IB Psychology Paper 3 2027: Complete HL Guide

Paper 3 is 1 hour 45 minutes, 30 marks, and 30% of your HL grade. It is the only paper where you cannot prepare the content in advance, which is exactly why preparing the background knowledge matters so much.

You get a resource booklet you have never seen. Several sources, all drawn from one HL extension inside one context. A mix of quantitative and qualitative, experimental and non-experimental. The data may be real or written for the exam.

Then four questions.

Question	Marks	What it asks	Approximate time
1	3	Explain one issue that limits what can be read from the data in a named source	10 minutes
2	6	Analyse the findings from a named source and state a conclusion linked to the claim	35 minutes for questions 2 and 3 together
3	6	Discuss credibility, bias, or how far findings transfer
4	15	Judge how far a stated claim holds, using at least three sources and your own understanding of the extension	55 minutes

That is one suggested plan rather than an official allocation. Adapt it to your writing speed. The one thing worth protecting is the time for question 4.

What the booklet will contain

All the sources tie to one extension in one context. If source 1 is about technology and learning, all of them will be.

Common, though not promised: one or two graphs, one source with reported statistics often in a table, at least one qualitative source often using thematic analysis with quotes, and sometimes a survey.

Suggested order on the day: read the claim first, then the booklet, then answer questions 1 to 3, then question 4.

Questions 1 to 3 and question 4 are different skills

Questions 1 to 3 test whether you can read one source closely. Question 4 tests whether you can hold several sources at once and make a judgement.

That is a genuinely separate skill, and practising questions 1 to 3 harder will not develop it. Both need their own preparation.

Question 4 is judged on three things at once

Half the paper sits here, and all three of these have to be visible in your answer.

Your own understanding of the HL extension, used to interpret the evidence and to say how far the claim holds. This is assessed in its own right. An answer that analyses the sources beautifully but shows no background psychology cannot reach the top band, no matter how sharp the methodology is.

Critical analysis of at least three sources, argued from their design, measures and samples. Three is not a suggestion. With fewer than three interpreted meaningfully, 10 to 12 becomes your ceiling.

Evaluation of competing points of view, ending in a reasoned conclusion. The requirement here has two verbs: identify a different view, then evaluate it. Most answers state an opposing view and move on, which meets half the requirement.

The mid-band trap

Most answers that stall at 10 to 12 do the same thing. They walk through source 1, then source 2, then source 3, describing each one and evaluating its method. It reads as competent, and it is really a summary.

The top band asks you to argue to a conclusion, using the sources as evidence inside your argument rather than as the structure of it. Your paragraphs should be organised around claims about the psychology, not around the order the sources appear in the booklet.

Three sources must be interpreted meaningfully. They do not need three separate paragraphs.

The structure for question 4

Unpack the claim, build your evidence paragraphs, step back to the evidence base, weigh a rival view, then reach a verdict.

Inside each evidence paragraph, four moves.

Source evidence. Something specific, from one source or several. Theory. What your own knowledge says it means. Evaluate. What that evidence is actually worth. Point. What it does to the claim.

That last move is the one people leave out, and it is what keeps every paragraph anchored to the question.

Where your own knowledge actually does work

Background knowledge is not there to be recited. It changes what can legitimately be concluded from a source you have never seen, in four ways.

It supplies a mechanism: the source shows an effect, and you explain why that effect would happen. It supplies a moderator: the source shows an average, and you know a variable that would make the effect stronger for some people and weaker for others, which narrows the claim. It supplies an alternative explanation: the source shows a difference, and you know another route to that same difference. Or it challenges the measurement: the source measured something and called it a construct.

If your knowledge sits in its own paragraph without touching a source, it is not interpreting anything.

The verdict

A flat yes or no is almost always the weakest answer available, because real evidence rarely lines up cleanly.

Note the wording of the question. How far the claim holds, not whether. A conclusion saying the claim holds under these conditions, for these people, on this outcome, but not in the strong form it was written in, is usually more defensible than a straight verdict either way.

A strong verdict is reasoned, stated plainly, and matches the argument you actually made. The most common failure is inconsistency: four paragraphs showing the evidence is weak, then a conclusion that the claim is well supported.

Nothing should appear in your conclusion for the first time.

The errors that cost the most

Using only two sources. Caps you at 12, regardless of everything else.
Own knowledge in its own paragraph, touching nothing.
Identifying an opposing view without evaluating it.
A conclusion that contradicts the paragraphs.
Treating each source as an isolated mini-essay with no relationship between them.
Summarising every source in the booklet. Three used properly beats all of them described.
A long introduction. There are no marks for defining terms or scene-setting.
Accepting a source's own conclusion. Researchers overreach, and spotting it is worth real marks.

Where to go next

If you want the full paper worked through with graded exemplars at each band, that is what the PsychExcel Paper 3 Complete Guide is built for.