Your Chapter 4 is currently forty pages of pasted SPSS tables with a sentence between them, and you know it is not a results chapter. Your supervisor has said “this needs to be written up properly” without explaining what that means. Submission is close. This page is the missing instruction: what a results chapter actually contains, and how to get there from the output you already have.
Rebuild Chapter 4 in Tesify — structure first, then the prose →
What this costs if you leave it
A results chapter made of raw output is one of the fastest ways to trigger a corrections round. In South Africa that is expensive in a specific way: your work is assessed in writing by independent external examiners with no oral in which to explain yourself, as our guide to external examination sets out. Corrections arrive months after submission, when your momentum is coldest, and if the round pushes you past the academic year you pay another year’s registration fees. The rebuild below is two or three evenings. The corrections round it prevents is months.
Step 1: Decide what the chapter is actually for
A results chapter reports what you found. It does not explain why, compare with literature, or draw implications — that is the discussion. Getting this boundary right immediately halves your problem, because a great deal of what students write into Chapter 4 belongs in Chapter 5 and can simply be moved rather than deleted.
The test for a sentence: if it contains “because”, “this suggests”, “consistent with previous studies” or “this may be due to”, it is discussion. If it states what the data showed, it is results. Some faculties combine the two chapters, particularly in qualitative work — check your own faculty guide and two recently passed dissertations in your department before you restructure anything.
Step 2: Choose which tables survive
SPSS produces a table for everything, including intermediate steps nobody needs to see. Most of it goes. Keep three categories:
- The sample description — one table of demographic frequencies and percentages.
- The descriptive statistics for your key variables — means, standard deviations, and scale reliabilities.
- One table per research question or hypothesis, containing the test result that answers it.
Everything else — assumption-check output, full correlation matrices you only mention in passing, iteration histories, case-processing summaries — goes to an appendix or nowhere. A tight results chapter for a master’s dissertation typically has a handful of tables, not thirty. If a table is not referred to in your text, it should not be in the chapter.
Step 3: Rebuild every surviving table by hand
Do not paste. Retype the numbers you need into a properly formatted table in your document. This is tedious and it is not optional — pasted SPSS tables carry the wrong fonts, the wrong rules, columns nobody needs and labels like VAR00007. The conventions most South African faculties expect, following APA-style practice:
- A table number, then a brief italic title on the line below it.
- Horizontal rules only — above and below the header row and at the foot. No vertical lines, no shading, no grid.
- Consistent decimal places within a column, and readable variable names rather than SPSS codes.
- A note beneath the table defining abbreviations and giving the sample size.
- Every table referred to in the text by number — “as shown in Table 4.3” — never as “the table below”.
Retyping has a second benefit students discover halfway through: it forces you to look at every number, and it is where transcription errors and impossible values get caught before an examiner catches them.
Step 4: Report each test correctly in running text
A table alone is not a finding. Each test needs a sentence in the text that states the result in the conventional format for that test. The standard shapes:
- Independent-samples t-test: group means and standard deviations, then t(degrees of freedom) = value, p = value, with an effect size such as Cohen’s d.
- One-way ANOVA: F(between df, within df) = value, p = value, with eta squared, followed by which post-hoc comparisons were significant.
- Correlation: r(df) = value, p = value.
- Chi-square: χ²(df, N = sample size) = value, p = value.
- Regression: the model fit and its significance, then each predictor’s coefficient with its significance.
Three formatting conventions catch almost everybody. Statistical symbols are italicised; the numbers are not. Values that cannot exceed one — p and r — conventionally drop the leading zero. And very small probabilities are written as p < .001 rather than as p = .000, which is what SPSS displays and which is never literally true.
The decimal comma trap — this one is specifically South African
South African English uses the decimal comma, and official sources write it that way: Stats SA publishes “85,6% of households”, not “85.6%”. Your statistical software, however, will almost certainly output a decimal point, and the APA-style conventions your faculty guide points to are written for a decimal-point convention.
The result is a chapter that mixes both, sometimes within a single table, and it looks careless. Decide which convention your faculty guide requires, apply it to every number in the dissertation, and — this is the part people miss — when you quote a figure from a South African source such as Stats SA or the CHE, keep it in whichever convention you have chosen and stay consistent rather than reproducing the source’s punctuation inside your own sentence. If your guide is silent, follow the convention used in recently passed dissertations from your department.
Step 5: Structure the chapter around your research questions
Order the chapter the way you ordered your questions, not the way you happened to run the analyses. The reliable structure:
- A short introduction saying what the chapter covers and in what order.
- Sample description.
- Preliminary analyses — data screening, missing data, assumption checks and scale reliabilities, briefly. Where an assumption failed and you changed test, say so here; our guide on what to do when your data fails the normality test covers how to word that.
- One section per research question or hypothesis, each with the relevant table and its reporting sentences, each ending with a plain statement of whether the hypothesis was supported.
- A short summary of findings that hands over to the discussion.
Write a signposting sentence at the head of each section. An examiner reading cold should never have to work out which question a section answers.
How to do the rebuild in three evenings
Evening one: build the skeleton. Create the section headings from your research questions, and under each one write only the table number and a placeholder sentence for the finding. You now have a chapter shaped correctly and entirely empty, which is a far better position than forty pages shaped wrongly.
Evening two: build the tables. Work through the skeleton, retyping each surviving table to standard. Move everything else to an appendix as you go.
Evening three: write the prose. Each section gets its reporting sentence, its reference to the table, and its plain answer to the question — while moving every interpretive sentence you find into a notes file for Chapter 5, where it becomes free material rather than wasted work.
That sequence is exactly what Tesify is built to hold: the chapter structure lives in the project, so each evening starts on a plan rather than on a blank screen, and the reference list keeps pace with the writing instead of becoming a submission-week rescue. You can start free and see whether the workflow fits. And if the underlying problem is that your evenings are already full, our system for writing a dissertation while working full-time is the other half of it.
Objections, answered honestly
What does Tesify cost in rand?
Pricing is shown at sign-up and charged in your card currency. Start free and see the plans before paying anything — and distrust any page quoting a rand figure that may have aged, including this one, which is why we do not print one.
Is it a plagiarism risk to write my chapter with an AI tool?
Reporting your own results in your own words, with editing support, is not what integrity policies target; submitting text you did not write is. South African universities regulate disclosure rather than prohibiting assistance — our guide to AI policies at South African universities sets out what UCT, Wits, Stellenbosch and Unisa each require. Declare your use and tell your supervisor.
What happens to my data?
Your project remains your own work. Read the privacy policy at sign-up as you would for any tool touching unpublished research — and note that nothing here requires uploading your dataset. You are writing about numbers you have already computed, so raw participant data can stay where your ethics clearance says it must.
My supervisor has not replied in weeks. Should I wait before rebuilding?
No. The structure above is standard and will not be wrong in any faculty. Rebuild, then send the improved chapter — a specific document gets a faster response than a general question. If silence persists, our escalation guide sets out the order to raise it in.
FAQ
Can I paste SPSS tables into my dissertation?
No. SPSS output is a working format. Retype the numbers you need into properly formatted tables that follow your faculty’s conventions.
How many tables should a results chapter have?
Typically a sample-description table, a descriptives table and one per research question. If a table is never mentioned in the text, move it to an appendix.
Should I interpret my findings in Chapter 4?
Not in the conventional five-chapter structure — interpretation belongs in the discussion. Some faculties, and much qualitative work, combine them; check your own guide.
Do I report non-significant results?
Yes, always, in the same detail as significant ones. Reporting only what reached significance misrepresents your study and is a serious integrity problem.
Why does SPSS show p = .000?
Because it rounds to three decimals. A probability is never exactly zero, so write p < .001 instead.
Do I include assumption-check output?
Report the checks briefly in the text and put the detailed output in an appendix. What matters is that you checked, what you found, and what you did about it.
Should I use decimal commas or decimal points?
Whichever your faculty guide requires — but apply it to every number in the dissertation. South African English uses the comma and official statistics follow it; APA-style guidance assumes the point. Mixing the two is the visible error.
How long should a results chapter be?
Shorter than students expect. It is the most economical chapter in the dissertation, because interpretation lives elsewhere. Length is set by the number of research questions, not by page targets.
Can I put charts in as well as tables?
Yes, where a chart genuinely shows something a table cannot. Do not present the same data both ways, and rebuild charts to your faculty’s standard rather than pasting the software’s default styling.