Memory Research

Read out digits and participants were asked to recall-added a digit each time. Average digit span is 7+/-2.

STM has limited capacity.

+ = lab experiment/has been confimed by other research (replication).

- = conducted a long time ago so may have less control.

Participants asked to learn trigrams whilst completing a distractor task for different lengths of time (3,6,9,12,15 or 18 seconds).

As time increased, recall decreased.

STM has limited duration (18 seconds).

+ = lab experiment- control- reliable.

- = stimulus material is artificial.

Participants given lists of words either sounding the same or meanng the same. Recall tested immediately (STM) or after 20 minutes (LTM).

More errors were made with acoustic words in short term and more errors in similar sounding words after delay.

Shows that coding is acoustic in STM and semantic in LTM.

+ = lab study.

- = artificial tasks cannot be generalised to everyday tasks.

Showed participants yearbooks after 15 and 48 years. Did free recall and recognition tasks (recognise names from the faces).

Recognition and free recall was higher after 15 years. Dropped after 48 years.

LTM can last a very long time (forever).

+ = high external validity (real-life)

- = no control over confounding variables.

Gave participants lists of words and asked them to recall them in any order.

Words were remembered more at the start and the end. Couldn't remember middle words.

Shows there are two separate stores- STM and LTM.

+ = lab experiment/ shows support for MSMM

- = artificial tasks- not true to everyday memory tasks.

Underwent brain surgery for epilepsy- hippocampus was removed.

LTM never improved but STM was intact.

Shows separation between LTM and STM.

+ = supports MSMM

- = evidence from brain damaged patients   is not generalisable.

Dual Task=

Verbal reasoning task with additional distractor task.

More mistakes made when central executive was used for both.

Shows central executive has a limited capacity.

+ = supports WMM.

- = artificial tasks.

Word-length effect-

Participants learnt long and short words.

Short words easier to recall.

Shows phonological loop has a limited capacity.

+ = supports different subsystems in WMM.

- = artificial tasks.

Motorcycle accident.

Poor STM for verbal but not visual.

Shows STM has more than one store- visuo-spatial sketchpad was fine but phonological loop was damaged.

+ = supports existence of separate visual and acoustic systems.

- = evidence from brain damaged patients is not generalisable,

Brain infection.

Episodic memory affected (couldn't remember past events) but not procedural (could still play piano). Semantic memory relatively unaffected.

Shows LTM is made up of separate stores.

+ = supports existence of separate LTM stores.

- = evidence from brain damaged patients isn't generalisable.

Participants learned trigrams. Couterbalanced distraction task.

Remembered first set but not subsequent sets.

Old trigrams interfered with new ones.

+ = lab experiment/ has beem confirmed by other research.

- = artificial tasks.

Learnt list of words then given a second list to learn. Asked to recall the 1st list.

More difficult when words were similar.

New words interfered with old words.

+ = lab experiments controlled.

- = artficial materials.

Divers learned words on land and underwater. Then recalled words either on land or underwater.

Recall better when the context is the same, e.g. learnt underwater, recalled underwater.

Shows support for the encoding specifity principle.

+ = lots of evidence adds to validity.

- = in real life, environments are not that different (extreme)

Gave participants words to learn either when they had taken an antihistamine drug or not. Asked to recall either on or off the antihistamine.

Recall was better in the same state.

Shows support for the encoding specificity principle.

+ = lots of evidence adds to validity.

- = unethical to make participants drowsy.

Showed film clips of accidents and changed the verb in the question, e.g. about how fast were the cars going when they hit each other?'

When 'smashed' was used the speed estimates were higher.

This shows leading questions affect memory.

+ = real life applications.

- = use of students.

Showed clips of accident and changed 'a' or 'the' in the question-'Did you see the broken headlight?'

Participants were more likely to see the broken headlight when 'the' was used.

This shows leading questions canaffect memory.

+ = real life applications.

- = artificial tasks.

Showed clips from different perspectives of a crime and allowed participants to dicuss.

Found that 71% od participants mistakenly recalled aspects that they hadn't seen.

Memory Conformity= think others are right.

+ = supported by evidence.

- = demand characteristics/ not real life.

Participants in waiting room hear an argument. Two different conditions (man walks out with):

1. High Anxiety= knife and blood.

2. Low Anxiety= pen and grease.

Less recall in high anxiety condition.

This shows anxiety decreases recall because of weapon focus.

+ = controlled.

- = not real situation/ surprise rather than weapon focus.

Real-life shooting witnesses- saw shop owner shoot a thief.

Participants interviewed 5 months after the incident

High anxiety= better recall.

Optimum level for arousal- can increase performance.

+ = based on real life.

- = field study lacks control.

Compared different interview methods.

Cognitive interview gave better recall.

Cognitive interview results in memories that are more accurate.

+ = research support for effectiveness.

- = time-consuming.