28
March 23, 2012 Intro to Cognitive Science 1 Inf1: Intro to Cogni-ve Science Recall, forge6ng, and eyewitness misiden-fica-on Alyssa Alcorn, Helen Pain and Henry Thompson

Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

March 23, 2012 Intro to Cognitive Science 1

Inf1:  Intro  to  Cogni-ve  Science    

Recall,  forge6ng,  and  eyewitness  misiden-fica-on    

Alyssa  Alcorn,  Helen  Pain  and  Henry  Thompson  

Page 2: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

1.  Today  we  will  cover:  

1.  A  short  discussion  of  retrieval  and  recall  (aka  how  to  get  the  stored  informa9on  back  out  again)  

2.  Forge>ng  3.  Memory  research  and  eyewitness  misiden9fica9on  of  crime  

suspects  

Next  week's  tutorial  will  discuss  seman9c  memory  (plus  reviewing  other  memory  topics)  

3/23/12 Intro to Cognitive Science 2

Page 3: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

2.  First,  a  criminal  act    

We  will  discuss  eyewitness  misiden9fica9on  at  the  end  of  the  lecture,  but  will  begin  today  by  witnessing  a  (staged)  crime.  

Please  watch  carefully.  Just  like  a  real  witness,  it  may  not  be  clear  now  what  details  will  be  relevant  later.  

Open  University  video  with  Professor  Graham  Pike  to  introduce  topic  of  eyewitness  misiden9fica9on  

hNp://www.open.edu/openlearn/body-­‐mind/psychology/misiden9fica9on-­‐can-­‐you-­‐iden9fy-­‐the-­‐criminal  

(first  50  seconds  only)  

3/23/12 Intro to Cognitive Science 3

Page 4: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

3.  “Remembering”  stored  informa-on  Storing  informa9on  in  memory  is  of  very  liNle  use  unless  the  

informa9on  can  be:  

 -­‐    accessed  again  (retrieved    or  recalled)      -­‐    at  an  appropriate  9me  (aka  remembering)  

Retrieval  is  the  process  of  using  one  or  more  cues  (in  any  modality)  to  recover  a  target  memory,  bringing  the  target  memory  or  informa9on  to  conscious  awareness  

Recall  is  when  a  person  must  generate  items  from  memory.    This  is  effec1vely  a  search  process.    

3/23/12 Intro to Cognitive Science 4

Page 5: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

3.  “Remembering”  stored  informa-on  Examples  include:    

•  Cued  recall  in  which  a  cue  prompts  recovery  of    the  target  informa9on    

–  Running  into  your  friend’s  flatmate  (cue)  and  trying  to  remember  their  name  (target)  

– Memories  invoked  by  senses  and  context  – Was  X  in  the  pub  last  night?  

•  Free  recall,  in  which  no  cues  are  given.    –  For  example,  an  experiment  which  instructs  “write  down  all  the  words  you  remember  from  the  list  you  just  studied”  

– Who  were  all  the  people  in  the  pub  last  night?  Free-­‐recall  tasks  are  suscep9ble  to  sequencing  effects  -­‐  the  original  

order  of  item  presenta9on  affects  the  likelihood  of  later  recall.    3/23/12 Intro to Cognitive Science 5

Page 6: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

4.  Primacy  and  recency  effects  in  free  recall  The  primacy  effect  is  that  the  first  items  presented  in  a  list  or  

sequence  tend  to  be  beNer  recalled  than  subsequent  items  

The  recency  effect  is  that  the  last  items  in  a  list  or  sequence  tend  to  be  recalled  well.  These  may  be  eliminated  by  doing  “filler  tasks”  (such  as  coun9ng)  between  hearing  the  items  and  recall.    

Thus,  items  in  the  middle  of  a  list  are  generally  the  least  likely  to  be  recalled.  

The  memory  experiment  described  in  the  sta9s9cs  lab  was  a  free-­‐recall  task  in  which  the  lists  of  words  to  be  recalled  were  designed  to  elicit  “false”  recall  of  an  item  not  on  the  list    

Par9cipant  data  exhibited  both  primacy  and  recency  effects  to  some  degree:  

3/23/12 Intro to Cognitive Science 6

Page 7: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

4.  Primacy  and  recency  effects  in  free  recall  

3/23/12 Intro to Cognitive Science 7

Page 8: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

5.  Why  do  we  some-mes  fail  to  retrieve  informa-on?  Forge<ng  can  be  conceptualised  as  a  failure  of  memory  

retrieval,  though  it  is  difficult  to  determine  whether  informa9on  is  ever  really  “lost”  completely.  

Forge>ng  or  retrieval  failure  can  be  divided  into  two  categories:  

•  Incidental  forge<ng  or  uninten9onal  forge>ng  (our  everyday  use  of  this  word)  

•  Mo1vated  forge<ng,  which  includes  both  conscious,  inten9onal  forge>ng  and  uninten9onal  forge>ng  triggered  by  some  mo9va9on.  For  example:    –  Consciously  deciding  (or  being  told)  to  forget  something  

–  Limi9ng  retrieval  of  some  experience  (usually  a  nega9ve  one)  as  a  form  of  emo1onal  regula1on  

3/23/12 Intro to Cognitive Science 8

Page 9: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

5.  Why  do  we  some-mes  fail  to  retrieve  informa-on?  In  all  of  these  forms,  losing  informa9on  or  failing  to  retrieve  is  actually  beneficial  and  adap9ve  for  our  cogni9on  overall.  It  is  a  form  of  op9misa9on  for  the  memory  system.  

 -­‐  We  are  most  likely  to  be  able  to  retrieve  items  and  events  that  we  need  to  recall,  with  less  interference  from  rarely-­‐needed  or  no  longer  needed  items  and  cues.  

 -­‐    We  are  not  overburdened  by  recalling  endless  stressful,  trauma9c,  and  nega9ve  events  

3/23/12 Intro to Cognitive Science 9

Page 10: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

6.  Forge6ng  over  -me    

A  fundamental  rule  for  most  organisms  (that  have  been  studied  so  far)  is  that  forge6ng  increases  as  -me  progresses  

3/23/12 Intro to Cognitive Science 10

In  one  of  the  earliest  psychological  studies  (s9ll  cited  today!)  Hermann  Ebbinghaus  (1885/  trans.  1913)  used  himself  as  a  par9cipant  to  study  memory  for  nonsense  syllables  over  9me  

http://en.wikipedia.org/wiki/File:Ebbinghaus2.jpg

Page 11: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

6.  Forge6ng  over  -me    

•  He  chose  nonsense  syllables  as  the  s9muli  to  be  learned  in  an  aNempt  to  make  sure  that  the  content  did  not  affect  his  memory  for  it  

•  In  other  words,  that  he  would  not  form  associa1ons  between  learned  content  and  prior  knowledge  or  images  that  could  serve  as  cues  

Ebbinghaus  learned  169  lists  of  13  nonsense  syllables  (like  “LEV”  and  “BUP”)  

•  He  then  re-­‐learned  each  list  aler  varying  9me  intervals,  from  several  minutes  later  up  to  one  month  later.  

•  He  used  the  amount  of  9me  required  to  re-­‐learn  each  list  as  a  measure  of  how  much  was  forgoNen.  

3/23/12 Intro to Cognitive Science 11

Page 12: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

7.  The  Ebbinghaus  forge6ng  curve  

While  our  intui9ve  sense  might  be  that  we  keep  forge>ng  things  fairly  steadily  over  9me,  Ebbinghaus's  (much-­‐replicated)  results  show  that  this  is  not  the  case.  

Forge>ng  is  not  a  linear  rela9onship  between  9me  and  content  retained,  but  close  to  a  logarithmic  curve  

•  This  means  that  a  lot  is  forgoNen  very  soon  aler  learning  

•  Then  the  rate  of  forge>ng  slows  down  and  almost  stabilises  over  a  longer  9me  

3/23/12 Intro to Cognitive Science 12

Page 13: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

7.  The  Ebbinghaus  forge6ng  curve  

3/23/12 Intro to Cognitive Science 13

Page 14: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

8.  Applying  research  to  second-­‐language  learning  In  a  long-­‐term  example  with  more  naturalis9c  content,  Bahrick  

(1984)  tested  reten9on  for  foreign  language  grammar,  reading  comprehension,  recall  and  recogni9on  vocabulary  

•  587  par9cipants  who  studied  Spanish  1  to  50  years  previously  •  Gathered  addi9onal  informa9on  about  their  language  learning  

–  Level  of  original  training,  grades  –  Use  of  spoken,  wriNen  Spanish  language  (rehearsal)  since  training  

Memory  showed  exponen9al  decline  in  reten9on  for  first  3-­‐6  years,  then  stabilisa9on  in  reten9on  for  up  to  30  years!  

Results  like  this  suggest  that  there  is  some  level  of  memory  “permastore”  affected  by  original  training,  but  NOT  by  subsequent  rehearsal.  

3/23/12 Intro to Cognitive Science 14

Page 15: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

8.  Applying  research  to  second-­‐language  learning  

3/23/12 Intro to Cognitive Science 15

Page 16: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

9.  Eyewitness  iden-fica-on  and  misiden-fica-on  One  par9cular  area  of  interest  for  applying  memory  research  (and  

where  memory  researchers  may  be  contacted  for  an  “expert  opinion”)  is  in  rela9on  to  crime.    

3/23/12 Intro to Cognitive Science 16

As  part  of  an  inves9ga9on,  a  witness  to  the  crime  may  be  called  on  to  provide  informa9on  about  its  circumstances  and/or  to  iden9fy  a  perpetrator  from  a  selec9on  of  suspects.  Unfortunately...  eyewitness  tes9mony  is  subject  to  many  errors  and  distor9ons  both  during  and  aler  the  event.  

Page 17: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

10.  Open  University  Video  (Graham  Pike)  hNp://www.open.edu/openlearn/body-­‐mind/psychology/misiden9fica9on-­‐can-­‐you-­‐

iden9fy-­‐the-­‐criminal      (play  from  0:50  to  2  minutes)  

“Who  thought  witness  was  in  first  row  (A,  B,  or  C)?      Second  row?    

 Third  row?      No  idea  at  all?”  

(play  rest  of  video)  

3/23/12 Intro to Cognitive Science 17

Page 18: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

11.  Key  points  from  the  video  Pike  makes  several  points  that  suggest  that  eyewitness  evidence  

is  really  not  very  reliable.  

e.g.,  77%  of  a  set  of  200  cases  overturned  aler  DNA  appeal  (i.e.  wrongful  convic9ons)  had  some  component  of  eyewitness  misiden9fica9on  as  a  factor  in  the  original  verdicts.  

Pike  men9ons  several  factors  contribu9ng  to  unreliable  eyewitness  accounts  or  iden9fica9ons,  which  crucially  involve    procedures  and  circumstances  aaer  the  event  

•  Use  of  leading  police  ques9ons  with  witnesses  •  Presen9ng  suspects  as  a  group  (simultaneous  presenta9on)  

rather  than  one-­‐by-­‐one  (sequen9al  presenta9on)  

The  following  video  explains  more  about    factors  influencing  misiden9fica9on,  including  official  procedures  and  also  features  of  the  witness,  perpetrator,  or  event  

3/23/12 Intro to Cognitive Science 18

Page 19: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

12.  Second  video  with  Graham  Pike    hNp://www.open.edu/openlearn/body-­‐mind/psychology/iden9fica9on-­‐evidence  

3/23/12 Intro to Cognitive Science 19

Page 20: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

13.  More  reasons  for  unreliable  iden-fica-ons  The  Memory  textbook  has  an  en9re  chapter  on  eyewitness  

tes9mony,  and  discusses  various  causes  in  some  detail.    

The  issues  of  poten9ally  poor  police  prac9ce  discussed  in  the  videos  come  down  to  the  issue  of  distor1ng  or  interfering  with  informa9on  when  it  is  retrieved.    

However,  other  processing  factors  may  also  play  a  role  •  ANen9onal  biases  •  Witness  characteris9cs  and  emo9onal  state  

–  A  very  young  or  very  old  witness  tends  to  be  less  reliable  –  An  extremely  frightened  or  stressed  witness  is  also  generally  less  reliable  

•  Change  blindness  •  “Seeing  what  we  expect  to  see”  3/23/12 Intro to Cognitive Science 20

Page 21: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

14.  Acen-onal  biases  in  eyewitness  misiden-fica-on  To  aAend  to  something  means  that  we  expend  cogni9ve  resources  

on  it  and  process  its  features  more  deeply.  

May  be  suscep9ble  to  change  blindness  when  items  are  not  aNended,  also  unlikely  to  process  and  store  much  informa9on  about  unaNended  or  minimally  aNended  aspects  of  situa9on.    

AAen1onal  biases  can  mean  that  a  witness  is  likely  to  aNend  to  and  encode  some  aspects  of  a  scene,  event,  or  perpetrator  at  the  expense  of  encoding  others.  For  example:  

•  The  weapon  focus  effect:  Witnesses  may  dispropor9onately  focus  on  a  weapon,  if  one  is  present  during  crime,  and  might  not  aNend  to  key  details  like  a  perpetrator's  appearance.  

•  Race  and  age  biases:  In-­‐depth  processing  of  in-­‐group  members  +  more  superficial  processing  of  out-­‐group  members  mean  may  be  harder  to  iden9fy  a  perpetrator  who  is  dissimilar  to  the  witness  

 (see    Simons  &  Levin  1998  “door  study”)  3/23/12 Intro to Cognitive Science 21

Page 22: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

15.  Seeing  (and  remembering)  what  we  expect  to  see  Expecta9ons  about  events  can  change  how  we  interpret  them,  

and  what  we  encode  (and  later  can  retrieve)  about  them.  

BartleN  (1932)  developed  a  theory  of  recalling  informa9on  as  consis9ng  largely  of  reconstruc-ng  the  details  of  the  event,  using  not  only  what  was  actually  observed,  but  also  “packages”  of  knowledge  called  schemas.  

•  These  are  scripts  or  groups  of  facts  about  situa9ons,  loca9ons,  and  events  

– What  happens  when  we  go  to  bank    – What  bank  robberies  and  bank  robbers  are  like  

– What  criminals  are  like  

•  They  are  stored  as  related  informa9on  in  long-­‐term  memory  and  appear  to  be  accessed  as  a  whole  package  

3/23/12 Intro to Cognitive Science 22

Page 23: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

15.  Seeing  (and  remembering)  what  we  expect  to  see  

Schemas  can  lead  us  to  form  expecta-ons  about  events  and  may  systema-cally  influence  our  recall.  Thus,  recalling  an  observed  event  is  likely  to  combine  event-­‐specific  informa9on  with  general  schema9c  informa9on  that  may  be  irrelevant  or  contradictory  to  the  actual  circumstances.  

 Perhaps  more  importantly  for  many  issues  of  witness  iden9fica9on,  schemas  may  bias  us  to  interpret  incomplete  and  ambiguous  informa9on  in  a  certain  way.  

3/23/12 Intro to Cognitive Science 23

Page 24: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

15.  Seeing  (and  remembering)  what  we  expect  to  see  For  example,  Tuckey  and  Brewer  (2003b)  report  a  study  in  which  

“eyewitnesses”  to  a  simulated  bank  robbery  interpreted  ambiguous  informa9on  according  to  their  schema  about  bank  robbers:  

•  Male,  wearing  dark  clothing  and/or  wearing  a  disguise  

•  Schema  content  study  reported  in    Tuckey  and  Brewer,  2003a  When  the  simulated  bank  robber  was  gender-­‐ambiguous  with  a  

completely  covered  head  and  face,  the  witnesses  were  more  likely  to  report  informa9on  congruent  with  their  bank  robbery  schema  (a  male  robber)  rather  than  simply  repor9ng  an  ambiguous  figure.  

3/23/12 Intro to Cognitive Science 24

Page 25: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

16.  Closing  thoughts  

There  is  some  debate  about  the  degree  to  which  empirical  studies  of  memory  apply  to  real  crimes  and  real  witness  iden9fica9on  situa9ons  (see  the  Memory  textbook,  chapter  13).    

While  Tuckey  and  Brewer's  experiments  are  a  laboratory  example  about  an  uncommon  type  of  crime,  consider  the  extremely  serious  implica-ons  of  cultural  schemas  that  tell  us  that  “criminals”  are  likely  to  be  males  from  ethnic  minori-es,  or  that  any  teenager  in  a  hooded  sweatshirt  is  an  ASBO  wai-ng  to  happen.  

3/23/12 Intro to Cognitive Science 25

Page 26: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

16.  Closing  thoughts  

Hopefully  our  memory  unit  has  begun  to  illuminate  how-­‐-­‐while  memory  is  largely  robust  and  adap9ve-­‐-­‐it  can  fail  in  significant  ways  both  for  individuals  like  Clive  Wearing,  and  as  an  informa9on-­‐processing  system  that  is  suscep9ble  to  considerable  distor9ons  and  processing  biases.  

Our  models  of  memory  and  our  aNempts  to  endow  ar9ficial  intelligences  with  human-­‐like  processing  capaci9es  must  consider  and  account  for  these  weaknesses,  as  well  as  strengths.  

3/23/12 Intro to Cognitive Science 26

Page 27: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

References  

Course  texts  

•  Memory  (Baddeley,  Eysenck,  &  Anderson,  2009)  Other  resources,  available  through  the  library  or  Google  Scholar  

•  Seman9c  memory  content  in  permastore:  Fily  years  of  memory  for  Spanish  learned  in  school  (Bahrick,  1984).    

•  Memory:  A  contribu9on  to  experimental  psychology  (1885,  translated  1913).    An  online  version  is  available  at  hNp://psychclassics.yorku.ca/Ebbinghaus/index.html  

3/23/12 Intro to Cognitive Science 27

Page 28: Recall,&forge6ng,&and& eyewitness&misiden-fica-on&

References  •  Professor  Graham  Pike's  videos  on  eyewitness  misiden9fica9on,  through  Open  University  

–  First  video  hNp://www.open.edu/openlearn/body-­‐mind/psychology/misiden9fica9on-­‐can-­‐you-­‐iden9fy-­‐the-­‐criminal  

–  Second  video  hNp://www.open.edu/openlearn/body-­‐mind/psychology/iden9fica9on-­‐evidence  

•  How  schemas  affect  eyewitness  memory  over  reprated  reteieval  aNempts  (Tuckey  &  Brewer,  2003a)  

•  The  influence  of  schemas,  s9mulus  ambiguity,  and  interview  schedule  on  eyewitness  memory  over  9me.  (Tuckey  &  Brewer,  2003b)  

3/23/12 Intro to Cognitive Science 28