Three studies run, every analysis done, and the results and discussion chapters of your psychology PhD still exist as SPSS output, R scripts and a folder of notes. The funded period ends in months, the supervisor wants a full draft, and each time you open the file the question is the same: which of the forty tests goes in, and what do you say about them?
The cost of leaving the chapters unwritten is measured in an unfunded writing-up year, a deferred submission, and a viva in which the examiner asks why an analysis that appears in the appendix was not in the results. Psychology examiners now read empirical chapters with the replication crisis in mind, and a results chapter that cannot distinguish confirmatory from exploratory analysis is examined as if the confirmatory analyses were exploratory too. This guide sets out how the empirical chapters of a UK psychology doctorate are built, what the field’s reporting standards require, and the workflow that turns analysed data into examinable chapters in weeks. Open a free Tesify workspace and build the first results chapter alongside as you read.
Why psychology chapters are examined harder than they were
The reason is public. In 2015 the Open Science Collaboration reported in Science replications of 100 experimental and correlational studies from three psychology journals: 97% of the original studies had statistically significant results and 36% of the replications did, with replication effects about half the magnitude of the originals and 47% of original effect sizes falling inside the replication’s 95% confidence interval. Four years earlier, Simmons, Nelson and Simonsohn had shown in Psychological Science that flexibility in data collection, analysis and reporting — researcher degrees of freedom — makes it unacceptably easy to accumulate significant evidence for a false hypothesis, and proposed six disclosure requirements for authors and four guidelines for reviewers. A doctoral examiner in 2026 has read both, and reads your results chapter for the degrees of freedom you did not disclose.
The defence is structural: a results chapter that separates preregistered from exploratory analyses, reports every test that was run, and gives effect sizes with confidence intervals, followed by a discussion that interprets in proportion to that evidence. The APA’s Journal Article Reporting Standards for quantitative research are the checklist most UK psychology departments now expect the thesis to meet, and they are the spine of what follows.
How the empirical chapters of a psychology thesis are shaped
Most UK psychology PhDs are built as a series of studies, and the thesis either gives each study its own chapter — introduction, method, results, discussion — or groups the studies into empirical chapters that share a method and closes with a general discussion. The choice is set by the programme and the supervisory team, and the DClinPsy thesis follows a different shape again. What every shape shares is that results and discussion are separated at the study level, and that a general discussion chapter integrates across studies. The table sets out what each part carries.
| Part | What it contains | What it must not contain | Length in a three-study thesis |
|---|---|---|---|
| Study results | Participant flow and exclusions, descriptives, assumption checks, confirmatory tests with effect sizes and confidence intervals, then clearly labelled exploratory analyses | Interpretation, theory, comparison with prior work | 3,000–5,000 words per study, plus tables |
| Study discussion | What the study found against its hypotheses, in proportion to the effect sizes; the study’s own limitations; what the next study was designed to resolve | The general theoretical argument; a limitations list that applies to every study | 1,500–2,500 words per study |
| General discussion | Integration across studies, the theoretical contribution, alternative explanations, limitations as a programme, implications and future work | Restatement of each study’s results in turn | 8,000–12,000 words |
The results chapter, to JARS standards
The APA quantitative reporting standards ask for specific things a thesis results section routinely omits. The standards call for the sample size, power and precision to be described, including any power analysis or the method used to determine precision; for inclusion and exclusion criteria, including any post-data-collection exclusion of participants; for the frequency or percentage of missing data, whether it is treated as missing completely at random, missing at random or missing not at random, and the methods actually used to address it; for the results of all inferential tests conducted, including exact p values where null hypothesis significance testing is used; for effect-size estimates and confidence intervals on those estimates; and for a distinction between primary and secondary hypotheses and a discussion of the implications of exploratory analyses. Built as a chapter, that becomes the following order.
- Participants and flow. How many were recruited, screened, excluded and analysed, and why, with the rule for exclusion stated as it was decided before the data were seen. If the study was preregistered — the Center for Open Science defines preregistration as specifying the research plan in advance of the study and submitting it to a registry — cite the registration and say where the analysis departed from it.
- Descriptives and assumptions. Means, standard deviations and correlations for every variable in the models, and the checks the chosen tests depend on, reported in a sentence each rather than a page.
- Confirmatory analyses, hypothesis by hypothesis. One subsection per preregistered or a-priori hypothesis, each giving the test, the exact statistic, the exact p value, the effect size and its confidence interval, and one sentence saying whether the hypothesis was supported. No interpretation.
- Exploratory analyses, labelled. Everything else you ran that the thesis relies on, under a heading that says so, with the same reporting and a note that these analyses generate rather than test hypotheses.
- Tables and figures that carry the numbers. A table per model, a figure per key effect with confidence intervals drawn, and the text pointing at them rather than repeating them.
The study discussion and the general discussion
The study discussion is short and disciplined: what was found against each hypothesis, in language proportionate to the effect size; one or two limitations that belong to this study specifically; and the bridge to the next study, stated as the question this study left open. It does not rehearse the theory. The general discussion is where the doctorate is examined. It integrates the studies into a single account, states the theoretical contribution in the terms of the framework set out in the literature review, takes the strongest alternative explanation seriously before dismissing it, treats limitations as features of the research programme rather than a list, and proposes future work someone could begin. The seven-move structure for a discussion chapter applies with two psychology-specific additions: a paragraph on the robustness of the effects across the studies, with the effect sizes side by side, and an honest statement of what a direct replication would need.
The discussion also has to handle the literature you cite. A finding that rests on a paper later retracted or failed to replicate is a viva question; our analysis of citing retracted papers shows how often it happens and how to check.
The workflow that gets the chapters written
The chapters are slow to write because the evidence lives in three places — the analysis output, the preregistration or analysis plan, and the notes on what each result means — and the writing is an act of reconciliation. The fix is to reconcile once, in a structured document, and write from that.
- Build the hypothesis ledger. One row per hypothesis per study: the hypothesis as preregistered or planned, the test, the statistic, exact p, effect size, confidence interval, supported or not, and a column for exploratory analyses that touch it. In Tesify, open a chapter for each study and paste the ledger at the top; every results subsection is written from a row.
- Draft the results subsections from the ledger. Because each row already holds the five numbers, the subsection is a sentence of method, a sentence of result and a sentence of support. Tesify drafts the connecting prose in your register from the ledger row and the table you paste beneath it; you check every number against the output.
- Write the study discussions against the ledger, not against your hopes. With the effect sizes on screen, the temptation to describe a small, uncertain effect as a finding is visible. The Tesify AI Editor flags overclaiming — significant results described as large, non-significant ones described as trends — so that the language matches the evidence before the supervisor sees it.
- Assemble the general discussion from the three study discussions. Put the effect sizes side by side in one table, write the integration paragraph first, then the contribution, the alternative explanation, the programme-level limitations and the replication paragraph. Our guide to turning notes into a chapter covers the same move for qualitative material.
- Reconcile before submission. Every number in the text is checked against its ledger row; every hypothesis in the introduction is found in the results; every exploratory analysis is labelled. This is the pass that removes the viva question.
The judgement stays yours throughout: which analyses were confirmatory, what the effects mean, how far the account can be pushed. What the workspace removes is the reconciliation between output, plan and notes that costs candidates a term. The hypothesis ledger also does something the field now expects: it makes the chain from preregistration to reported result auditable, which is what a variables and hypotheses table does for a management doctorate and what an examiner in psychology looks for first.
Write your results chapters in Tesify today: one workspace for the hypothesis ledger, the tables and the three empirical chapters, with the AI Editor checking that every claim matches its effect size. The free plan is enough to build the ledger and draft the first results chapter; the data, the analyses and the conclusions remain yours.
Frequently asked questions
How much does Tesify cost for a psychology doctoral researcher?
There is a free plan that is enough to build the hypothesis ledger and draft a results chapter. Paid plans add the capacity for a full thesis with several empirical chapters; pricing is on the site and there is no long contract.
Will my examiners accept results chapters written with an AI tool?
Examiners assess whether the analyses, the reporting and the interpretation are yours and correct. The tool drafts connecting prose from your ledger and checks your language against your effect sizes; it does not run analyses or decide what the results mean. Check your department’s policy on AI assistance and declare use where it asks you to.
Is unpublished participant data safe in Tesify?
Your workspace is private and unpublished thesis content is not used to train models. Participant-level data should stay in the environment your ethics approval names; what goes into the workspace is the hypothesis ledger, summary statistics and the chapter text, which is all the writing needs.
What reporting standard does a UK psychology PhD have to meet?
Most departments expect the APA Journal Article Reporting Standards for quantitative research, which call for power and precision, inclusion and exclusion criteria, missing-data handling, all inferential tests with exact p values, effect sizes with confidence intervals, and a distinction between confirmatory and exploratory analyses. Ethics reporting follows the British Psychological Society’s Code of Human Research Ethics, updated in April 2021.
Do I have to report analyses that did not work?
Yes. The reporting standards ask for the results of all inferential tests conducted, and the replication literature is precisely about what happens when they are omitted. Report them, label exploratory analyses as such, and let the discussion say what they mean.
Should each study have its own discussion as well as a general discussion?
In most UK psychology theses, yes. The study discussion is short and specific to that study; the general discussion integrates across studies and carries the theoretical contribution. A thesis with only a general discussion tends to lose the study-level limitations.
What if my studies did not replicate each other?
Report it and interpret it. A programme in which Study 3 fails to reproduce Study 1’s effect is a legitimate doctoral finding when the results are reported fully and the general discussion explains the candidates for the difference — sample, measure, power — with the effect sizes side by side.
How long should the results chapter of a psychology PhD be?
For a three-study thesis, 3,000–5,000 words of results per study plus tables, 1,500–2,500 words of study discussion per study, and a general discussion of 8,000–12,000 words, within whatever overall word limit your regulations set.