Reconstructive memory

Definition of reconstructive memory 

Memory isn’t passive retrieval but recreation of events every time its rmb

• incl omission/addition of details to recalled event based on indiv personal exp

• Recreation of memories influenced by various factors

  • Perception

  • belief

  • past exp

  • cultural factor

  • current context

Memory - inaccurate due to:

• leading qns used by police after event

• meant to guide person to the answer questioner wants

• info received after event can have retroactive interfering effect - later findings interfere w previous learning

Loftus and Palmer (for RM)

Aim:

• investigate if memory can be altered to be misleading post-event using leading qns

Method: 

• Lab expt

Procedure:

• 1st expt

  • 45 american students

  • Shown vids of car crashes lasting 5-30s

  • R used 5 verbs to ask question abt speed which cars contacted

    • Hit, smashed, collided, bumped, contacted

  • Variables

    • IV: misleading post-event info, operationalised as emotional intensity of verb in qn

    • DV: speed est. 

  • Note: p’s ans might be response-bias factors

    • misleading info provided - simply influenced answer person gave

    • BUT did not actl lead to false memory of event 

• Expt 2:

  • 150 p shown vid of car accident

    • asked to complete questionnaire

  • randomly assorted to 3 diff grps w diff qns

    • “smashed into eo”

    • “hit eo”

    • no critical qn (control)

  • After week, asked if they saw any broken glass

Results:

• Expt 1: less intense verb - est lower speed than more intense word

  • Smashed → collided → bumped → hit → contacted

• Expt 2:

  • emotional intensity of verb in leading qn → influence probability p reported seeing broken glass 

  • provide support for schema + reconstructive memory 

Conclusion:

• p utilised schema to est speed of car crash → rely on intensity of verb provided in critical qn

• pre-existing knowledge knowledge regarding verb intensity → misrep of actual car speed

• schema influence cognitive processing 

  • contribute to distortion of memory when receiving and encoding info

Loftus + Palmer eval

Strength

• Standardised procedure

  • Methodology to carry out expt - easy to follow + easily replicable  

• increase consistency + comparability in research → > reliable + valid results 

Limitations

• small sample size → lower pop valid 

loftus + pickrell (for reconstructive memory)

Aim:

• determine if false memories of autobiographical events - created through power of suggestion

Method:

• p - 3M, 21F

• before study

  • parent/sibling of p contacted → asked 2 qns

    • retell childhood memories of p

    • do you rmb time when p was lost in a mall

• p received questionnaire in mail

  • 4 memories they were aseked to write abt 

  • mail back questionnaire to psychologists

• 3 events were real, one was them getting lost in the mall

  • instructed if they didn’t rmb event → write “i dont rmb this”

procedure: 

• p - interviewed twice over 4 weeks

• asked to recall as much info they cld abt 4 events

• asked to level of confidence on scale of 1-10

• after 2nd interview - debriefed → asked if they cld guess which of memories was false

results:

• abt 25% of p “recalled” false memory

• ranked memory as less confident than other memories

• wrote less abt memory on questionnaire 

conc:

• memory is easily distorted 

• ext factors such as suggestion → significantly distory/create new memories

• underscore idea that memory is not a passive process of retrieval 

  • instead active reconstruction by various factors 

yuille and cutshall

against RM → argued memory might be more reliable

Aim: 

• investigate accuracy of recall of eye witness of real crime → in response to leading qn over time

Method + procedure:

• aft 4 months - interview 13 ppl who witnessed armed robbery in Canada

• incl 2 misleading qns

  • 1st qn (getaway car had no broken headlight)

    • 1st half of grp → saw “a” broken headlight on getaway car

    • other half → saw “the” broken headlight 

  • 2nd qn (panel colour on car)

    • 1st half → saw “the” yellow panel

    • other half → saw “a” yellow panel

Results:

• misleading qns had little effect on recall

• majority of p said there was no yellow panel + broken headlight 

Conc:

• eye witness reliable

  • recall large no. of details despite misleading qn

• directly involved w event had more emotional influence → rmb more than lab expt 

• argued it wld be hard to generalise findings as it was a unique event 

Yuille and Cutshall

strength:

• high eco valid 

• used to criticise reconstructive memory

limitation

• low reliability, cannot be replicated 

• EV cannot be controlled

• ungeneralisable 

loftus and pickrell eval

strength:

• high eco valid 

  • ppl talk abt childhood memmories

• memories can be verified by family members

limitations:

• doesn’t tell why some p more suspectible to false memories than others

• ethical concerns → deception used in study

• demand char

  • social desirabilty effect 

Reconstructive memory eval

Strength

• real-life eyewitness studies → memory can be more accurate than lab findings

• Field studies → people often recall central details of real events accurately

  • Despite exposure to misleading info → emotional + personal relevance strengthen memory

• Increase ext validity of reconstructive memory theory

  • Show while memory is reconstructive → still reliable in meaningful real world situations

Limitation:

• research is that its lab-based → findings lack eco valid 

• Lab → inaccuracies in recall → minimal real-world consequence 

  • p nto exposed to emotional stress of witnessing real crimes/accidents

• limits how findings can be generalised to real-life eyewitness testimony 

  • emotional + personal involvement affect memory diff

• demand char → affect recall accuracy

• p do not expect to mislead them → change ans to fit perceived expectation rather than reflecting genuine memory distortion

• recall inaccuracies lab-based studies - not rep true memory errors

  • p compliance, reduce int valid