Cognitive approach - case studies

Schema theory
Bartlett

aim: investigate how prior knowledge and cultural schemas affect the recall of a story

procedure:
- british ppt heard Native American folk tale (culturally unfamiliar)
2 conditions:

• repeated reproduction - recall story shortly after hearing it, then recall after longer intervals

• serial reproduction - one ppt recalls story to another, who recalls to another

changes in story were analysed

findings:

recall showed distortion:

• assimilation - story became more consistent with british cultural expectations (unfamiliar details changed to fit schemas)

• levelling - story became shorter (“unimportant” details were omitted

• sharpening - order of events changes, details + emotions added, more familiar terms used

conclusion:

• ppts kept main themes

• changed unfamiliar elements to match their cultural schema

• supports reconstructive memory

GRAVE

G - low, ppts were only british

R - low, no standardised instructions, no fixed delay before recall

A - high, supports reconstructive memory + how schemas shape recall

V - low internal (lack of control)

E - no ethical concerns

Schema theory
Brewer and Treyens

aim: investigate the role of schema in the encoding and retrieval of episodic memory

procedure:

• 86 uni ppts

• sat in professor’s office for 35s, told to wait

• room contained schema-consistent items, schema-inconsistent items and missing schema-expected items

assigned to one of 3 conditions:

• written recall + recognition = write everything remembered + rate items from a list (present/absent)

• drawing recall = given outline of room and asked to draw remembered objects

• verbal recognition = list of items read aloud, say if each was in the room

findings:

• schema-consistent items recalled better in written + drawing

• schema-inconsistent items not freely recalled, recalled in recognition tasks

• ppts showed false memories, recall missing schema-expected items

conclusion:

• memory is schema-driven + reconstructive (ppts relied on office schema to encode + retrieve info)

• schemas can cause false memories of expected items

• support schema theory

Tversky & Kahnemann (1986)

aim: to test how positive vs. negative framing affects decision-making

procedure:

307 US undergraduates chose between 2 programs in a hypothetical disease outbreak

conditions:

• +ve frame: options described as lives saved (A vs. B)

• -ve frame: options described as deaths (C vs D)

findings:

• +ve frame: 72% chose certain option (A) 28% chose risky option (B)

• -ve frame: 22% chose certain loss options (C) 78% chose risky option (D)

GRAVE:

G - low: western uni students (not generalisable to other ages/cultures)

R - high: highly standardised procedure (easy to replicate)

A - high: framing effects are used in marketing, public health messaging

V - high internal validity: controlled experiment, low external validity (it’s a hypothetical scenario)

E - hypothetical choices = minimal harm

Tversky & Kahnemann (1974)

aim: to test the anchoring effect, whether the first number seen biases people’s quick estimates of a calculation

procedure:

high school students split into 2 groups and given 5 seconds to estimate a multiplication

ascending condition: 1×2×3×4×5×6×7×8

descending condition: 8×7×6×5×4×3×2×1

researchers predicted the first number would act as an anchor

findings:

• ascending group estimated lower (median = 512)

• descending group estimated higher (median 2250)

• true answer is 40320, supports anchoring bias

GRAVE

G - low: small sample of high school students (may not represent full population)

R - high: simple, standardised experiment

A - low: task is artificial but it supports broader idea of anchoring biases

V - high internal validity (controlled conditions) but low ecological validity (estimating under 5s not realistic)

E - low concern: minimal risk

Glanzer and cunitz

aim: to test whether more time between words would increase recall of the start of a list (primacy effect) by allowing more rehearsal

procedure:

240 US army men

conditions:

• word read once, 3s

• word read once, 6s

• word read once, 9s

• word read twice, 3s

• word read twice, 6s

• word read twice, 9s

had 2min to write down all the words remembered

findings:

• longer intervals = increased recall for most words (supports primacy) but didn’t improve end of list recall (recency)

• repetition helped at 3s, no effect at 6s or 9s

• primacy due to rehearsal

Milner

aim: to understand the effects of HM’s brain surgery on memory

procedure:

case study using method triangulation:

• psychometric tests (IQ)

• direct observation

• interviews

• cognitive tests

findings:

• HM couldn’t form new episodic or semantic long term memory

• could hold brief info with rehearsal

• could learn procedural skills

• could make cognitive maps

• MRI showed damage in hippocampus, supports idea that memory is multiple systems

GRAVE:

G - low: it’s only one case

R - low: case studies aren’t replicable, but method triangulation increases reliability

A - high: important to understand different memory systems and role of hippocampus in LTM

V - high ecological validity

E - research maintained consent, confidentiality and protection from harm

Landry and Bartling

aim: test the WMM by seeing whether articulatory suppression reduces serial recall

procedure:

• 34 psych undergrads

• 7 letter strings using F K L M R X Q

Loftus and Palmer

aim: investigate whether the verb used in a leading question would influence the ppts estimates of the speed of the cars in a traffic accident

procedure:

• 45 students

• watched 7 short clips of car accidents

after each clip they:

• wrote brief description of the accident

• answered a questionnaire (including critical question abt speed)

critical question:

“how fast were the cars going when they [verb] each other?”

verb varied by condition

findings:

• more intense verb = higher the estimated speed

• stat test showed the differences were sat significant

• wording of question systematically influenced speed estimate

conclusion

• leading questions can distort memory

• supports idea that memory is reconstructive (can be influenced by post-event info)

GRAVE

G - low, student sample

R - high, lab experiment, standardised procedure

A - high, relevant to witness testimony, leading questions can distort what witnesses report

V - high internal, control of confounding variables, supports cause-effect rs between IV and DV

E - informed consent, no deception