Textbook Notes (368,125)
Canada (161,663)
Psychology (1,112)
PSYC 334 (10)
Chapter

PSYC334 Ch 15.pdf

9 Pages
89 Views
Unlock Document

Department
Psychology
Course
PSYC 334
Professor
Ronald R Holden
Semester
Winter

Description
PSYC  334   Chapter  15   The  Problem  of  Bias  in  Psychological  Assessment     -­‐ Mean  difference  of  1  SD  in  IQ  between  black  and  white  students   -­‐ Drops  to  .5  –  .7  when  SES  is  controlled  for   -­‐ Controversial     Cultural  test  bias  hypothesis  (CTBH):  differences  in  the  mean  test  scores  across  gender  or   ethnic  groups  are  due  to  artifacts  of  the  test  or  measurement  process  and  do  not  reflect  real   differences  among  the  groups  on  those  constructs.   -­‐ Flawed  psychometric  methodology,  not  actual  differences       Bias:  systematic  error  in  a  test  score.       If  a  test  has  a  bias  related  to  a  nominal  variable  (e.g.,  gender  or  race)  it  is  said  to  have  a  cultural   bias.     Native  Americans  tend  to  perform  worse  at  verbal  tasks  than  white  Americans.     Both  Hispanic  and  Native  Americans  are  better  at  visual-­‐spatial  tasks  relative  to  verbal  tasks.     15  :  1   PSYC  334       Explaining  mean  group  differences.     Most  common  explanations     1. Genetic     2. Environmental  (e.g.,  SES,  education,  culture)   3. Interaction  between  genes  and  environment   4. Defective  tests   15  :  2   PSYC  334   Recall  that  cultural  test  bias  hypothesis  posits  that  mean  group  differences  are  due  entirely  to   defective  tests  and  do  not  represent  any  real  difference  in  the  latent  psychological  construct  (or   ability,  etc).       Hypothesis  of  differential  validity:  tests  measure  intelligence  of  the  group  upon  which  the   questions  are  based,  and  are  therefore  biased  against  all  other  groups.  For  example,  most  tests   that  we  use  are  based  on  white,  Western,  middle-­‐class  knowledge  and  values.     ...     Bias   Fairness       Bias:  statistical  property  of  a  test.     Fairness:  moral,  philosophical,  or  legal  issue  on     which  reasonable  people  can  disagree.         -­‐ Determined  empirically  (through   -­‐ Determined  through  principles,  opinion,   discrepancies  in  data)   and  debate         Distinguish  between  bias  and  offensiveness.   -­‐ Minority  panel  to  review  offensiveness   -­‐ Hard  to  find  biased  items  on  gross  inspection         15  :  3   PSYC  334   Test  bias  and  inappropriate  administration  or  use   -­‐ Tests  need  to  be  administered  by  skilled,  trained,  and  sensitive  professionals   -­‐ Need  to  use  tests  for  what  they  are  intended     The  cultural  test  bias  hypothesis  has  no  way  of  controlling  for  nonintellectual  influences  acting  on   individuals  who  take  exams.     -­‐ Emotional,  motivational  factors   -­‐ Hard  to  separate  individual  from  group  performance     Psychological  processes  are  internal  in  the  organism.     -­‐ Therefore  not  directly  observable  or  measurable  (must  be  inferred  from  behaviour)     Cultural  test  bias  is  a  subset  of  a  problem  seen  in  general  psychological  testing:  there  will  always   be  some  uncertainty  –  i.e.,  we  can  never  guarantee  that  what  we’re  measuring  is  what  we  think   we’re  measuring.     Also  do  not  allow  tests  to  be  pooled  together  into  umbrella  categories.  To  say  that  “mental  tests”   are  culturally  biased  is  not  helpful.  Look  at  specific  tests.     Objections  to  the  use  of  educational  and  psychological  tests  with  minority  students.     Association  of  Black  Psychologists  (ABPsi)     ABPsi’s  stance  on  psychological  testing:  refuse  children  to  be  subjected  to  achievement,   intelligence,  aptitude,  and  performance  tests,  which  have  been  used  to...   -­‐ Label  Black  people  as  uneducable   -­‐ Place  Black  children  in  “special”  classes  and  schools   -­‐ Perpetuate  inferior  education  of  Black  people     -­‐ Assign  Black  children  lower  educational  tracks  than  White  children   -­‐ Deny  Black  people  higher  education  opportunities   -­‐ Destroy  intellectual  growth  and  development  of  Black  people       Most  common  objections  to  the  use  of  educational  and  psychological  tests  with  minorities   1. Inappropriate  content:  different  upbringing  means  exposed  to  different  things.   2. Inappropriate  standardization  samples   3. Examiner  and  language  bias:  intimidating,  accuracy  of  communication?   4. Inequitable  social  consequences:  assignment  to  dead-­‐end  educational  tracks.   5. Measurement  of  different  constructs:  do  not  measure  minority  intelligence.   6. Differential  predictive  validity:  predict  outcomes  for  White  children  but  not  for  minority   children.     7. Qualitatively  distinct  aptitude  and  personality:  since  majority  and  minority  groups  have   different  aptitude  and  personality  characteristics,  testing  of  the  groups  should  begin  with   different  definitions  for  the  different  groups.     Differential  predictive  validity:  occurs  when  a  test  accurately  predicts  outcomes  for  one  group   but  makes  inaccurate  predictions  for  another  group.     15  :  4   PSYC  334       .......     Most  models  of  uncovering  test  bias  come  from  the  decision-­‐making  process  after  testing  has   completed,  and  do  not  look  at  the  test  itself.  This  then  becomes  a  matter  of  validity.  Recall  the   issue  of  differential  predictive  validity.       Differential  validity:  occurs  when  deficiencies  in  the  test  itself  give  rise  to  different  meanings  of   scores  earned  by  members  of  different  identifiable  subgroups.       I  guess  this  is  similar  to  differential  predictive  validity,  which  is  a  more  specific  subtype  of   differential  validity.       Cultural  loading:  degree  of  cultural  specificity  in  items  or  a  test  as  a  whole.     Cultural  loading  is  not  the  same  as  cultural  bias.  A  test  can  be  culturally  loaded  without  being   culturally  biased.       Greater  cultural  specificity  does  increase  chances  of  bias,  but  the  terms  are  not  synonymous.       Culture-­‐free  intelligence  tests  have  been  developed,  but  are  not  psychometrically  sound.   -­‐ Intelligence  can  be  tied  to  knowledge  that  is  most  important  and  relevant  in  that  culture               15  :  5   PSYC  334   Inappropriate  indicators  of  bias:  mean  differences  and  equivalent  distributions.     Mean  differences  as  bias:  some  claim  that  when  there  are  no  a  priori  differences  between   groups,  differences  in  the  means  of  the  groups  identify  b
More Less

Related notes for PSYC 334

Log In


OR

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


OR

By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.


Submit