Cambridge AS & A Level Psychology 9990/12 is the May/June 2026 Paper 1 (Approaches, Issues and Debates), variant 2. It lasts 1 hour 30 minutes, is worth 60 marks and tests the AS Level core studies. All ten questions are compulsory, and every one names a core study. The biggest question is Question 9, worth 12 marks, on how Hölzel et al. scanned their participants and how the study compares with Dement and Kleitman, followed by the 10-mark evaluation of Piliavin et al. This guide goes through the paper question by question and shows what the official mark scheme gave marks for. The paper and its mark scheme are free on the Tutopiya Learning Portal.

For an overview of all four 9990 papers from this series, see A Level Psychology 9990 May/June 2026 past papers: what came up and how it was marked.

9990/12 at a glance

The rubric is: answer all questions. You write on the question paper.

Question	Section	Core study and task	Marks
1	A	Hassett et al. (monkey toy preferences): the other toy category, what “play” means, a conclusion	5
2	A	Dement and Kleitman (sleep and dreams): three features of the sample, one weakness of it	5
3	A	Milgram (obedience): ethical ways a business could get staff to follow safety rules	4
4	A	Baron-Cohen et al. (eyes test): an assumption of the cognitive approach, a methodological strength	5
5	A	Fagen et al. (elephant learning): two ethical everyday applications	4
6	A	Saavedra and Silverman (button phobia): consent, minimising harm, confidentiality	6
7	A	Pozzulo et al. (line-ups): the individual and situational explanations debate	4
8	A	Andrade (doodling): the background to the study	5
9	B	Hölzel et al. (mindfulness and brain scans): the scanning (4), a similarity and a difference with Dement and Kleitman (8)	12
10	B	Piliavin et al. (subway Samaritans): two strengths and two weaknesses, one about quantitative data	10
		Total	60

Question-by-question walkthrough

Question 1: Hassett et al., monkey toy preferences (5 marks)

• 1(a) The other toy category (1 mark): plush. Soft, stuffed or “plushies” were accepted, but naming one example of a plush toy was not, and only your first answer counts.
• 1(b) What “play” means (2 marks): a full definition has two parts, that play is voluntary and that it is fun or engaging for the individual. One part earns 1 mark. “Play is playing” earns nothing.
• 1(c) One conclusion (2 marks): the scheme looked for four ideas: hormones (a biological influence), a behavioural preference or bias, a cognitive bias, and social processes. Three or more in your conclusion earned 2 marks; two or fewer earned 1. For example, toy preferences in monkeys reflect hormonally influenced behavioural and cognitive biases that social processes then shape. Results were not credited as a conclusion, “environment” was not credited, and saying the monkeys chose like children earned only 1.

Question 2: Dement and Kleitman, the sample (5 marks)

• 2(a) Three features of the sample other than being adults (3 marks): the list was definitive. Nine participants in total; mostly male (seven men); two women; five studied intensively over 6 to 17 nights; four used only to confirm results over 1 or 2 nights. Opportunity sampling and being told to avoid caffeine and alcohol were also credited. “Four withdrew” and “equal numbers of males and females” were not.
• 2(b) One weakness of the sample (2 marks): 1 mark for the weakness and 1 for an example from the study. A credited answer: the sample lacks generalisability, because so few women took part and women may dream differently or report dream content differently. “Small sample” on its own did not earn the example mark.

Question 3: Milgram, safety rules at work (4 marks)

One mark per ethical suggestion that draws on Milgram. Credited ideas included a manager dressing more authoritatively (such as a technician’s coat), firm verbal prods delivered in a stern tone (“it is very important that everyone follows these rules”), using an authority figure such as a boss, manager or expert, and paying staff a small amount to complete a safety course. Focusing on one specific safety rule was fine.

Suggestions that were not ethical scored nothing: hiring a stooge, cutting pay and physical punishment. Simply copying details from Milgram’s procedure was not credited either; the suggestion had to be for the workplace.

Question 4: Baron-Cohen et al., the eyes test (5 marks)

• 4(a) An assumption of the cognitive approach with an example (3 marks): 2 marks for a full assumption, 1 for a brief one, and 1 for a linked example. A full assumption: information is processed through the same route in all humans (input, process, output), in a similar way to a computer. Another: people differ in cognitive processing such as attention, language, thinking or memory, with one process named. A matching example: participants saw a pair of eyes (input), recognised the emotion (processing) and chose the emotion word (output); two of the three stages had to be right. Simply saying the autism group “did worse” on the test earned no example mark.
• 4(b) A methodological strength (2 marks): 1 mark for the strength and 1 for an example. “Quantitative data” or “standardised” on its own could not earn the first mark; you had to say what it achieves. Credited pairs: standardised so it could be replicated and tested for reliability (everyone saw the same 36 pairs of eyes); quantitative data that made comparisons objective (each participant scored out of 36); and improved validity over the original test that avoided a ceiling effect (four answer options per pair of eyes instead of two). Ethical and application strengths were not credited.

Question 5: Fagen et al., everyday applications (4 marks)

For each application, 1 mark for what it is (clearly based on Fagen et al.) and 1 for how it would be done. For example, teachers helping students learn new skills (what) by using a token system to motivate them to finish work on time (how), or pet owners and vets training animals to accept treatment, such as moving an eye dropper gradually closer with a treat (how). Training service dogs and helping zookeepers carry out procedures with less stress were also credited.

The application needed a specific population or scenario: “people can…” or “animals can…” was not enough. “Secondary positive reinforcement” or “reward” on its own was not a creditworthy how, using rewards to train animals for entertainment was not credited, and the same idea applied to a different group counted once.

Question 6: Saavedra and Silverman, ethics (6 marks)

Each part gave 1 mark for explaining the guideline and 1 for an example from the study.

• 6(a)(i) Consent (2 marks): participants are given enough information before a study to decide whether to take part. Example: both the boy and his mother consented to the therapy and assessments. “Parents gave consent” was not accepted.
• 6(a)(ii) Minimising harm (2 marks): participants should leave in the same state they entered, should not face more harm than in everyday life, and any harm should be reversed. Example: exposure to buttons through imagery was rated as less distressing over time, while behavioural (in vivo) exposure was rated as more distressing. “Exposure therapy” on its own could not earn the example mark.
• 6(b) How confidentiality was followed (2 marks): data should not be identifiable as one participant’s. The example mark needed a direct detail, such as that all we know is that the child was Hispanic and from Florida. “His name was never revealed” could earn the outline mark only.

Question 7: Pozzulo et al., individual and situational explanations (4 marks)

1 mark for outlining each side and 1 for an example of each from the study. The individual side explains behaviour by factors within the person (dispositional); the situational side by factors in the external environment. The scheme’s examples: adults’ identification rate for human faces in target-present line-ups was 0.66, showing that people’s memories differ (individual); children’s correct rejection rate for human faces in target-absent line-ups was 0.45, suggesting social pressure led them to pick a face (situational). Any example showing individual differences in recall, including arguments about children’s or adults’ memory, counted for the individual side. Circular definitions such as “individual factors” were ignored.

Question 8: Andrade, the background (5 marks)

One mark per correct background point. Credited points included: doodling might aid concentration or arousal; daydreaming had been linked to boredom; doing a concurrent task such as doodling had been assumed to impair concentration on the main task; arousal must be maintained to concentrate; the effects of boredom on processing had been overlooked; a secondary task might reduce mind-wandering, and could help if it does not compete for the same cognitive resources (one visual and one auditory task, for example). One mark was also available for defining daydreaming, multi-tasking or doodling.

The trap here: anything about what Andrade did or found was ignored, and so were references to the central executive or the role of daydreaming, because those were discussed after the study.

Question 9: Hölzel et al. and Dement and Kleitman (12 marks)

• 9(a) How and how often participants were scanned (4 marks): up to 3 marks for how and 2 for how often. How: all at the same centre; an MRI scan; a standard head coil; whole-brain data sets; voxel-based morphometry (VBM) to measure grey matter. Only 1 mark in total was available for the general mechanics of MRI, such as lying still in a tube. How often: the MBSR group was scanned two weeks before and two weeks after the programme, and the control group was scanned twice, about two months apart. “Once before and once after MBSR” earned only 1 mark. The scheme also accepted the average gaps between scans (56 days for the MBSR group, 66 days for controls, within 2 days).
• 9(b) One similarity and one difference, not about the sample (8 marks): each half was levelled from 1 to 4.

Level	Similarity or difference (each half)
4	Explained, using both studies as examples
3	Explained with only one study, or both studies used briefly
2	Brief with some use of a study, or explained with no study evidence
1	Brief, with no use of either study

Credited similarities included both being experiments, both collecting quantitative data, both lacking mundane realism and both using scientific machinery. Differences included the brain measurement technique, external validity, ethics and design. The scheme’s model difference: Hölzel et al. used MRI to see changes in grey matter density, while Dement and Kleitman used EEG to record brain wave activity, so one looked at brain structure and the other at real-time function.

Watch the caps: a different aim or different durations scored Level 1; “both biological” with correct assumptions was capped at Level 3, and without them at Level 1; a point not explicitly stated as a similarity or difference was capped at Level 2; and “easy to analyse” does not explain why quantitative data is useful. Anything about the sample or generalising from it scored zero.

Question 10: evaluate Piliavin et al. (10 marks)

You had to give two strengths and two weaknesses, with at least one about quantitative data (the named issue). Strengths the scheme listed: reliability, quantitative data, external validity and the use of observation. Weaknesses: internal validity, the sample and ethics.

The scheme’s detailed example: quantitative data, such as how quickly people helped and how many of each race helped, allowed meaningful comparisons of helping between victim types. For ethics, it listed contexts that count: passengers could not consent; they could not easily withdraw because they were stuck in the carriage; witnessing a collapse could be distressing; the victims were actors (deception); comments were noted down (privacy); and with about 4450 passengers there was no debrief. A point with no context, such as “limited controls so no cause and effect”, was only a brief point.

The essay was marked with five levels:

• Level 5 (9–10): very good evaluation including the named issue, with both strengths and both weaknesses in detail.
• Level 4 (7–8): good evaluation including the named issue, three or four points with most in depth.
• Level 3 (5–6): mostly appropriate evaluation that may not include the named issue; two points in detail, or all four briefly.
• Level 2 (3–4): weak evaluation of a strength or a weakness; brief points, some without context.
• Levels 1 and 0: little or no evaluation, or nothing creditable.

Where the marks are on 9990/12

Split by what each question asks you to do:

Skill	Questions	Marks
Recall study details (name, identify, describe)	1(a), 2(a), 8, 9(a)	13
Outline or explain a concept with an example from the study	1(b), 1(c), 2(b), 4, 6, 7	21
Apply a study to everyday life	3, 5	8
Compare two studies	9(b)	8
Evaluate a study	10	10
Total		60

• Explaining with study evidence is the biggest skill. Twenty-one marks go to outlining a term or guideline and backing it with a detail from the named study, and most of those halve without the example.
• Only 13 marks reward pure recall, and Question 8 shows how strict that recall is: the background, not the procedure.
• The two Section B questions are 22 marks, and both reward explanation and evaluation more than description.

How to self-mark 9990/12 with the mark scheme

1. Learn the layout. A slash or “or” separates ways of making the same point; semicolons or a (1) separate different points. Text in brackets is context for the examiner and is not needed for the mark.
2. Mark the two-part answers in halves. For 2(b), 4(b), 6 and 7, award the identification or definition mark first, then check the example really comes from the named study.
3. Check the “what and how” in Question 5. Strike out applications with no specific group or setting.
4. Cross out off-topic material in Question 8 (anything Andrade did or found) before you count points.
5. Level each half of 9(b) separately. Ask whether you explained the point and whether you used both studies, then apply the caps for sample points and unstated similarities.
6. Find the named issue in Question 10 first. If quantitative data is missing, your mark cannot go above 6. Then count how many of the four points are detailed and in context.
7. Log each lost mark as a missing study detail, a missing example or a missing explanation, and revise that study next.

Sit 9990/12 as a timed mock

The paper allows 90 minutes for 60 marks, one and a half minutes per mark. A workable plan:

Block	Marks	Time
Section A, Questions 1 to 8	38	52 minutes
Question 9(a) and 9(b)	12	17 minutes
Question 10 essay (plan the four points first)	10	16 minutes
Check examples, the named issue and that 9(b) avoids the sample	–	5 minutes
Total	60	90 minutes

Decide before you write Question 10 which point covers quantitative data, and put it first.

Other May/June 2026 9990 papers

• 9990 Paper 22 (Research Methods)
• 9990 Paper 32 (Specialist Options: Approaches, Issues and Debates)
• 9990 Paper 42 (Specialist Options: Application and Research Methods)

Practise and download

• Download 9990/12 and its mark scheme on the Learning Portal →
• All 9990 May/June 2026 papers: overview and marking patterns
• Try a free A Level Psychology quiz

Need help with a 9990 topic? Book a free trial lesson with an A Level Psychology tutor →

Frequently asked questions

Which core studies came up on 9990 Paper 12 May/June 2026? +

Ten: Hassett et al. (monkey toy preferences), Dement and Kleitman (sleep and dreams), Milgram (obedience), Baron-Cohen et al. (eyes test), Fagen et al. (elephant learning), Saavedra and Silverman (button phobia), Pozzulo et al. (line-ups), Andrade (doodling), Hölzel et al. (mindfulness and brain scans) and Piliavin et al. (subway Samaritans).

How many marks is 9990/12 and how long is it? +

It is out of 60 marks and lasts 1 hour 30 minutes. All ten questions are compulsory: eight short questions in Section A worth 38 marks and two longer questions in Section B worth 22.

How was the 10-mark Piliavin et al. essay on 9990/12 marked? +

With a five-level grid. Level 5 (9 to 10 marks) needed very good evaluation that included the named issue, quantitative data, and covered both strengths and both weaknesses in detail. Level 4 also needed the named issue, so an answer without it could not go above Level 3 (6 marks).

How was the similarity and difference question on 9990/12 marked? +

Question 9(b) gave up to 4 marks for the similarity and 4 for the difference, each levelled separately. Level 4 needed the point explained using both Hölzel et al. and Dement and Kleitman. Description with no explanation was capped at Level 3, and anything about the sample scored nothing.

Why did some answers to 9990/12 Question 8 score zero? +

Question 8 asked only for the background to Andrade's doodling study. The mark scheme ignored anything about what Andrade did or found, so answers that described the procedure or results earned no marks for that material.

Where can I download 9990/12 May/June 2026 and its mark scheme? +

Both are free to download and print on the Tutopiya Learning Portal, on the Cambridge AS & A Level Psychology 9990 past papers page.

Ready to Excel in Your Studies?

Get personalised help from Tutopiya's expert tutors. Whether it's IGCSE, IB, A-Levels, or any other curriculum — we match you with the perfect tutor and your first session is free.

Book Your Free Trial