Ask a NET candidate in month one what they are studying and you will hear personality, or social, or cognitive. Ask them about research methodology and statistics and you will usually get a sentence about doing it later. I have watched this happen for years, and later almost always means the last fortnight.
Key terms
Weighting — how many marks a unit has historically carried.
Previous-year analysis — reading past papers for the pattern rather than for the answers.
Why it gets left
Two reasons, and neither is laziness.
The first is that methodology feels like the scaffolding rather than the building. It is what you use to study psychology, so it reads as preliminary.
The second is that it is the only unit where you cannot bluff. You either know why an ANOVA rather than a t-test, or you do not. There is no partially-remembered theorist to lean on.
Why that is the wrong order
Commercial preparation sites that track the paper year on year put research methodology and statistics at roughly 15–20% of the marks — among the heaviest single blocks in the paper. Check that against the papers yourself rather than taking anybody's word for it, mine included; the method for doing so is below.
More to the point: it is the most learnable unit in the syllabus. Personality theory rewards wide reading over a long time. Statistics rewards understanding a decision tree, once. The return on a week is not comparable.
Read the papers, not the predictions
The single most useful thing you can do costs an afternoon and no money.
Take the last ten years of papers. For each question, mark which unit it came from. Do not attempt the questions — you are not revising, you are counting. At the end you will have a weighting table built from evidence rather than from somebody's blog, and you will have noticed which topics recur.
Ten years matters. Three years is short-term noise, and the whole point is to see past it.
What to do first inside the unit
Start with design, not with tests. Almost every statistics question in this paper can be answered by somebody who knows what design produced the data — how many groups, related or independent, what kind of measurement. The test follows from the design. Learn it in that direction and you are reasoning; learn it backwards and you are memorising a list that will not survive the exam hall.
Then reliability and validity, which is the pair candidates reverse under pressure more than any other.
In short
- Research methodology and statistics is reported at roughly 15–20% of the paper. Verify it against the papers yourself.
- It is the most learnable unit per hour spent.
- Count the last ten years of questions by unit. Do not attempt them — count them.
- Learn design first; the test follows from the design.
- Ten years, not three. Three is noise.
The syllabus, pattern and eligibility for each exam — each fact dated and linked to the body that published it — are on the exams pages.