Social Approach

1. What is the Social Approach?

Idea: Social psychologists study how our behaviour, thoughts (cognitions) and emotions are shaped by other people — groups, authority, personal space, helping.

• Social roles = ways we behave as members of a group (student, child, worker, manager…). We switch roles to meet expectations — e.g. a manager is formal in the office but casual with their mother.
• Others affect us even when not present — excitement before a party comes from imagining friends.
• Social pressure = influence of a person/group on another person/group. Can be direct (orders, threats, demands) or indirect (social approval/reward, need to conform).

Key term	Meaning
obedience	following a direct order from a person/people in authority
authority	person/organisation with power who can give orders & require obedience
social roles	ways we behave as members of a social group
social pressure	influence of a person/group on another person/group
destructive obedience	obedience with potential to cause psychological/physical harm or injury
legitimacy	how worthy of obedience an authority is perceived to be (e.g. police badge + uniform)
individual–situational	debate: personality/physiology (individual) vs environment (situational) — or interaction

2. Core Study 1 — Milgram (1963): Obedience

Milgram, S. (1963). Behavioural study of obedience. Journal of Abnormal and Social Psychology, 67, 371–378.

① Psychology being investigated

Obedience & social pressure — esp. destructive obedience. Obedience is a basic part of social structure (we need authority systems to live together) but can cause harm/aggression.

Link: Holocaust — ~11 million innocent people murdered under Nazi orders (incl. 6 million Jews). It needed obedience from many ordinary citizens. At Nuremberg Trials some said they were “just following orders”.

② Background

• Theory: the “Germans are different” hypothesis — Germans had defective personal traits ⇒ extreme obedience (an individual explanation).
• Milgram (born into a Jewish family) challenged this → proposed a situational explanation.
• Pre-study: he asked psychology students & colleagues how many would give the max shock → they predicted < 3% (many said nobody would).

③ Aim

To investigate how obedient people are to orders from an authority — specifically whether they obey even when it physically harms another person (destructive obedience), using a lab procedure where a researcher orders the participant to give electric shocks to a “victim”.

④ Method

• Controlled observation in a lab — variables & measurements controlled, behaviour observed & recorded.
• Called an “experiment” but no manipulated IV and no control condition — every participant had the same procedure. (Later variations — e.g. voice feedback from the victim — allowed comparisons.)
• Obedience operationalised = maximum voltage given in response to orders. Observers also noted body language, verbal comments & protests.

⑤ Sample

• 40 men, aged 20–50, from New Haven, USA; recruited by a newspaper advert ⇒ volunteer (self-selected) sample.
• Mix of unskilled workers, white-collar workers (professional jobs) & professionals.
• Volunteer sample = quick & easy, but may not be representative.

⑥ Procedure

1. Each was promised $4.50 just for turning up (not conditional on finishing). Done at Yale University in a modern lab → made it seem legitimate.
2. Arrived individually, met another “participant” = a stooge / confederate (likeable, middle-aged, trained by Milgram).
3. Told it was a study of punishment on learning. Drew paper slips for “teacher” / “learner” — rigged: real participant was always teacher.
4. Learner strapped to a chair with electrodes in another room. Teacher shown the shock generator: switches 15 V → 450 V in 15 V steps, labelled “slight” … “moderate” … “danger: severe shock” … last two XXX.
5. Told shocks were painful but not dangerous; given a 45 V sample shock. Then seated behind a wall — could hear but not see the learner.
6. Memory task: teacher reads word pairs, tests learner. Each mistake → shock, +15 V each time. (Learner was a stooge following a pre-set plan of wrong answers; he never actually got shocked.)
7. Experimenter (31-yr-old biology teacher, grey technician’s coat, serious) stayed in the room.

⑦ Standardised prods & stopping rule

Up to 300 V learner was silent. At 300 V and 315 V he banged on the wall, then no more noises / no answers. Experimenter said no response = wrong answer ⇒ still shock. If the teacher protested, 4 verbal prods in fixed order:

• Ends when participant refuses to go on, or gives the max 450 V.
• One-way mirrors let observers record behaviour/comments.
• Debrief + interview: deception explained; teacher rated how painful the 450 V was on a 0–14 scale; met the learner again to see he was unharmed.

⑧ Results

• Most were convinced it was real. Mean pain rating of 450 V = 13.42 / 14 ⇒ they believed they caused serious pain.
• Still, extremely high destructive obedience: mean voltage = 368 V; all gave at least 300 V; 65% (26/40) gave the max 450 V.
• Prediction was 0–3% ⇒ startling contrast.

⑨ Qualitative results

• Signs of tension, increasing with shock level: sweating, shaking, groaning; 14/40 showed nervous laughter / smiling.
• One participant had a violent seizure from stress ⇒ could not continue.
• Protests e.g. “I don’t think this is very humane…”, “I’m gonna chicken out… I’ll hurt his heart” — but the verbal prods usually persuaded them to continue.
• Afterwards: visible relief (wiped faces, sighed, shook heads). A small minority stayed calm throughout.
• Extra: 43 Yale undergrads did it unpaid — results “very similar” (not in the original paper).

⑩ Conclusions

• Supports a situational explanation of obedience. Factors for high obedience: perceived legitimacy (Yale, academic setting, uniform/lab coat); participants were paid so felt obliged; both believed they and the learner had freely volunteered.
• Two main conclusions: ① Individuals are far more obedient to authority than expected — true for most people. ② Despite high obedience, people find destructive acts stressful ⇒ conflict between need to obey authority and need to avoid harming others.

⑪ Strengths & weaknesses

Strengths ✓	Weaknesses ✗
Controlled extraneous variables (e.g. age/appearance of stooge) → standardised procedure (same prods every time) ⇒ reliable	Participants all men from one local area ⇒ low generalisability (no male–female comparison)
Realistic shock generator + sample shock ⇒ believed it was real ⇒ valid	Voltage doesn’t fully explain the experience of participants
Sample had varied ages & backgrounds (incl. professionals) ⇒ better external validity	Observations (behaviour & comments) are subjective qualitative data — though they gave richer understanding of the tension
Voltage = quantitative, objective ⇒ easy to compare, shows overall obedience	Lab = artificial; major ethical issues (below)
Interviews/observation = qualitative data explained behaviour

⑫ Ethical issues

• No informed consent — told it was about memory & punishment.
• Deception throughout — e.g. rigged role allocation, fake shocks.
• Right to withdraw — some asked to stop, but many felt obliged to continue because of the verbal prods.
• Not protected from psychological harm — visible extreme distress, yet only 1 trial was stopped; possible lasting distress about their own behaviour.
• Positive: all debriefed & reassured no real harm done.
• Harmed reputation of psychology; public may distrust and refuse future research. Caused outrage when published.

⑬ Issues & debates + application

• Application to everyday life: we follow orders even when uncomfortable (e.g. military, hospitals where staff follow orders against conscience) ⇒ organisations can introduce whistle-blowing policies & training so staff can safely challenge harmful orders.
• Individual vs situational: findings challenge “certain personalities are destructively obedient” — all gave 300 V ⇒ blame situational variables (legitimacy of setting, authority of researcher). But not everyone gave 450 V ⇒ individual differences exist too.
• Another real example: Hofling et al. (1966) — nurses obeyed a phone order from an unknown “Dr Smith” to give 20 mg of a drug when the label said max 10 mg. Nearly all obeyed.
• Temporal validity: would his findings hold today?

3. Core Study 2 — Perry et al. (2015): Personal Space

Perry, Mankuta & Shamay-Tsoory (2015). OT promotes closer interpersonal distance among highly empathic individuals. Social Cognitive and Affective Neuroscience, 10(1), 3–9.

① Psychology being investigated

Interpersonal distance (personal space) = the bubble of space around a person which they prefer others not to enter — it moves with them, bigger/smaller depending on the situation.

• Too far away ⇒ signals we feel uncomfortable / less responsive. Too close ⇒ other person feels uncomfortable or threatened.
• Depends on how well we know the person + culture (Beaulieu 2004: contact cultures stand closer; no-contact cultures touch less & stand further apart).
• Hall (1966) — 4 zones (middle-class Americans) ↓

• Intimate — romantic partners / very close family; all senses involved.
• Personal — most everyday interactions; can see, touch & hear.
• Social — formal interactions; louder voice, body movement, eye contact.
• Public — public figures (speech); loud voice & big body movements.

OT affects people differently depending on their situation, individual attributes & culture. Empathy ability may change how people process social cues ⇒ may also link to personal-space preferences.

② Background & Aim

• Looked at personal-space preference for different social figures (strangers vs friends) & the effect of OT.
• Tested if people with different empathy are influenced by OT in same or different ways = an interaction effect (shown with “×”, e.g. OT × empathy × condition).
• Used a placebo to provide a control condition.

③ Method & design

Laboratory experiment at the University of Haifa, using a mixed design (independent groups + repeated measures) for 2 experiments.

IV	Levels	Design
Empathy	High / Low (naturally occurring, not manipulated)	independent groups
Treatment	OT / placebo (saline) — manipulated	repeated measures
Condition (Exp 1 only)	stranger, authority, friend, ball	repeated measures

DV: personal-space requirement. Exp 1 = preferred distance between participant & approaching person/object. Exp 2 = preferred distance & angle between two chairs. (No “condition” IV in Exp 2 — always imagined the same person.)

④ Sample

• 40 male undergraduates, Univ. of Haifa, aged 19–32 (mean 25.29); for course credit or payment.
• Normal vision, no psychiatric/neurological history; 5 left-handed.
• Split by the Interpersonal Reactivity Index (IRI) — 28-item self-report, four 7-item empathy subscales:

Group	n	IRI score	Mean age
High IRI	20	≥ 40	23.9
Low IRI	20	< 33	25.9

⑤ Procedure

1. Attended uni for Exp 1; a week later, same day & time, did Exp 2.
2. Randomly given 24 IU intranasal OT or saline placebo, self-administered with a nasal dropper under supervision.
3. Double-blind: neither participant nor experimenter knew which they got.
4. Then did the IRI online, waited 45 min in a quiet room reading nature magazines (limits social interaction).
5. Order of experiments counterbalanced (ABBA): half OT week 1 → saline week 2; other half reverse.

⑥ Experiment 1: Comfortable Interpersonal Distance (CID)

Computerised version of an older paper-based CID test. Imagine yourself at the centre of a room as someone approaches along a radius; press space bar where you want them to stop. The approaching one could be a close friend, stranger, authority figure or rolling ball (control). Animation stops when the figures collide or space bar is pressed.

• 24 trials per figure ⇒ 96 trials. DV = percentage of remaining distance from the total.

⑦ Experiment 2: Choosing rooms

• Told they would later sit in a room with another participant to discuss personal topics & that their room choices would be used to design it (deception — no discussion ever happened).
• Shown pairs of coloured room pictures (2 chairs, table, cupboard, plant, lamp, clock) for 2 s; pick the preferred room by keyboard.
• Experimental stimuli: distance between chairs 20–140 cm (steps of 20 cm) & angle (0° both face forwards, 45° each, 90° facing each other).
• Control stimuli: distance between table & plant 200–320 cm (steps of 20 cm); same angles (non-social objects).
• 21 chair-distance pairs, 21 table-plant pairs, 3 angle options, ×7 repeats = 84 pairs; each repeated twice = 168 pairs.

⑧ Results — Experiment 1 (CID)

DV = mean % of distance from centre. Higher % = prefers greater personal space.

Condition (IV)	Stranger	Authority	Ball	Friend
Mean distance	39.82%	34.12%	20.20%	12.46%

⇒ Need for personal distance ↑ the less well known the approaching figure is.

Interaction: treatment × empathy

OT decreased preferred distance in the high empathy group but increased it in the low empathy group ⇒ OT has a differential effect rather than the same effect on everyone.

	OT	Placebo
High empathy	23.29%	26.11%
Low empathy	30.20%	26.98%

Interaction: treatment × condition × empathy (high empathisers)

	OT	Placebo
Stranger	39.73%	38.55%
Authority	30.55%	33.92%
Ball	14.42%	20.96%
Friend	8.49%	11.02%

• Placebo: high empathisers differed significantly between friend–authority & friend–stranger.
• OT: same differences plus ball–stranger & ball–authority ⇒ the ball becomes an “invitation” to social interaction — a cue enhanced by OT for high empathisers.

⑨ Results — Experiment 2 (rooms)

• Differences in chair distance but not in preferred angles.
• High empathy → closer chairs after OT than placebo; low empathy → opposite.
• No interaction for table–plant (control) ⇒ OT doesn’t change general preferences, only those with a social context ⇒ supports social salience hypothesis.
• Comparing both experiments: moderate correlation between average distance in CID (placebo) & chair distance (placebo).

⑩ Conclusion

OT enhances social cues in opposite ways for different empathy levels (supports social salience): low empathy → prefer ↑ distance; high empathy → prefer ↓ distance. Also confirms earlier work: people need less distance from close friends than from strangers.

⑪ Strengths & weaknesses

Strengths ✓	Weaknesses ✗
Counterbalanced order ⇒ no order effects	Lab/university = artificial ⇒ low ecological validity
Double-blind ⇒ less demand characteristics & researcher bias	Imagining strangers on a screen ≠ real people ⇒ lacks mundane realism (no facial expression, tone of voice, gestures)
Extraneous variables controlled (waiting room limited social contact)	Only 40 young male students from one place ⇒ low generalisability
Standardised computer program (e.g. 2 s images, computer-generated) ⇒ reliable

⑫ Ethical issues

• ✓ Informed consent obtained; approved by University of Haifa ethics committee & Hadassah Medical Centre; no side effects reported; painless; fully debriefed.
• ✗ Deception about Exp 2’s purpose (discussion never happened) — unlikely to cause psychological distress.

4. Core Study 3 — Piliavin et al. (1969): Subway Samaritans

Piliavin, Rodin & Piliavin (1969). Good Samaritanism: an underground phenomenon? Journal of Personality and Social Psychology, 13(4), 289–299.

① Psychology being investigated

• Bystander = person present but not directly involved. Bystander apathy = a bystander doesn’t show concern for someone in need.
• Trigger: murder of Kitty Genovese (New York, 1964) — attacked near her home; reports said ~38 neighbours saw/heard but didn’t prevent it.

② Background

• Darley & Latané (1968): those who thought others also witnessed an emergency (e.g. someone having a seizure) were less likely to help than those who thought they were alone ⇒ diffusion of responsibility (perceived obligation to help is reduced when others are present).
• Alternative: modelling — if we see others helping, we may imitate (Bandura).
• We may help similar people more; help is more likely when victims are not seen as responsible for their situation (Schopler & Matthews 1965) — creates more sympathy.
• Earlier work was lab-based ⇒ Piliavin et al. tested real-life helping.

③ Aim

To study bystander apathy & diffusion of responsibility in a natural setting, and the effect of 4 variables on helping behaviour (“Good Samaritan” = someone who offers help to others in difficulty — from a Bible story):

① type of victim · ② race of victim · ③ behaviour of a model · ④ size of bystander group

④ Method & design

• Field experiment — realistic place (New York subway); IVs manipulated, DV measured. Independent groups (trials on different days/participants).
• Standardisation between trials (victim identically dressed & behaved) although in the field.

IV	Levels
Type of victim	drunk or ill (cane)
Race of victim	black or white
Model’s behaviour	close to / distant from victim; helped early or late
Size of group	naturally occurring number of passengers (not manipulated)

DV = level of helping: quantitative — time for first passenger to help & total number who helped (+ race, gender & location of helper). Qualitative — verbal remarks made by passengers.

⑤ Sample

• Passengers on the Harlem–Bronx underground line (New York), weekdays 11 a.m.–3 p.m. ⇒ opportunity sample.
• ≈ 4450 participants: ~45% black, 55% white. Mean 43 passengers per carriage; mean 8.5 in the critical area.

⑥ Procedure

1. 4 teams of student researchers; each team: 2 male + 2 female. Boarded by different doors.
2. Female confederates sat in the adjacent area next to the critical area and recorded data (2 observers). Male confederates were the victim and the model.
3. Same route each trial: 7.5-minute gap between two stations. At about 70 s the victim staggered forward & collapsed, lying on the floor looking up. If nobody helped, the model helped him up at the next stop.
4. Victim stood at the pole in the middle of the critical area; the model stayed standing.

Victims & models

• Victim played by different males aged 26–35 (3 white, 1 black), dressed identically in casual clothes (jacket, old trousers, no tie).
• 38 / 103 trials: drunk — smelled of alcohol, bottle in brown paper bag. 65 / 103: cane — sober, black cane.
• Models: white males aged 24–29, informal dress. When helping: raised victim to sitting & stayed with him till the next stop.

Condition	Model’s behaviour
Critical / early	in critical area, helps at 70 s
Critical / late	in critical area, helps at 150 s
Adjacent / early	in adjacent area, helps at 70 s
Adjacent / late	in adjacent area, helps at 150 s
No model	does not help until after the trial/next stop

⑦ Results

• Much more helping than in lab studies: 78% of victims got spontaneous help (before the model acted / no-model); in 60% of cases more than one person helped.
• Type of victim: cane helped in 62/65 trials; drunk in 19/38. In all but 3 cane model-trials help came before the model could act.
• Race: cane — equal help for black & white. Drunk — some same-race helping (more help to own race); black victims got less help overall (not statistically significant).
• Sex of helper: 90% of helpers were male. Females said e.g. “It’s for men to help him”, “I’m not strong enough” — maybe a gender difference or because the victim was always male.
• Modelling: hard to analyse since most help was spontaneous; early (70 s) model intervention was slightly more likely to produce helping than late (150 s).

% trials with help	White cane	White drunk	Black cane	Black drunk
No model	100%	100%	100%	73%
Model trial	100%	77%	— *	67%

* no model trials for the black “victim” in the cane condition.

Diffusion of responsibility — predicted vs found

⑧ Cost–benefit model (alternative explanation)

• Witnessing an emergency raises arousal (disgust, aversion, or even sympathy & courage). To reduce this unpleasant arousal people respond.
• Cost–benefit model: weigh advantages vs disadvantages of helping — beneficial ⇒ help; risks too great ⇒ don’t.
• Evidence: more comments made in trials without helping, and more with a drunk victim; in ~20% of trials passengers moved away from the critical area. Less helping due to disgust ⇒ arousal prompts commenting/moving away instead.

⑨ Conclusion — factors affecting helping

• Type of victim — cane helped more than drunk.
• Gender of helper — men more likely to help than women.
• Similarity of victim to helper — maybe own race (esp. if drunk).
• Duration of the emergency — the longer it lasts, the less likely anyone helps & the more likely they cope with arousal another way.
• Many people helped spontaneously, even in a group; no evidence of diffusion of responsibility.

⑩ Strengths & weaknesses

Strengths ✓	Weaknesses ✗
Field experiment ⇒ high ecological validity; passengers unaware ⇒ behaved naturally	Less control over extraneous variables (weather, train delays) ⇒ ↓ validity & reliability
~4500 participants of mixed ethnicity & gender ⇒ more representative ⇒ ↑ generalisability	Can’t be sure participants took part only once (same route) ⇒ may have seen > 1 condition
Quantitative measure (number of helpers + time to help) = objective; 2 observers ⇒ more reliable	If passengers suspected a set-up ⇒ demand characteristics
Qualitative remarks gave depth	All from New York ⇒ can’t predict other countries (unrepresentative)
Standardised: same route, same victim clothing/behaviour

⑪ Ethical issues

• No consent — didn’t know they were in research. No debriefing afterwards.
• Deception — they believed the victim genuinely needed help.
• Possible psychological distress: guilt at not helping, concern for the victim.
• Ethics committee test: today this would need ethics approval (the study was 1969).

5. Issues, Debates & Research Methods

① Application of psychology to everyday life

Study	Real-life use
Milgram	Destructive obedience can occur in nearly anyone due to situational variables (legitimacy of authority) ⇒ e.g. whistle-blowing policies in hospitals to report mistakes by senior staff.
Perry	OT effect is complex, not universally positive ⇒ using it as a treatment (autism, social anxiety) could worsen social difficulty.
Piliavin	Shows situational factors that make bystanders help (same sex/race, can’t walk past, victim not disgusting/embarrassing).

② Individual vs situational explanations

• Milgram: mainly situational (legitimacy, authority). But not everyone gave full voltage ⇒ individual differences also matter.
• Perry: situational — relationship to person/object affects need for space; individual — empathy explains why people exposed to the same OT still differ.
• Piliavin: situational (type of victim, model, similarity, duration) — plus individual factors such as sex of helper.

Debate	Meaning
Cultural differences	Much research uses narrow samples (e.g. Western uni students: educated, young, wealthy) ⇒ results may not apply elsewhere; cross-cultural research highlights differences.
Reductionism vs holism	Reductionism = explaining a phenomenon by its most basic elements (neurotransmitters, genes) or only some factors (e.g. only cognitive). Holism = view the whole; many interacting factors; no single cause.
Determinism vs free-will	Determinism = behaviour governed by processes beyond our control (genes, upbringing). Free-will = people can choose & are responsible. e.g. crime: unavoidable result of genes/upbringing vs entirely their own choice.
Idiographic vs nomothetic	Idiographic = describes the individual (qualitative data, case studies, unstructured self-report — psychodynamic, brain damage). Nomothetic = establishes general laws for all people (experiments/observations, quantitative — e.g. Milgram, Piliavin, drug trials). Complementary: idiographic for new areas, then test on a larger scale.

③ Research methods at A Level

Method	Points
RCT (Randomised Control Trial)	Similar people randomly allocated to experimental group (intervention, e.g. drug) or control (alternative/none; often placebo); results analysed statistically. + reduces bias, especially double-blind (neither participants nor researchers know the condition) ⇒ ↑ validity. − hard to run large trials with similar participants (age, sex, SES, health); less effective for psychological treatments (many uncontrollable factors) than drugs (more universal biochemical effect).
Questionnaires	Postal/online: private ⇒ more relaxed/honest, reach many people quickly. ✗ lost mail; self-selecting (higher return for interesting topics) ⇒ not representative. Rating scales (e.g. Likert 1–5) show degree of opinion; forced/fixed-choice (yes/no, + “don’t know”) — both quantitative, easy to compare, but can be badly designed/misread (scale read the wrong way round). Use open questions if depth is needed.
Psychometrics	Self-report with entirely objective scoring (no researcher judgement). Used for IQ, personality, memory, happiness; organisational & clinical use. + standardised, reliable, quantitative, easy to compare with norms, less scoring bias. − only a snapshot (mood affects performance), practice effects, may lack face validity, may be biased to certain social/cultural groups.
Temporal validity	Type of external validity: do findings stay true over time? High when replicated at a later date. e.g. Piliavin: “it’s for men to help him” — gender attitudes may have changed. Think: would Milgram’s findings hold today?

6. Exam-style Questions & Cheat-Sheet

Question type (marks)	What to write
Outline one assumption [2]	“Behaviour can be influenced by the actual, implied or imagined presence of others” + a short example.
Outline a quantitative / qualitative result, Milgram [2 each]	Quant: 65% (26/40) gave 450 V; mean 368 V; all ≥ 300 V. Qual: sweating, shaking, groaning; 14/40 nervous laughter; one seizure.
Identify 2 features of Milgram’s situation [2]	Prestigious Yale lab (legitimacy) · experimenter’s uniform/lab coat · paid participants · believed learner volunteered.
Explain real-life application, Milgram [4]	Situational variables cause destructive obedience ⇒ whistle-blowing policies / training, e.g. hospitals/military.
Evaluate controlled observation (Milgram) [6]	+ control of extraneous variables, standardised ⇒ reliable; objective data. − artificial, subjective observation, ethics.
Describe Perry’s sample [2]	40 male Haifa undergraduates, 19–32 yrs (mean 25.29), split into 20 high / 20 low IRI.
State ‘personal space’ [1]	The area around a person in which they prefer others not to enter.
Perry IVs (both experiments) [4] / DV in Exp 1 [2]	Empathy (high/low, not manipulated) & treatment (OT/placebo, manipulated). DV = % of remaining distance when the participant pressed the space bar.
Evaluate Perry: 2 strengths + 2 weaknesses [10]	Double-blind, counterbalanced, standardised vs ecological validity, mundane realism, all-male sample.
2 ways Piliavin was standardised [4]	Same route & timing (collapse at ~70 s) · victim identically dressed & behaved · models similar · same procedure for 4 teams · same 2 victim types (drunk/cane).
Diffusion of responsibility [4 + 3]	Define (↓ personal responsibility when others present) + example; results contradict it: no fall in helping with bigger groups (78% spontaneous help).
Field experiment: 1 strength + 1 weakness [4]	+ ecological validity (natural behaviour); − less control of extraneous variables / no debrief / demand characteristics if suspected set-up.

Activity answers from the chapter