You ran the tests for your quantitative research dissertation, pasted the tables, reported the p-values, and still received a disappointing grade. Why? Because simply running tests isn’t enough. Many statistics chapter marking rubrics penalise students who treat their dissertation results chapter like a software manual rather than a cohesive scientific argument.
As a consultant, I regularly decode dissertation marking criteria. Examiners aren’t rewarding the sheer volume of your dissertation statistics. Four carefully selected analyses always beat ten poorly justified ones. Here is what they are actually evaluating.
Quick answer
Examiners score five things in a statistics chapter: whether you justified your test, checked assumptions, interpreted rather than described results, considered effect size and answered each hypothesis. Running the software is the easy part.
Methodological justification
Did you explain why you chose a specific test? A strong statistics chapter proves your chosen method specifically fits your data type and research design.
Assumption checking
Did you demonstrate an awareness of mathematical rules? Simply running an SPSS dissertation analysis without checking normality or evaluating Levene’s test loses marks immediately.
Interpretation vs. description
Markers want accurate SPSS results interpretation, not a regurgitated list of numbers.
Weak:
“ANOVA showed p = .021.”
Stronger:
“The one-way ANOVA indicated a statistically significant difference between the groups, followed by appropriate post-hoc analysis to identify where the differences occurred.”
The second demonstrates actual statistical reasoning, whereas the first just repeats output.
Effect size and practicality
Did you consider the magnitude of the finding? Relying solely on statistical significance while ignoring effect size reveals a shallow understanding of your statistical analysis dissertation project.
Alignment with hypotheses
Every analysis must clearly answer a stated hypothesis. Extraneous testing looks like fishing for significance.
Rubric checklist
| What markers score | What earns marks | What loses marks |
|---|---|---|
| Methodological justification | Explains why the test fits the data type and design | Runs a test with no reasoning |
| Assumption checking | Reports normality and Levene’s test results | Skips assumptions entirely |
| Interpretation | Explains what the result means for the hypothesis | Repeats numbers from the output |
| Effect size | Reports and discusses magnitude | Relies on the p-value alone |
| Hypothesis alignment | Each test answers a stated hypothesis | Extra tests that look like fishing |
The final takeaway
When figuring out how to write a statistics chapter, remember this: the goal is not to show the examiner how well you navigate software. The goal is to demonstrate that you can make, justify, and interpret appropriate statistical decisions to answer your research questions. See also the 15 most common SPSS mistakes researchers make.
Want your results chapter checked against your marking criteria?
We review your analysis, interpretation and write-up so every section earns marks.
Get SPSS helpFrequently asked questions
How many statistical tests should I include in my results chapter?
As many as your hypotheses need and no more. Four carefully selected analyses that are well justified beat ten poorly justified ones, and extra tests can look like fishing for significance.
What is the difference between describing and interpreting results?
Describing repeats the numbers from the output. Interpreting explains what the result means for your hypothesis, including the test used, the direction of the difference and what follow-up analysis showed.
Do examiners expect effect sizes?
Yes, in most marking schemes. Relying only on statistical significance suggests a shallow understanding, so report the effect size and say whether the finding is practically meaningful.
Comments are closed