check row labels every time.
Mistake 8: Missing effect sizes
p-values without effect sizes tell examiners half the story. Add Cohen's d, η², partial η², or odds ratios as appropriate. Many universities now require them.
Mistake 9: Unlabelled or misleading charts
Default SPSS charts with default titles look unprofessional. Customise labels, remove clutter, and ensure chart type matches data type.
Mistake 10: Qualitative results as transcript summary
Listing interview responses without thematic analysis is not a results chapter. Organise by themes, support with quotes, and demonstrate analytical coding.
Self-review protocol before submission
- 1Outline maps every objective to a results subsection.
- 2Every table referenced before it appears.
- 3Statistics verified against output.
- 4Non-significant findings included.
- 5Interpretation language checked for overclaim.
How supervisors and examiners respond
Fixable presentation errors trigger revision requests. Integrity concerns around selective reporting trigger deeper scrutiny of all analyses. Presentation quality signals overall care.
Professional data analysis support
If test selection, SPSS output interpretation, or results chapter writing is blocking your dissertation timeline, ReportLift data analysis support helps you run valid tests, interpret findings correctly, and report results to examiner and journal standards.