Learning Approach

1. What is the Learning Approach?

Idea: concerned with how humans & animals acquire or change their behaviour — learning from others, learning by association, learning from rewards & punishments.

Learning = new, permanent changes in behaviour following experience. Not the same as memory (times-tables = a memory task) and not basic biology (breathing, blinking).

• new responses acquired = the relatively permanent change
• learning may happen without new behaviour being shown yet = behavioural potential
• the environment influences & provides opportunities for learning (experiences)

Not all learning is good: bad habits are learned too, often with no effort — simply by being in a certain environment. We keep learning across the whole lifetime (e.g. older adults learning new technology).

Main assumptions

1. Blank slate — each life begins empty; observable changes in behaviour result from interactions with the environment.
2. Humans & animals learn through social learning, operant conditioning and classical conditioning.
3. These are stimulus–response processes: experiences in the environment (stimuli) shape our behaviour (response).

Process	Learning is through…	Core study
Social learning	observing & imitating a role model	Bandura et al. — aggression
Operant conditioning	consequences of our actions (reinforcement / punishment)	Fagen et al. — elephant trunk wash
Classical conditioning	association between stimuli (+ evaluative learning)	Saavedra & Silverman — button phobia

2. Core Study 1 — Bandura, Ross & Ross (1961)

Bandura, A., Ross, D., & Ross, S. A. (1961). Transmission of aggression through imitation of aggressive models. Journal of Abnormal and Social Psychology, 63(3), 575–582.

The psychology being investigated — social learning & aggression

Children copy adults. The immediate social setting encourages the child to imitate what he/she is watching, which makes it more likely the child does what others around them do. Or watching a behaviour can give a new response the child can reproduce independently ⇒ it should generalise to new settings and be produced without the model.

Learning from others can lead to aggressive or non-aggressive behaviour. Bandura expected: aggressive model ⇒ more aggression; non-aggressive model ⇒ less aggression than normal.

Background

• Previous research: children imitate a model when the model is present. Bandura wanted to know if social learning theory explains aggression when the model is no longer there.
• Children are rewarded differently for imitating adults. In the USA in the 1960s: boys rewarded for sex-appropriate behaviour, punished for inappropriate (e.g. cooking, ‘playing mother’); same for girls.
• ⇒ two predictions: (1) children imitate same-sex models more; (2) boys imitate aggression more readily (seen as masculine-type behaviour).

Aim

To investigate whether a child would learn aggression by observing a model and reproduce it in the model’s absence, and whether the sex of the model was important.

Method — research method & design

• Laboratory experiment — not the children’s normal place; situation controlled.
• Design = independent measures (different children in each level of the IV) — but children were matched for aggression in threes.
• 3 IVs: model type (aggressive / non-aggressive / none) · model sex (same / different from child) · learner sex (boy / girl).
• DV = behaviour the child displayed, measured via controlled observation; measures of aggressive behaviour recorded.

Sample

72 children aged 3–6 years (36 boys + 36 girls) from the Stanford University nursery school.

Group	Children	Boys / girls
Control (no model)	24	12 boys + 12 girls
Aggressive model	24	12 + 12 → each of 4 cells = 6
Non-aggressive model	24	12 + 12 → each of 4 cells = 6

Within the model groups, half saw a same-sex model, half an opposite-sex model (⇒ only 6 children per experimental condition).

Procedure

Pre-test: children were watched in nursery by the experimenter + a teacher who knew them well. Rated on 4 measures — physical aggression, verbal aggression, aggression to inanimate objects, aggression inhibition (anxiety) — each on a 5-point scale. Then assigned to 3 matched groups.

Inter-rater reliability: 51 children were rated by both observers (rest by one) — similar ratings ⇒ r = 0.89 (high).

Stage 1 — Modelling: experimenter + child entered the observation room: child at a table in their ‘play area’ doing potato prints & sticker pictures. Opposite corner: table & chair, Tinkertoy set (wooden building kit), mallet, 5 ft (152 cm) Bobo doll (inflatable clown that bounces back when hit) — where the model sat. Experimenter stayed in the room so the child wouldn’t refuse to be alone/leave (appeared to work quietly at desk).

• Non-aggressive condition: model assembled Tinkertoys for 10 min.
• Aggressive condition: Tinkertoys for only 1 min, then attacked Bobo — laid on its side, sat on it & punched the nose, picked it up & hit head with mallet, tossed in the air & kicked. Sequence 3 times over 9 min, with aggressive comments (“Kick him”) and 2 non-aggressive comments (“He sure is a tough fella”).
• Of the model groups: half saw a same-sex model, half opposite sex. Control group saw no model ⇒ no aggression.

Stage 2 — Mild annoyance: all children were deliberately mildly annoyed (shown nice toys, then taken to another room). Two reasons:

• watching aggression may reduce the observer’s aggression (even if learned) — needed evidence of learning
• to make sure even non-aggressive & control children would be likely to express aggression, so any reduction could be measured

Stage 3 — Test of aggression (delayed imitation): child observed for 20 min through a one-way mirror (transparent for researcher, normal mirror for the child). Room: 3 ft (92 cm) Bobo, mallet & peg board, 2 dart guns, tether ball with a face hanging from the ceiling + non-aggressive toys (tea set, crayons & colouring paper, ball, 2 dolls, 3 bears, cars & trucks, plastic farm animals). Toys always presented in the same order.

Scoring: 5-second intervals ⇒ 240 response units per child. Response measures:

Category	What counted
Imitative physical aggression	striking Bobo with mallet, sitting on & punching its nose, kicking it, tossing in the air
Imitative verbal aggression	repeating ‘Sock him’, ‘Hit him down’, ‘Kick him’, ‘Throw him in the air’, ‘Pow’
Imitative non-aggressive verbal	repeating ‘He keeps coming back for more’ or ‘He sure is a tough fella’
Partially imitative	mallet aggression (hitting other objects with the mallet) · sits on Bobo (laid on side, sat on it, no attack)
Aggressive gun play	shooting darts / aiming gun & firing imaginary shots at objects
Non-imitative aggression	physical aggression to objects other than Bobo + hostile remarks not in the imitation list (‘Shoot the Bobo’, ‘Cut him’, ‘Stupid ball’, ‘Horses fighting, biting’, ‘Knock over people’)

Also counted: non-aggressive play + sitting quietly not playing; the children’s remarks about the situation were recorded.

Reliability of scoring: a male model scored all children — he didn’t know which condition they were in (unless he was the model), though it was often obvious for aggressive-model children. A second scorer independently rated half the children ⇒ r ≈ 0.9 for the different categories.

Results

• Children exposed to aggressive models imitated their exact behaviours and were significantly more aggressive — physically and verbally — than non-aggressive-model or control groups.
• 1/3 of aggressive-condition children also copied the model’s non-aggressive verbal remarks; none of the non-aggressive/control children made such remarks.
• Boys: with a male model, imitative physical aggression = 25.8 vs 7.2 for girls ⇒ boys imitated physical aggression of a male model far more. With a female model, girls imitated less (5.5) than with the male model.
• Girls imitated more verbal aggression of the same-sex model than boys did (girls with female model 13.7 vs boys with male model 12.7 — though not significantly so).
• Children were more likely to imitate a same-sex model than an opposite-sex one — stronger for boys than girls.
• Non-aggressive model children: much less mallet aggression than aggressive/control groups (esp. girls). Aggressive model did not affect gun play or punching Bobo, but non-imitative physical & verbal aggression was higher after an aggressive model.
• Non-aggressive play: girls played more with dolls, tea sets & colouring; boys did more exploratory play & gun play. No sex differences with farm animals, cars or tether ball. Non-aggressive-model children played more with dolls and spent over twice as much time sitting quietly, not playing.

Response category / participants	Aggressive model		Non-aggressive model		Control
	F model	M model	F model	M model
Imitative physical aggression
Female participants	5.5	7.2	2.5	0.0	1.2
Male participants	12.4	25.8	0.2	1.5	2.0
Imitative verbal aggression
Female participants	13.7	2.0	0.3	0.0	0.7
Male participants	4.3	12.7	1.1	0.0	1.7
Mallet aggression
Female participants	17.2	18.7	0.5	0.5	13.1
Male participants	15.5	28.8	18.7	6.7	13.5

Response category / participants	Aggressive model		Non-aggressive model		Control
	F model	M model	F model	M model
Punches Bobo doll
Female participants	6.3	16.5	5.8	4.3	11.7
Male participants	18.9	11.9	15.6	14.8	15.7
Non-imitative aggression
Female participants	21.3	8.4	7.2	1.4	6.1
Male participants	16.2	36.7	26.1	22.3	24.6
Aggressive gun play
Female participants	1.8	4.5	2.6	2.5	3.7
Male participants	7.3	15.9	8.9	16.7	14.3

Table 4.1 — mean aggression scores (Bandura et al.). Higher number = more of that behaviour.

Qualitative results — the children’s remarks

• Remarks about the aggressive models showed differences in reactions to male vs female models and between boys and girls.
• Some comments reflected knowledge of sex-typed behaviour — about the female model: “Who is that lady? That’s not the way for a lady to behave…” / “…She was punching and fighting but no swearing.” (disapproving)
• Comments about the male model were not disapproving — seen as appropriate & approved by boys and girls: “Al’s a good socker, he beat up Bobo…” / “He’s a good fighter like Daddy.”

Conclusion

Results strongly suggest observation and imitation can account for learning specific acts — without reinforcement of either the model or the observer. All 4 hypotheses supported:

• observed aggression is imitated (aggressive-model > non-aggressive-model / no model)
• observed non-aggression is imitated (non-aggressive-model < no model)
• children more likely to copy a same-sex model — may depend on how sex-typed the behaviour is
• boys more likely to copy aggression than girls

Strengths

• Lab experiment ⇒ extraneous variables controlled (e.g. nice toys → another room; same exposure time; behaviour standardised in each condition) ⇒ more valid (differences due to the models) & more reliable (same exposure for each child).
• Inter-observer reliability checked for the initial rating and data recording — very high.
• Pre-testing + matched pairs design ⇒ differences due to the models, not individual differences between children ⇒ ↑ validity.
• One-way mirror ⇒ children unaware of being watched ⇒ natural behaviour, not demand characteristics ⇒ ↑ validity.
• Observation split into 5-second intervals + clearly defined categories (imitative / non-imitative) ⇒ ↑ validity & reliability.
• Both quantitative & qualitative data: quantitative = objective record of acts per category; qualitative = (more subjective) reasons behind copying some behaviours but not others.

Weaknesses

• Small sample — only 6 children per condition (even though matched).
• Children quite similar — all from the same university nursery (academically able parents) ⇒ biased sample ⇒ lowers validity / generalisability.
• Could have asked children for self-reports of emotions while watching / reacting to Bobo ⇒ better explain sex-typing effects.
• No follow-up to see how long the learned behaviour lasted (if imitation = learning, change should be relatively permanent).

3. Core Study 2 — Fagen, Acharya & Kaufman (2014)

Fagen, A., Acharya, N., & Kaufman, G. E. (2014). Positive reinforcement training for a trunk wash in Nepal’s working elephants: demonstrating alternatives to traditional elephant training techniques. Journal of Applied Animal Welfare Science, 17(2): 83–97.

The psychology being investigated — operant conditioning

Operant conditioning = learning from the outcome (consequence) of our behaviour. A behaviour with a good consequence is more likely to be repeated ⇒ it has been reinforced.

Type	What happens	Example
Positive reinforcement	reward added for desirable behaviour ⇒ repeated	food, praise; praising child for ‘please’ & ‘thank you’
Negative reinforcement	something unpleasant removed/avoided in response to a stimulus	feel cold (stimulus) → put on coat (response) → warm ⇒ keep doing it
Primary reinforcer	fulfils a biological need; occurs naturally	water, food, shelter, sleep
Secondary reinforcer	a stimulus associated with primary reinforcers; must be learned	money (buys food, water, shelter)

• Shaping: behaviour built up gradually — child’s first vocalisations sound only slightly like words; with praise they become clearer & more accurate.
• Behaviour chains: simple actions combined into sequences — riding a bicycle: first learn to pedal, then steer, then balance on two wheels.
• Limit: the theory does not explain underlying, unseen reasons (thoughts, feelings). Most behaviour = learned through trial and error.

Background

Traditional elephant training: captive elephants are given ‘free’ (unlimited) contact with handlers (mahouts); behaviour managed using punishment (Figure 4.3) — pain or fear to shape behaviour, e.g. pain inflicted by a bamboo stick.

Because of concerns for captive animal welfare and keeper safety ⇒ growing interest in reward-based training (positive reinforcement) using ‘protected’ contact (a barrier between elephant & handler). Positive reinforcement methods improve elephants’ psychological well-being (Desmond & Laule, 1991).

Secondary positive reinforcement (SPR) training: animal learns to associate a specific sound (‘marker’ / ‘sound-marker’) with food. Once conditioned, the marker rewards wanted behaviour.

• Marker is better than food alone: it precisely shows the moment of correct behaviour; preparing/delivering food causes a delay.
• Food rewards still given at intervals to maintain the link between secondary & primary reinforcer.
• SPR already used successfully with pandas, primates, antelope; advantage = ability to shape captive animals’ behaviour to improve health & well-being.

Why a trunk wash? Diagnosing/treating illness in elephants = important for captive welfare. Tuberculosis (TB) = serious respiratory disease, a big concern in captive elephants, can pass between animal & human. Best detected by a sample from the trunk (‘trunk wash’) — but getting elephants to do it correctly is hard; many samples are insufficient for testing. Fagen et al. wanted to avoid traditional methods such as punishment, so they tested SPR.

Aim

To see whether free-contact, traditionally trained elephants can be trained to participate in a trunk wash using positive reinforcement.

Method — research method & design

Controlled observation of a small group of captive elephants trained over a period of weeks. Researchers watched behaviour in response to a specific stimulus and used a behavioural checklist to record responses as a percentage pass ⇒ also called a structured observation.

Sample

• 5 female elephants: 4 juveniles (aged 5–7, born at the stable) + 1 adult (estimated in her 50s); all at the same elephant stable in Nepal.
• Chosen from others as docile, not pregnant / not looking after a calf, and their mahouts were willing.
• All traditionally trained, in free contact with mahouts; none had previous SPR experience.
• Daily life: mostly grazing in the jungle under mahouts’ control; rest of day leg-chained to a stake with 6–8 feet (1.83–2.4 m) freedom. Diet: fresh grasses, grain, nutritional supplements; water at a river during grazing.

Procedure

• Method = SPR: chopped banana = primary reinforcer, short whistle blow = secondary reinforcer.
• Training in indoor sessions (7:30 a.m.–10:00 a.m. and 4:00–7:00 p.m.) by a trainer, mahout present for safety. Mahouts stood to the side, did not speak or signal.
• No elephant went longer than 2 days without a session. Elephants could choose not to engage (turn/walk away from trainer).
• Purpose: teach a voluntary trunk wash in several steps in response to a cue. After establishing marker–reward (banana–whistle), three methods:

Method	How it works
Capture	wait for animal to perform a behaviour naturally, then ‘capture’ it by marking with a reward
Lure	for non-natural behaviours, animal is ‘lured’ into a body position by placing a reward in a certain place
Shaping	after capture/lure, reward only the ‘best’ behaviours — incrementally closer to the goal behaviour

Task	Role in trunk wash	Description	Method
Trunk here	saline/water instilled into trunk	end of trunk placed in trainer’s hand	Lure (banana in trainer’s hand)
Trunk up	saline runs to base of trunk	elephant lifts trunk upward	Lure (banana held above head)
Bucket	trunk in bucket ready to exhale	end of trunk placed in bucket	Lure (banana at bottom of bucket)
Blow	exhale to give the sample	exhales through trunk into bucket	Capture + Shaping (wait for natural exhale, reward, then reward more forceful exhales)
Steady	hold any position for the required time	hold position previously asked of them	Shaping (extend how long position held, with reward)

• 3 other tasks (targeting, trunk down, trunk out) were introduced but discarded — not essential to the trunk wash.
• After each task was learned it was paired with a verbal cue: a one-syllable word with no meaning in English or Nepali (avoids meaning for elephants or mahouts).
• Behaviour chaining: behaviours put together in small sequences — performing the first correctly earns the chance to do the second & gain a reward. First paired: bucket + blow; then longer sequences until the whole trunk wash.
• Desensitisation to a syringe at the trunk-here position: over repetitions the syringe moved closer → touching trunk → inserted → inserted with increasing fluid until the elephant tolerated the full 60 ml needed for sample collection.
• Offered drinks before each session so they wouldn’t drink the saline/water; one elephant preferred saline (would drink it after rejecting water) ⇒ switched to a water solution.
• No time limit on any stage — success of each individual elephant decided it ⇒ training plans varied by individual.

Recording & passing: an assistant recorded session length (min) and the total number of cues/‘offers’ for each behaviour. After session 10, elephants were tested about every 5 sessions on previously taught behaviours. Pass = 80% (≥ 8/10 correct cues). Whole sequence passed at ≥ 80% ⇒ training complete.

Results

• The 4 juveniles learned the trunk wash; the adult did not.
• Elephants 2 & 4 never passed their steady test, but passed their full trunk-wash tests.
• Elephant 5 (adult) did not pass blow-into-bucket, desensitisation to syringe or steady.
• Some tasks harder: trunk-here needed more offers/cues than bucket or blow-into-bucket.
• Gradual improvement: mean success rate 36% after 10 sessions → 89.3% after 35 (never 100% as 90% was given as default to behaviours in a passed sequence).

Elephant	Sessions	Mean duration (min)	Total training (min)
1 (juvenile)	30	12.42	373
2 (juvenile)	25	10.29	257
3 (juvenile)	35	13.27	389
4 (juvenile)	35	11.11	451
5 (adult)	n/a	n/a	424

Table 4.4 — results of training sessions for the trunk wash (as printed in the textbook).

Conclusion

Juvenile, free-contact, traditionally trained elephants can be trained to take part in a trunk wash using only SPR. Training can be done with the elephants’ voluntary participation, avoiding punishment, and produce reliable results.

Strengths

• Controlled observation can be replicated easily using the same observation schedule ⇒ can be tested for reliability.
• Attempts to keep validity: mahouts asked not to speak/signal; verbal cues had no meaning in Nepali ⇒ limited demand characteristics on the elephants’ responses.
• Structured observation ⇒ quantitative data (correct responses to cues/offers) ⇒ objective analysis of whether SPR worked; easier to analyse than unstructured observations.

Weaknesses

• Small sample of 5 female elephants, captive, from one stable ⇒ hard to generalise; may not be representative.
• Elephants chosen for temperament (docile) ⇒ further limits representativeness.
• Practical problems: large size = risk to elephants & humans; hard to control distractions — tourists & other elephants during training may have affected concentration/willingness.
• Reliability issue: training sessions were flexible (mood & ability of each animal). If repeated with other elephants, trainer experience and elephants’ personalities/conditions (stress levels, previous SPR experience) could affect results.

4. Core Study 3 — Saavedra & Silverman (2002)

Saavedra, L. M., & Silverman, W. K. (2002). Case study: disgust and a specific phobia of buttons. Journal of the American Academy of Child and Adolescent Psychiatry, 41(11), 1376–1379.

The psychology being investigated — classical conditioning & phobias

Classical conditioning = learning in which an unconditioned response becomes linked to a previously neutral stimulus to create a learned association. Investigated by Ivan Pavlov (1927): dogs salivated in a lab (a different experiment) — salivation = unconditioned (uncontrolled) response to food; dogs began to anticipate food when they saw researchers. Pavlov paired food with neutral stimuli (e.g. bells) ⇒ after several trials, the bell (CS) made them salivate (CR).

Term	Meaning	Pavlov
NS	neutral stimulus — initially has no effect	bell
UCS	unconditioned stimulus — already produces a response	food
UCR	unconditioned response — innate (instinctive) reaction to UCS	salivation
CS	NS after pairing with UCS (sometimes after only once, more often repeatedly)	bell
CR	newly learned response, similar to the UCR	salivation to the bell

• Some psychologists believe behaviour such as phobias can be learned (and unlearned) the same way as any other behaviour.
• Expectancy learning (subtype): a neutral/non-threatening object becomes associated with a potentially threatening outcome ⇒ the person expects the negative outcome ⇒ fear in a previously non-threatening situation.
• Evaluative learning: another type of classical conditioning — association between a neutral stimulus and a negative evaluation (e.g. disgust) rather than fear ⇒ disgust, as well as fear, may be a target emotion in treating phobias. Fear/anxiety becomes associated with a neutral stimulus such as buttons.
• A strictly behaviourist view of classical conditioning = little role for thinking & feeling. This study explores cognition & emotion in changing phobic behaviour (evaluative learning).
• Operant link: reward (e.g. sticker after a check-up) ⇒ child happy to return. Phobias maintained by avoidance: anxiety is reduced when you avoid the object ⇒ avoidance is reinforced (negative reinforcement). Treatment: praise/rewards for staying calm when approaching the phobic stimulus ⇒ better tolerance.

Background

• Focus: how phobias can be treated by applying different learning theories — operant conditioning (reinforce desirable behaviour) vs evaluative learning (classical conditioning) to reduce feelings of disgust.
• Hepburn & Page (1999): adults with a blood phobia — treating disgust as well as fear helped them make progress.
• De Jong et al. (1997): children with a spider phobia — no attempt to change disgust, yet disgust declined alongside fear.

Aim

To examine the role of classical conditioning in fear and avoidance of a particular stimulus. For the case-study participant: to see if a type of exposure therapy could reduce the disgust and distress associated with buttons.

Method — research method & design

• Clinical case study — one participant whose life history & treatment were studied in depth.
• Data via self-report: boy and mother interviewed about onset & subsequent behaviour; boy observed throughout treatment to see if phobic behaviour improved; results measured on a 9-point scale of distress = the ‘Feelings Thermometer’ (0–8).

Sample

1 participant: a 9-year-old Hispanic American boy; with his mother he sought help from the Child Anxiety and Phobia Program, Florida International University, Miami. Met the criteria for a specific phobia of buttons; symptoms for about 4 years before the study.

Procedure

• Boy & mother gave informed consent. Interviewed to see whether any trauma or abuse could explain the phobia.
• Onset at age 5: he knocked over a bowl of buttons in front of his class and teacher — found it distressing; aversion steadily increased.
• When interviewed the phobia was interfering with normal functioning: he could no longer dress himself; preoccupied with avoiding touching buttons or clothing that could have touched buttons.
• Researchers created a hierarchy of feared stimuli with the participant (each item = more fear). Most difficult = small, clear plastic buttons (rated 8); handling them or touching someone wearing them = the most unpleasant task.

Stimuli (Table 4.5)	Distress rating (0–8)
1. Large denim jean buttons	2
2. Small denim jean buttons	3
3. Clip-on denim jean buttons	3
4. Large plastic buttons (coloured)	4
5. Large plastic buttons (clear)	4
6. Hugging Mom when she wears large plastic buttons	5
7. Medium plastic buttons (coloured)	5
8. Medium plastic buttons (clear)	6
9. Hugging Mom when she wears regular medium plastic buttons	7
10. Small plastic buttons (coloured)	8
11. Small plastic buttons (clear)	8

Intervention 1 — contingency management (positive reinforcement therapy): a behaviour-focused approach: the boy was rewarded for showing less fear and for actually handling the buttons. His mother gave the positive reinforcement only after he completed a gradual exposure to buttons. Sessions lasted 20–30 min. Researchers observed how he approached buttons (e.g. whether the number of buttons handled increased) and measured his subjective distress on the 9-point Feelings Thermometer.

Intervention 2 — imagery exposure therapy (the main focus): interviews showed he found buttons touching his body disgusting and believed they smelled unpleasant ⇒ basis for disgust imagery exercises.

• Unlike in vivo exposure (physically handling / being exposed to feared stimuli), imagery exposure uses visualisation.
• Combined with cognitive self-control strategies. He imagined buttons falling on him and considered how they looked, felt and smelled; talked about how the images made him feel.
• Exposures progressed from larger → smaller buttons, in line with the fear hierarchy. Each session: self-report distress on the Feelings Thermometer.

Results

Positive reinforcement therapy: successful completion of all exposure tasks on the hierarchy; boy observed approaching buttons more positively (handled larger numbers in later sessions). BUT his subjective distress increased significantly between sessions 2 and 3 and kept rising (Fig 4.7). By session 4 items such as hugging Mom wearing buttons had increased in dislike from original scores.

Behaviour improved but disgust, fear & anxiety actually increased — consistent with evaluative learning: despite repeated exposure to the phobic object, evaluative reactions (disgust) remain unchanged or even increase.

Imagery exposure therapy: appeared successful in reducing distress. E.g. imagery of ‘hundreds of buttons falling all over his body’: before = most fearful & disgusting (8), 5 midway, just 3 after the exposure.

Follow-ups (6 months & 12 months): boy reported minimal distress about buttons; no longer met diagnostic criteria for a specific phobia; feelings no longer affected normal functioning; he could wear small, clear plastic buttons on his school uniform daily.

Conclusion

Positive reinforcement therapy was successful in changing observable behaviour; imagery exposure therapy was successful in reducing feelings of fear and disgust. In particular:

• emotions and cognitions relating to disgust are important when learning new responses to phobic stimuli
• imagery exposure can have a long-term effect on reducing distress associated with specific phobias as it tackles negative evaluations

Strengths

• High validity: participant studied over time using several methods of data collection.
• Standardised measures — the Feelings Thermometer before, during & after therapy.
• Quantitative data let researchers compare phobic reactions at start vs end of treatment to assess improvement.
• Lots of qualitative data (background from boy & mother, e.g. the button incident at school) ⇒ understand reasons underlying psychological disorders.
• The subjective hierarchy and ratings were appropriate for the aim (understand evaluative learning in an individual) — a generic scale wouldn’t show personal progress.

Weaknesses

• Small sample (one person) ⇒ hard to generalise; a button phobia makes the case even less representative.
• Case study = building rapport ⇒ less anonymity / objectivity ⇒ higher risk of bias (may compromise validity); researcher bias in selecting/reporting a participant who may be likely to have a positive outcome.
• Study could be seen as subjective: the boy created his own hierarchy & gave highly individual ratings.
• Demand characteristics: boy knew he was in therapy meant to improve symptoms ⇒ may have affected the ratings he gave to the different exposure levels.

5. Issues, Debates & Approaches

The application of psychology to everyday life

• Bandura: children everywhere are exposed to aggression — real (domestic violence, news) or fictional (cartoons, films, games). Shows models can influence children’s behaviour, especially boys ⇒ countries restrict TV, films & computer games — e.g. the 9:00 p.m. ‘watershed’ in the UK + age certification for films & games.
• Fagen: elephants can learn welfare-friendly behaviours using positive reinforcers instead of physical punishment/intimidation ⇒ ↑ captive welfare & ↑ safety of those working with animals. SPR trunk-wash encourages protected contact (safety barriers) and helps identify a disease transmissible between humans & animals.
• Saavedra & Silverman: therapy based on classical conditioning can treat specific phobias — a distressing mental health condition that reduces quality of life. Disgust imagery exposure challenges fearful associations ⇒ long-term improvement from exposure therapies.

Nature versus nurture

Study	Nurture	Possible nature
Bandura	situational influence of models on acquiring aggression (and suppression by the non-aggressive model); boys may have acquired stereotypes or been rewarded for masculine-type behaviour	boys’ greater imitation could be due to hormones / brain development
Fagen	human training shaped behaviour with rewards ⇒ elephants learned trunk washing	adult elephant not shaped: decline in natural ability to learn with age, personality difference, trunk weakness or vision problems
Saavedra	classical conditioning = purely nurture: phobias are not innate or inherited; products of experiences with neutral stimuli; treatment follows the same principles (+ cognitive therapy)	—

Individual and situational explanations

• Bandura: imitation suggests situational factors (the model is part of the situation; male vs female model differences). Individual factors: in operant conditioning boys & girls may be rewarded differently for the same behaviour — a daughter praised for not fighting, a son for ‘sticking up for himself’ ⇒ different acquisition even with same models.
• Fagen: most elephants changed behaviour through situational changes (mahouts’ SPR) — but not the whole group ⇒ individual differences (age, personality) affect ability to learn new behaviours.

The use of children in psychological research

• Bandura: children didn’t appear to be able to consent or withdraw; children are vulnerable ⇒ potential for distress. Head teacher was thanked (aware of the procedure) but no sign of parental consent. Guidelines: get parents’/guardians’ consent as well as the child’s own.
• Practical advantages of children: exposed to much less violence than adults ⇒ fewer extraneous factors (e.g. a bad day at work); more naïve (less likely to guess the aim) ⇒ greater potential for representative effects of the procedure.
• Saavedra: 9-year-old asked for consent + mother consented; study potentially highly distressing (real & imagined exposure); boy vulnerable (recognised mental health condition). But aim = improve quality of life ⇒ temporary distress may be justified.

The use of animals in psychological research

Advantages

Disadvantages

Bandura-type study on animals removes risk of short-term distress / long-term harm to children. Children might have thought they were ‘supposed’ to copy the adult; animals wouldn’t. Animals’ prior exposure to aggression can be controlled. Fagen: already in captivity & familiar with handlers; near-daily training over weeks would be impractical with humans; intermittent food rewards motivated the elephants (harder to motivate humans). No evidence of harm; housing/care sufficient (Fig 4.9).

Animals have biological sex differences but do not acquire sex-typed behaviours from cultural factors as children do ⇒ can’t test social influences. Psychologists must meet housing requirements for captive animals — is it right to keep animals in captivity for research?

6. Research Methods Toolbox

Study	Method / design	Data	Sample
Bandura	lab experiment, independent measures (matched), 3 IVs; controlled observation (one-way mirror)	quantitative (counts per 5-s interval) + qualitative (remarks)	72 children, 3–6 y, Stanford nursery
Fagen	controlled / structured observation, behavioural checklist (% pass), 80% pass rule	quantitative (% correct)	5 female elephants, Nepal stable
Saavedra	clinical case study: interviews, observation, Feelings Thermometer (0–8)	qualitative + quantitative (self-report)	1 boy, 9 y, Miami

7. Practice, Project & Exam Questions

Project — sex-stereotyped TV advertising

Media (celebrities, actors in TV ads) can be role models that teach stereotyped behaviours through observation & imitation.

1. Choose a media source (magazine for young people / 1–2 hrs of TV with ads).
2. Make a table: name of product | stereotypical | non-stereotypical | neutral (tick one) — e.g. laundry detergent ✗ stereotypical, banking services = neutral, sports drink = non-stereotypical.
3. Watch/review the ads (if in a group, agree how to categorise).
4. Write a short report: Aim (what were you trying to find out?) · Method (how did you collect data?) · Results (what did you find?) · Conclusion (did you find what you expected?).

Exam-style questions — how to score

Q [marks]	Answer pointers
1. Why social imitation is quicker than operant conditioning [2]	no need to experience reinforcement/trial & error yourself — just watch a model; a behaviour can be learned in one observation instead of building up through repeated rewards.
2. Why informed consent from children and parents [3]	children are vulnerable / may not understand the risks; study could cause distress (aggression, annoyance); consent lets them withdraw ⇒ protect from harm; parents/guardians legally responsible.
3. Two sex-typed behaviours [2]	(1) boys imitated physical aggression more (esp. of male model); (2) girls played more with dolls / tea sets / colouring; (3) comments: ‘ladies shouldn’t fight’.
4a. Choice of species & strain [2]	use the least sentient / most appropriate species & strain for the aim (e.g. captive, docile elephants already used to humans & handlers).
4b. Why positive reinforcement [2]	behaviour followed by a reward (banana) is repeated; avoids punishment ⇒ better welfare; voluntary participation.
4c. One benefit of Fagen [2]	method to diagnose TB quickly & voluntarily ⇒ treat elephants & protect humans; welfare-friendly alternative to punishment.
5. Two findings (Fagen) [4]	4 juveniles learned the trunk wash, adult didn’t (describe + data); mean success 36% → 89.3%; trunk-here needed more cues; E2 & E4 never passed steady.
6. Evaluate strengths & weaknesses (Fagen) [8]	3–4 each from the lists in Section 3 — each point: state → explain → link to study (small sample / replicable & reliable / mahouts silent ⇒ validity / quantitative data) + conclude.
7. Why Saavedra & Silverman was carried out [2]	to examine the role of evaluative learning / classical conditioning in treating a phobia; to see if imagery exposure could reduce disgust & distress.
8. Operant conditioning [1 + 2]	(a) learning through the consequences of our actions; (b) contingency management — boy rewarded by mother for handling buttons after gradual exposure ⇒ approached buttons more.
9. Conclusions (Saavedra) [2]	positive reinforcement changed behaviour; imagery exposure reduced fear & disgust (long-term); emotions/cognitions about disgust matter in learning new responses.

8. Full Chapter Cheat-Sheet

	Bandura et al. (1961)	Fagen et al. (2014)	Saavedra & Silverman (2002)
Theory	social learning	operant conditioning (SPR)	classical conditioning (evaluative learning), operant
Aim	learn aggression by observing a model? sex of model?	free-contact elephants trained to do trunk wash by positive reinforcement?	role of classical conditioning in fear & avoidance; can exposure therapy reduce disgust?
Method	lab experiment	controlled (structured) observation	clinical case study
Sample	72 children (36 B + 36 G), 3–6 y	5 female elephants (4 juveniles + 1 adult), Nepal	1 boy, 9 y, Hispanic American
Key procedure	aggressive / non-aggressive / no model → mild annoyance → 20-min test	whistle + banana; lure / capture / shape; chain; 80% pass	hierarchy; contingency mgmt; imagery exposure; Feelings Thermometer
Key result	aggression imitated; boys > girls; same-sex model preferred	4 juveniles learned; adult didn’t; 36% → 89.3%	behaviour ↑ but distress ↑ with reinforcement; imagery 8 → 5 → 3; no phobia at 6 & 12 months
Strength	controlled, matched, one-way mirror, reliable	replicable; mahouts silent; quantitative	valid, multiple methods, standardised scale
Weakness	small sample (6 per cell); similar children; harm	small, captive, docile, one stable; flexible sessions	one participant; bias; demand characteristics
Ethics	aggression & annoyance; no parental consent shown	well treated, voluntary, aim = TB diagnosis	consent from boy + mother; anonymity; distress justified by aim
Application	media restrictions, watershed, certificates	welfare, protected contact, TB detection	phobia therapy (disgust imagery)