Most of the output is not for you
Statistical software prints everything it computed, because it does not know which analysis you are writing up. Your assignment needs a small subset of that, and pasting the whole block in does not read as thorough. It reads as though you could not tell which part mattered.
For nearly every test at coursework level, the write-up needs four things: what test was run and on what, whether the assumptions held, the test statistic with its degrees of freedom and its p value, and an effect size. Everything else exists for the analyst rather than for the reader.
So the discipline is to identify those four before you look at anything else, then rebuild them as a table in whatever style manual your course uses. Software output pasted directly is almost never formatted the way your manual requires, and that is a straightforward loss of marks for a fixable reason.
What the numbers mean, in sentences
Here is the plain-language version of the things you will meet most often. Say these out loud until they sound obvious, because the moment they do, writing the results section becomes mechanical.
- p value: how surprising your result would be if there were genuinely no effect. Small means surprising, not important.
- Significant: your result was surprising enough to report under the threshold you set. It does not mean large, and it does not mean it matters clinically.
- Effect size: how big the difference actually is. This is the number that answers so what, and rubrics increasingly ask for it.
- Confidence interval: the range of values your data is compatible with. A wide one means you learned less than the point estimate suggests.
- Degrees of freedom: bookkeeping that reflects sample and design. Report it, do not interpret it.
- Assumption tests: checks on whether the test was appropriate at all. A failure here matters more than any result underneath it.
Significance is not importance
This is the misunderstanding that produces the most confidently wrong sentences in student papers. A significant result means you found something unlikely to be chance. It says nothing about whether the difference is big enough for anybody to care.
With a large sample, tiny differences become statistically significant routinely. A two-point difference on a hundred-point scale can be highly significant and completely irrelevant to practice. That is exactly why effect sizes matter, and why a good results section reports both and then says what the combination means.
The other direction catches people too. A non-significant result is not evidence that there is no difference. It means your study did not detect one, which may be because there is nothing there or because your sample was too small to see it. Writing that honestly is more impressive than overclaiming, and graders notice.
When an assumption fails
It happens constantly with real data and it is not a disaster. What is a disaster is proceeding silently as though the check passed, because anybody reading carefully will notice, and it undermines everything that follows.
There are three honest responses and all of them are acceptable. Use a test that does not require the assumption, which usually means a non-parametric alternative. Transform the data, and say that you did. Or proceed while stating the violation as a limitation and explaining why the test is still defensible.
Any of those earns marks. Ignoring it loses them. Rubrics at graduate level increasingly award credit specifically for identifying and handling assumption violations, which means a failed check is an opportunity rather than a problem.
Write the justification before you run anything
The choice of test is graded, and it is graded more heavily than the execution. So write down, in ordinary language, why this test and not another: how many groups, whether they are independent or related, what kind of measurement you have, and what the assumptions require.
That paragraph lands directly in your methods section, generally earns marks on its own, and leaves you able to respond when somebody in a seminar asks why an alternative was rejected. Having no answer to that question damages you considerably more than picking a marginally imperfect test would have.
It also protects you from the most common failure, which is running the test you happen to know rather than the one your design requires. If the whole area is a fog rather than a detail, having somebody walk through one worked example with you is usually cheaper and more durable than having the assignment produced.
Questions people actually ask
Do I paste the SPSS output straight into my paper?
No, and it costs marks nearly every time. Software output isn't formatted to any style manual, and it contains far more than your write-up needs. Rebuild the small number of values your analysis requires as a properly formatted table, in the style your course uses, and leave the rest out entirely.
What does p equals .06 mean? Did I fail?
It means your result didn't cross the threshold you set in advance, and that's a finding rather than a failure. Report it accurately, note that it approached the conventional threshold if you like, and discuss what might explain it, including sample size. Reframing it afterwards as significant is the version that gets marked down.
Why does my rubric want an effect size?
Because a p value only tells you a result was surprising, not whether it was large enough to matter. With a big sample, trivial differences become significant routinely. Effect size answers the question your reader actually has, which is whether anybody in practice should change what they do, and graduate rubrics increasingly require it.
My assumption test failed. What now?
Choose one of three honest routes: switch to a test that doesn't require the assumption, transform the data and say so, or proceed while stating the violation as a limitation with a reason. All three earn marks. Only silence loses them, and at graduate level handling a violation well is often specifically rewarded.
