Dissertation results chapter help

The results chapter is the one place a dissertation cannot hide a weak decision. Every earlier choice, the study design, the variables, the sample, becomes visible the moment the numbers go on the page. Supervisors send this chapter back more than any other, and usually for the same reasons: output pasted from SPSS without interpretation, tables that do not match the reporting style, or a finding stated more strongly than the test supports.

It is also the chapter where the writing and the statistics cannot be separated. Wording a result correctly requires knowing exactly what the test did, which is why handing it to a general academic editor rarely fixes it. If you would rather work through the structure yourself first, our guide to structuring a results chapter section by section sets out the order and what belongs in each part.

1. 1

   The analysis, run and checked

   Your data cleaned, the correct test run in SPSS or R, and the assumption checks documented alongside it.

2. 2

   Tables and figures

   Publication-quality tables and figures, numbered, captioned, and formatted to your department's style guide.

3. 3

   The written results

   Each finding reported with its test statistic, p-value, and effect size, worded to claim exactly what the data supports.

4. 4

   Question-by-question mapping

   Every result traced back to the research question or hypothesis it answers, so the chapter reads as one argument.

A result is not reported until three things are on the page: what the test found, how large the effect was, and how certain the estimate is. A bare p-value answers only the first, which is why chapters built on significance alone come back marked for revision. Reporting a t-test properly means the test statistic, the degrees of freedom, the exact p-value, and a standardised effect size such as Cohen's d, with a confidence interval where the style guide asks for one.

• Descriptive statistics before inferential ones, so the reader knows the sample before reading claims about it.
• Exact p-values rather than a bare threshold, except where convention sets a floor such as p < .001.
• An effect size beside every significance test, so magnitude is not confused with certainty.
• Assumption checks reported, including the ones that failed and what was done about them.
• Missing data and exclusions stated with the rule that was applied, before any result depends on them.

The conventions themselves are covered in our guide to reporting statistics in APA style, and the distinction that trips up most chapters is set out in what a p-value does and does not tell you.

The most common structural error is a results chapter that starts interpreting. Results report what was found; the discussion explains what it means and how it sits against the literature. Mixing them costs marks twice, because the results read as speculation and the discussion has nothing left to say.

The line is easier to hold with a rule: in results, a sentence may state what the test showed and point to the table that shows it, and nothing else. Why the finding matters, whether it replicates earlier work, and what its limitations are all belong in the next chapter. We write to that boundary, and flag anything in your existing draft that crosses it.

If your data is collected and the chapter is all that remains, this is the right scope. If the analysis itself has not been decided yet, or the study design still needs settling, start with a statistical consultant on the full project. For a dataset that needs analysing from scratch, see analysis of an existing dissertation dataset. Doctoral projects with advanced modelling are covered by statistical analysis support for a PhD, and if the work is in SPSS specifically, see hands-on SPSS support.