Cognitive Approach

0. Chapter at a glance

Core study	Psychology investigated	One-line idea
Andrade (2010) — doodling	attention & memory	doodling can improve concentration on, and memory for, a (boring) conversation
Baron-Cohen et al. (2001) — Eyes test	theory of mind & social sensitivity	revised “Reading the Mind in the Eyes” test shows adults with Asperger/HFA recognise emotions less well
Pozzulo et al. (2011) — line-ups	false positive responses & eyewitness testimony	children make more errors than adults when the culprit is absent — due to social, not cognitive, factors

1. What is the Cognitive Approach?

Cognitive psychology = what we do with information: how it gets in (senses, thoughts), what we do with it once inside, and how we get it back out.

Memory processes: encoding (convert sensory info so it can be stored) → storage (the brain’s record of the event) → retrieval (getting it back, by recognition or recall).

Studies attention, memory, decision-making, language development — and applies this, e.g. how cognition affects behaviour, or how accurate memory is when giving evidence about a crime.

Two main assumptions

① Human vs computer information processing

Cognitive psychologists study the events between detecting a stimulus and making a response — compared to a computer:

	People	Computers
INPUT	Eyes (light, colour, movement)	Camera, scanner
	Ears (sound)	Microphone
	Skin (pressure)	Touch pad, keyboard, mouse
PROCESSING	Thinking & decision-making in the brain	CPU — control centre
	Short-term memory	RAM — fast, temporary, lost if computer crashes
	Long-term memory	Hard drive — “permanent” memory
OUTPUT	Voice	Speakers
	Body (hands writing, drawing, moving)	Screen, printer

Similarities: we take info in (encode) in different ways · have more than one memory store · produce outputs (movements, sounds). Difference: computers are less likely to forget but less good at guessing.

② Individual differences in cognitive processes

Psychologists make generalisations about people but are also interested in individual differences = ways people are dissimilar. Computers vary too (processing capacity, memory size, interactive screens, “skill learning”).

• Direct cognitive differences: information-processing speed differs → quicker reading / arithmetic.
• Effects on behaviour: some people speak faster; keen sportsmen & women react faster than less sporty people.
• Effects on emotion: stimuli such as noises make some people grumpy.

Key words — approach basics

2. Core Study 1 — Andrade (doodling)

The psychology being investigated — attention & memory

Doodling = absent-mindedly sketching patterns/shapes while thinking about something else. Teachers tell learners off for it — is the complaint justified?

Doodling uses some of our information-processing capacity, even though it feels “unconscious”. It could take cognitive resources away from the primary task by dividing attention → worse performance.

Whether we attend to a stimulus decides whether we remember it. Attention is required for the encoding stage of memory.

Theory ① — doodling HURTS (divided attention)

Divided attention is easier when tasks are simple, well-practised & automatic. e.g. an inexperienced driver stops talking when the road gets difficult. Driving = primary task; talking = concurrent task (happens at the same time).

So doodling should worsen performance — but if both tasks are highly practised/automatic it may not.

Theory ② — doodling HELPS (focused attention / arousal)

• Doodling may improve performance by raising arousal and enhancing focused attention (Do & Schallert, 2004) → reduces daydreaming so you stay focused.
• Daydreaming uses important cognitive resources (the central executive) → inhibits tasks needing attention & memory. Bored + daydreaming = poorer performance.
• Doodling may help maintain arousal (Wilson & Korn, 2007) — like giving yourself something physical to do while you think: raises arousal if sleepy, lowers it if agitated/bored → better primary-task performance.

Based on the working memory model (short-term memory): a primary listening task uses auditory memory; a concurrent visual task (doodling) uses visuo-spatial memory → the two do not directly compete.

Background & Aim

Andrade’s definition: doodling = sketching patterns & figures unrelated to the primary task.

She used several strategies to increase boredom → creates a situation where doodling could affect attention “for better or worse”. Her expectation: doodling would interfere less with overall processing than devoting more central-executive function to daydreaming → attention improves → recall increases.

Primary task (listening to a message) = auditory; doodling = visuo-spatial → not in direct competition for one memory resource.

Method

Design: laboratory experiment (not a normal place to take phone messages; controlled). Independent measures — participants in either the control group or the doodling group.

Sample: 40 members of a participant panel at the UK Medical Research Council unit for cognitive research · general population · aged 18–55 · paid a small sum · 20 per group, mainly females (2 males control, 3 males doodling). One doodling participant didn’t doodle → replaced.

They were asked to join the study immediately after another unrelated study for a different researcher → intended to enhance boredom.

Procedure

	Control group	Doodling group
Response sheet	lined paper (could doodle if wanted, but no instructions about doodling)	A4 sheet, alternating rows of squares and circles (10 per row) + wide margin on left to write names
Instruction	write names of party-goers	shade the shapes with a pencil — “just something to relieve the boredom”; neatness/speed doesn’t matter; also write names

• All listened to a monotonous mock phone call about a birthday party (2.5 min, monotonous voice, 227 words/min).
• Tape had 8 names of people attending, 3 names + 1 cat who could not attend, 8 place names, and irrelevant details.
• Standardised instructions: write down names of people who will definitely/probably come (excluding yourself); ignore names of those who can’t come; do not write anything else.
• Listened at a comfortable volume. Then sheets collected; experimenter talked to them for 1 min incl. an apology for misleading them about the memory test.

Operationalising the DV (recall): plausible mishearings (e.g. “Greg” for “Craig”) = correct. Names on tape but not party-goers (e.g. John) = false alarms. Completely new names generated = false alarms. Other words about people (e.g. “sister”) = ignored.

Finally: surprise test (places then people, or vice versa) → debrief + asked whether they suspected a memory test.

Results

Doodling group shaded a mean of 36.3 shapes (large individual range — up to 110). No control participant doodled spontaneously.

Task	Measure	Control	Doodling
Names (monitored)	correct	4.3	5.3
	false alarms	0.4	0.3
	memory score	4.0	5.1
Places (incidental)	correct	2.1	2.6
	false alarms	0.3	0.3
	memory score	1.8	2.4

• Control: correctly recalled mean 7.1 (SD 1.1) of the party-goers’ names; 5 people made a false alarm.
• Doodling: mean 7.8 (SD 0.4) party-goers’ names; 1 person made a false alarm.
• Overall doodlers recalled a mean of 7.5 names + places vs 5.8 for controls = 29% more.
• Demand characteristics check: 3 doodlers + 4 controls suspected a memory test; none said they actively tried to remember. Doodlers still better even when these were excluded.

Conclusion

Doodling helps concentration on a primary task: the experimental participants, for whom doodling was a concurrent task, performed better than controls who just listened.

Because doodlers did better on both monitored & incidental info, there are two possible explanations:

1. Attention effect — doodlers noticed more target words. Doodling ↑ arousal → ↓ daydreaming → maintained attention on the primary task.
2. Direct memory effect — doodling improved memory directly, e.g. encouraging deeper information processing.

Strengths & weaknesses

✓ Strengths	✗ Weaknesses
Lab experiment, independent measures → limits uncontrolled variables (comfortable volume; recorded message → target words said the same way every time).	Participant variables may confound: number of shapes shaded differed (0–110) between individuals. Independent measures → individual differences matter more.
Standardised boredom (monotone tape, dull quiet room, they expected to go home) → all equally likely to daydream → ↑ validity + reliability.	Sample bias: aged 18–55 so representative of adult ages BUT all volunteers from a recruitment panel (time to spare, interest in psychology) → lowers validity.
Operationalisation of doodling via shading sheets → otherwise individuals would doodle differently / not at all → ↑ validity.	Demand characteristics: some suspected a memory test — but roughly equal per condition, didn’t try to remember, removing them didn’t alter the conclusion → unlikely to reduce validity.
Natural doodling encouraged: lined paper discouraged control doodling (none doodled); only 1 doodler replaced → effective.	No self-report of daydreaming → can’t tell attention vs memory explanation. (Only ask whether they daydreamed, not what about — that would invade privacy.)
Quantitative data (numbers of names/places) = objective record of memory.	Lab = low ecological validity (not a normal place to take messages).

Ethical issues

• No fully informed consent — participants couldn’t know about the unexpected place-names test.
• Unexpected test could distress them if they couldn’t remember → possible breach of protection from harm.
• Mitigation: debrief + apology for misleading them. A debrief is not a substitute for good ethical procedure.

3. Core Study 2 — Baron-Cohen et al. (Eyes test)

The psychology being investigated — theory of mind & social sensitivity

ASD (autism spectrum disorder) occurs in ≈ 1% of the population. People with ASD share difficulties in social sensitivity, communication, coping with change, and may have narrow interests. ASD now covers both HFA + Asperger syndrome (AS) — but HFA and AS were the terms used when this study was published.

Baron-Cohen et al. suggest people with ASD lack / have an underdeveloped cognitive process called a theory of mind = ability to realise others have different feelings, beliefs, knowledge and desires from your own. They find it hard to understand that others have their own plans, thoughts, points of view.

Most theory-of-mind tests are designed for children. Baron-Cohen et al. (1997) developed the “Reading the Mind in the Eyes” task for adults — identify the emotional state others are in. It tests the first stage of theory of mind (assigning an appropriate mental state to another person) → a good measure of social cognition. If limited → social sensitivity reduced.

Background

The 1997 first version: participants saw photos of eyes and chose the emotion from 2 options. Adults with ASD correctly identified fewer emotions than the non-ASD group — but there were practical problems, solved in the revised version:

Problems with ORIGINAL Eyes test	Solutions in REVISED Eyes test
Forced choice with only 2 options, always semantic opposites (e.g. sympathetic / unsympathetic) → task too easy	Options ↑ to 4; not opposites
Only 25 sets of eyes → many ASD participants scored 24/25 → ceiling effect	36 sets of eyes in the final analysis
Used both basic + complex emotions; basic ones too easy	Only complex emotions
Emotion in some photos could be solved from direction of gaze (e.g. ignoring)	Those sets of eyes deleted
Imbalance of male and female faces	Equal number of male + female faces
Participants might not have understood the words	Given a glossary

Aim & hypotheses

5 hypotheses:

1. ASD participants will score significantly lower on the revised Eyes test than the control group.
2. ASD participants will score significantly higher on the AQ than controls.
3. Females in the “normal” groups (2 & 3) will score higher on the Eyes test than males.
4. Males in the “normal” group (3) will score higher on the AQ than females.
5. AQ and Eyes-test scores will be negatively correlated.

Method

Design: laboratory experiment (testing environment not like everyday life), specifically a quasi-experiment: participants not randomly assigned — they’re allocated by a given characteristic (having ASD or not). Independent groups design: Group 1 (ASD) vs control groups 2, 3 & 4.

IV = type of participant (ASD vs control). DV (2 key measures): score on the revised Eyes test; and IQ (for ASD + IQ-matched groups) measured with the short WAIS-R (block design, vocabulary, similarities, picture completion).

Sample — four groups

Group	N	Average	Who / how recruited
1: AS/HFA	15	age 29.7; IQ 115	all males; self-selecting via adverts in Autistic Society magazine + support groups; diagnosed in specialist centres with DSM or ICD criteria
2: Adult comparison	122	age 46.5	“normal” adults, no AS/HFA diagnosis; from adult community + education classes in Exeter and public library users in (UK); males + females
3: Student comparison	103	age 20.8	“normal” students, University of (highly selective → not representative of general population); males + females
4: IQ-matched	14	age 28; IQ 116	randomly selected from general population; IQ matched to Group 1

Groups 2, 3 and 4 are control groups.

Procedure

1. Making the revised test: started with 40 sets of eyes, target + foil words chosen by the first two authors (Baron-Cohen & Wheelwright). Each set had 1 target word + 3 foils, developed using groups of 8 judges.
2. Judges’ rule: at least 5 of the 8 had to agree the target was the most appropriate. If more than 2 chose a foil → new target word / new foils / both, and retested until criterion met.
3. After combining Groups 2 + 3, 4 items gave inconsistent results (controls often chose foils) → removed → 36 sets (18 male, 18 female faces), each with 4 emotion choices (e.g. target reflective; foils aghast, irritated, impatient).
4. Each participant first read a glossary of the emotion words (could refer back whenever needed) → shown a practice item → then the sets of eyes: pick the correct word from 4.
5. Eyes test given to all (with a researcher, quiet room, Exeter or ). Gender recognition test given only to Group 1 as a control task.
6. AQ completed at home and returned by post by all except Group 2.

Results

• Eyes test: AS/HFA correctly identified significantly fewer target words than all 3 comparison groups. No participant checked the glossary for more than 2 definitions.
• Groups 2 + 3: sex differences in Eyes scores were apparent but not significant.
• Gender recognition: all AS/HFA scored 33 or above out of 36.
• AQ: AS/HFA scored significantly higher than Student + IQ-matched groups. Smaller but significant male–female AQ difference in the Student group.
• Correlation: significant negative correlation AQ × Eyes (−0.53); no correlation IQ × Eyes.

Group	N	Eyes test mean (SD)	AQ mean (SD)
1 AS/HFA	15	21.9 (6.6)	34.4 (6.0) [N=14]
2 Adult comparison	122	26.2 (3.6)	—
3 Student comparison	103	28.0 (3.5)	18.3 (6.6) [N=79]
4 IQ matched	14	30.9 (3.0)	18.9 (2.9)

* Not all participants returned the AQ — N in square brackets = number who did.

Conclusion

• The AS/HFA group had a specific deficit in a cognitive process that should help identify emotions in others — i.e. theory of mind.
• Validated by the AS/HFA group identifying gender from the eyes accurately → the Eyes test detected a subtle, specific impairment in otherwise intelligent participants.
• Sex difference in comparison groups: males had lower Eyes scores + more autistic traits (higher AQ) than females — a pattern similar to ASD but less marked. Some differences weren’t significant → more research needed.
• The revised Eyes test is a more sensitive measure of adult social intelligence than earlier versions → lets future research discriminate individual differences more meaningfully.

Strengths & weaknesses

✓ Strengths	✗ Weaknesses
Lab experiment, standardised administration by a researcher → confounding variables controlled (quiet room → fewer distractions; everyone read the glossary first) → ↑ internal validity, and can be repeated → check reliability.	Quasi-experiment → couldn’t randomly allocate. A factor other than ASD that differs between groups could be a confounding variable. (Partly resolved: 2 control groups + IQ-matched group.)
Validity ↑ vs previous versions: changes took normal performance below the ceiling.	Ecological validity: real eyes aren’t static or isolated from the rest of the face → can’t apply results straight to everyday life. Improve: videos + whole faces.
Gender recognition control task + all images standardised (size, colour — B&W) → valid.	Small sample: only 15 AS/HFA (all male, self-selected) → may not represent everyone with AS/HFA → caution generalising. Also still still images (emotions are usually detected on live, moving faces).

Ethical issues

• All could give informed consent; data confidential → few ethical issues in the study itself.
• Possible issue from findings: identifying “normal” controls vs “impaired” ASD performance could portray neurodiverse groups negatively.
• Ethical strength: identifying differences gives greater understanding of ASD experience → potential improvements in society, e.g. school or workplace.

4. Core Study 3 — Pozzulo et al. (line-ups)

The psychology being investigated — false positives & eyewitness testimony

Human memory is like a computer’s in some ways — but a computer reproduces an error-free copy unless it crashes or gets a virus. Human memory can be distorted by other information we’re exposed to during and after encoding. Usually this doesn’t matter — but when memory is key evidence in a criminal trial, it’s vital.

False memories = post-event information from an eyewitness; the witness confidently believes it. Even a question containing extra information can distort recall (Loftus & Palmer: video of a car accident; adults estimated faster speeds when asked “smashed into each other” than “hit each other”).

Children can be witnesses too — sometimes the only witness → the legal system relies on eyewitness testimony. But are children reliable witnesses?

Background

• Cognitive effects, e.g. post-event information and the way questions are asked, cause errors in children’s decision-making.
• Pozzulo & Lindsay (1997): children were less likely than adults to say “I don’t know”, even when they knew they could.
• Pozzulo & Lindsay (1998): when the culprit is not in the line-up, children are more likely than adults to identify an innocent person = false positive response.
• Line-ups can create mistakes (→ miscarriages of justice): a witness may be more likely to make errors if asked to decide from a line-up and is then shown another line-up containing the same suspect.

Social factors that may push children into incorrect decisions:

1. Asked to pick the culprit → child thinks they must choose someone; saying “not there” seems like a “non-response” (“If I’m given a selection I must make a choice”).
2. The adult asking = an authority figure (esp. a police officer) → compliance.
3. Children feel more pressured than adults; want to comply; may even fear getting into trouble if they don’t choose.

Aim & predictions

Social factors may have a greater effect on child witnesses than adults. Two eyewitness tests: identification (pick the culprit from a target-present line-up) and rejection (correctly reject the faces when the culprit is absent = target-absent line-up).

To separate social vs cognitive effects, Pozzulo et al. needed to minimise cognitive effects that could impair children’s decisions → they compared identification/rejection of a cartoon character (familiar → ≈100% accuracy expected for children and adults, cognitively easy, only needs matching) vs humans (unfamiliar → higher cognitive demand).

In a target-absent task, matching isn’t possible so a selection must be made → harder → children rely heavily on social factors → more errors than adults.

Method

Design: laboratory experiment (not a police station — real line-ups happen somewhere official). 3 independent variables:

IV	Levels	Design
Age	young children vs adults	independent measures
Line-up type	identification (target-present) vs rejection (target-absent)	repeated measures
Level of cognitive demand (familiarity of target)	cartoon (familiar, low demand → differences due to social demand) vs human (unfamiliar, higher demand)	repeated measures

DV: whether the participant identified the correct face if present or the empty silhouette if not. Children responded by pointing (recorded by experimenter); adults recorded responses on a sheet. Photoarrays were black & white — close-cropped faces for cartoons, head-and-shoulder images for humans.

Sample

	Children	Adults
N	59 (21 female, 38 male)	53 (36 female, 17 male)
Age	4–7 years, mean 4.98	17–30 years, mean 20.54
Recruited	pre-kindergarten / kindergarten classes in 3 private schools, Eastern Ontario, Canada	Introductory Psychology Participant Pool, Eastern Ontario University

Parents/guardians (and adults) confirmed participants were familiar with the 2 target cartoons: Dora the Explorer and Go, Diego, Go!

Procedure — making the stimuli

	Videos (6 s, colour, no sound; 2–3 s of the face)	Photoarrays
HUMAN targets (2 Caucasian students)	female brushing hair; male putting on coat	photographed in different clothes to the video; 4 foils per target chosen by 3 raters for similar facial features, hair length + colour; cropped to face, neck + top of shoulders
CARTOON targets	Dora talking to the audience; Diego putting on gloves for safety — no other characters/sound; 2–3 s of the face	still images of the 2 targets + 4 foils from the internet (3 raters judged ~10 cartoons per target; 4 most similar chosen); cropped to face only (clothing was similar in every episode)

• Target-present: target + 3 foils. Target-absent: 4 foils. Every line-up also contained a blank silhouette (a choice can be made even if the target is judged absent).
• Simultaneous array — target (if present), all foils + blank silhouette shown at once.
• Target position randomised; for target-absent the 4th foil was in the same position as the original target.
• All black & white — so bright cartoon colours couldn’t become the focus of recognition rather than the identity of the character.

Procedure — running the experiment

• Each participant watched 4 videos (cartoon target-present, human target-present, cartoon target-absent, human target-absent) in random order, each followed by a line-up. Position of target / matching foil counterbalanced. Shown on 13-inch laptops.
• Written consent from parents/guardians; adults completed a consent form. Adults came to the lab, told it was a study about memory.
• Children tested by 3 female experimenters at the school, briefed as “researchers from the university doing a project on TV shows and computer games”. Told they could change their minds about taking part and wouldn’t get into trouble. Did craft work first to make them comfortable.
• Experimenters wore professional-casual clothes (sweater/blouse, dress-pants) — not overly formal because that could suggest authority and add social pressure.

Results

Two key comparisons: ① children vs adults (identification + rejection accuracy) ② children’s accuracy for cartoons vs humans. Adults generally better than children at most tasks; cartoon responses more accurate than human; little difference between children and adults for target-present cartoons.

Group	Target	Identification rate (target-present)	Correct rejection rate (target-absent)
Children	Dora	1.00	0.80
	Diego	0.97	0.67
	Human female	0.24	0.47
	Human male	0.21	0.43
Adults	Dora	1.00	0.96
	Diego	0.89	0.92
	Human female	0.46	0.72
	Human male	0.85	0.67

Other responses in Table 3.6: wrong choices in target-present = foil or false rejection (chose the blank); wrong choices in target-absent = false identification of a face. E.g. children, human female target-present: foil 0.38, false rejection 0.38; human male: foil 0.45, false rejection 0.34. Adults, human female: foil 0, false rejection 0.54; human male: foil 0.15, false rejection 0.

Line-up	Comparison	Statistical conclusion	As predicted?
Target-present (identification)	Children: cartoon vs human	significantly more accurate with cartoons	—
	Adults: cartoon vs human	significantly more accurate with cartoons	—
	Children vs adults, cartoons	children have similar accuracy	yes ✓
	Children vs adults, humans	children significantly less accurate	yes ✓
Target-absent (rejection)	Children: cartoon vs human	significantly more accurate at rejecting cartoons	—
	Adults: cartoon vs human	significantly more accurate at rejecting cartoons	—
	Children vs adults, cartoons	children significantly less accurate	yes ✓
	Children vs adults, humans	children significantly less accurate	yes ✓

All 4 predictions supported.

Conclusion

• As predicted, children easily found the cartoon in the target-present line-up (≈100% — a ceiling effect). It’s cognitively easy: just match an existing memory to a face.
• Any error in the target-absent cartoon line-up must be due to social, not cognitive, factors — it can’t be a faulty memory for the cartoon faces. Most likely: the child’s expectation that they should make a selection rather than say the target is absent.
• The child–adult difference in rejecting human faces (like cartoons) can also be explained by social demands.
• Social vs cognitive demands are relative, not absolute — both tasks are affected to some extent by both. All we can say: for children, social factors play a larger part in target-absent than target-present line-ups.

Strengths & weaknesses

✓ Strengths	✗ Weaknesses
Standardised environment — adults’ + children’s experience as identical as possible → controlled.	Not real — the line-up wouldn’t feel as important as a real case; emotional experience differs → lower ecological validity (particularly as the study looked at social effects, which emotion can influence). Needs research in realistic settings.
DV clearly operationalised (choice of individual in the line-up) = accurate measure of accuracy.	Small differences between groups: children pointed, adults wrote — seeing the options on paper might affect adults differently. (Necessary to make the task accessible to children.)
Quantitative data → statistics comparing conditions for effects of age + target type.	Demand characteristics controlled by giving minimal info (adults: memory study; children: TV/computer games) → ↑ internal validity — but thinking it’s a memory study may affect recall differently from thinking it’s about TV.

Ethical issues

• Ethics concern wider society as well as participants: findings make a valuable contribution to how child witnesses are used → improve lives. But benefits must be weighed against potential damage to participants.
• If the children had been victims of real crimes, considerable potential for harm — but the experimental situation was not crime-related, so the children weren’t at risk.
• Safeguards: consent from parents/guardians + informed consent from the children in a child-friendly way · comfort and right to withdraw ensured · ongoing monitoring of fatigue, anxiety and stress.

5. Issues, Debates & Approaches

① Application of psychology to everyday life

Study	Application
Andrade	Doodling could help us focus (lectures; waiting for a friend’s car). BUT it must be undemanding (just shading). Deliberately drawing something specific could be counterproductive — drawing becomes the primary task and distracts from the lecture. Helps students use doodling effectively.
Baron-Cohen	Eyes carry lots of information for attributing emotion → possible programme to teach people with ASD to interpret emotions; Eyes test could be developed to aid initial diagnosis and signposting to clinical staff.
Pozzulo	Children aren’t as reliable witnesses as adults in police line-ups (unknown if culprit is present; children more likely to make false positives). But mainly due to social, not cognitive, reasons → improve reliability of child evidence → next step: research ways to combat social factors that reduce accuracy.

② Individual and situational explanations

Study	Situational	Individual
Andrade	Doodling affected recall → a situational effect on information processing. She deliberately ensured participants were bored → situational causes of doodling itself.	Huge differences in shapes shaded (0–110) + not everyone doodles → individual causes behind doodling behaviour too.
Baron-Cohen	—	ASD/HFA people did significantly worse than “normal” → reading emotions from the eyes is an individual skill (not external environment). Standardised environment across all groups supports the individual explanation.
Pozzulo	Poor child performance in target-absent line-ups due to social demands (e.g. feeling pressured).	It may appear to be individual (poorer cognitive abilities of children) but the evidence shows this is unlikely.

③ Nature versus nurture

Do we have our characteristics because of nature (genetics) or nurture (experiences/influences after conception)? Long-running debate.

ASD: no known single cause; both genetic and environmental arguments. Evidence emerging that genes are involved specifically with theory-of-mind problems. Researchers also look at environmental factors: maternal speech, early interactions with siblings and peers.

④ Children as participants

• Baron-Cohen did not use children — adult male pictures of eyes + relatively sophisticated words wouldn’t suit children. Alternative: Baron-Cohen’s Sally and Anne test — questions after a short scene acted out with dolls.
• Pozzulo shows effective ways to research children: practical changes (researcher recording the open filler task; pointing to record line-up responses); ethical changes (right to withdraw, consent explained child-friendly, settling them down, monitoring fatigue/anxiety/stress).

6. Exam-style Questions — Answer Pointers

7. Full Chapter Cheat-Sheet

	Andrade (2010)	Baron-Cohen et al. (2001)	Pozzulo et al. (2011)
Topic	doodling · attention + memory	Eyes test · theory of mind	line-ups · false positives
Method	lab expt, independent measures	lab quasi-expt, independent groups	lab expt; age = independent; others repeated
Sample	40 panel volunteers (18–55)	15 AS/HFA males + 122 + 103 + 14 controls	59 children (4–7) + 53 adults
IV	doodle vs no doodle	participant type (ASD vs control)	age · line-up type · cartoon vs human
DV	memory score (correct − false alarms) names + places	Eyes test score (/36); AQ; IQ	correct pick / silhouette
Key result	doodlers recalled ≈29% more (7.5 vs 5.8)	21.9 vs 26.2 / 28.0 / 30.9; r(AQ,Eyes) = −0.53	children ≈ adults (cartoon ID); worse on human ID + all rejection
Conclusion	doodling aids concentration (attention or deeper processing)	ASD has specific theory-of-mind deficit; revised test valid	child errors = social not cognitive
Weakness	no daydreaming measure; volunteers	static eyes; small ASD sample	not real ⇒ ecological validity
Ethics	deception, no informed consent	consent ✓; negative portrayal risk	child consent, withdrawal, monitoring