TASK 4: OUR LEGAL JUDGMENT IS ALWAYS FAIR. ISN’T IT?
LEARNING GOALS:
1. What are implicit attitudes and stereotypes?
2. How can we test for implicit bias?
3. Which biases/factors play a role in the legal system?
4. How do these biases influence legal decision making in court?
5. Can we reduce the influence of this bias?
Greenwald & Krieger (2006) – Implicit Bias: Scientific Foundations
Explicit belief: consciously endorsed.
- An intention to act is conscious if the actor is aware of taking action for a particular reason
o One may dissemble and deny that they’re taking an action for a particular reason, so
conscious intentions based on explicit beliefs may be hard to verify
o But a deceitful actor is nevertheless capable of asserting the belief or identifying the
intention that provides the basis for action, even when unwilling to do so
- The science of implicit cognition suggests that actors do not always have conscious,
intentional control over the process of social perception, impression formation, and
judgment that motivate their actions
Implicit attitudes & implicit stereotypes are especially relevant to bias and discrimination.
- Attitude: an evaluative disposition; the tendency to like or dislike, or to act favourably or
unfavourably toward, someone or something
o Explicit expressions of attitudes occur frequently, whenever we say we like or dislike
something or someone
Attitudes can also be expressed through favourable or unfavourable action,
such as by voting for or against a particular presidential candidate
o Implicit attitude indicator: an action that indicates favour or disfavour toward some
object but is not understood by the actor as expressing that attitude
As an example, consider how people form impressions of a liked or disliked
person’s spouse, child, or sibling. Someone who knows nothing about the
person’s relative other than the relative’s relation to the person may find
that they like or dislike the relative. This attitude toward the relative is likely
to match the attitude toward the person
Here, the “implicit” designation indicates that the attitude expressed
toward the candidate determined the attitude toward the relative,
even though the liking or disliking for the relative may be
experienced as an independent attitude
o Dissociations between implicit and explicit attitudes: the discrepancy between
implicit and explicit attitudes toward the same object
Commonly observed in attitudes toward stigmatized groups, including
groups defined by race, age, ethnicity, disability, and sexual orientation
The IAT is used to reveal such attitudinal dissociations
- Social stereotype: a mental association between a social group or category & a trait
o The association may reflect a statistical reality, but it need not. If the association
does reflect a statistical reality, members of the group will be more likely to display
the trait than will members of other groups
A perfect or near-perfect correlation, which might be a defining trait – such
as physical stamina for basketball players – is of little psychological interest &
is often not even considered part of a stereotype
, o Implicit stereotype: the introspectively unidentified (or inaccurately identified)
traces of past experience that mediate attributions of qualities to members of a
social category
Stereotypes can involve associations of either favourable or unfavourable
traits with a group
- The difference between attitudes and stereotypes
o For stereotypes, the content of the ascribed trait, rather than its evaluative valence,
is central
o For attitudes, the opposite holds
Bias/response bias: a displacement of people’s responses along a continuum of possible judgments.
- Need not indicate something unwise, inappropriate, or even inaccurate
o Biases can be either favourable or unfavourable
- Implicit biases are discriminatory biases based on implicit attitudes or implicit stereotypes
o Implicit biases are especially intriguing & problematic, because they can produce
behaviour that diverges from a person’s avowed or endorsed beliefs or principles
o An attitudinal bias might be that one has a more favourable attitude toward one race
group than toward the other
o If, among equally qualified renters, one assumes that members of one race will be
more conscientious in paying rent than those of another, this may be a bias rooted in
stereotype
- Ingroup bias: favouritism toward groups to which one belongs
o Bias is a problem only when it is directed against some group. Thus it may be
considered acceptable to be biased in favour one one’s siblings, children,
schoolmates, and friends
o The intuition that biases in favour of one’s smaller ingroups (such as family & friends)
are acceptable typically does not extend to believing that biases favouring one’s
larger ingroups (one’s race, sex, ethnicity, religion, or age group) are appropriate
The illegality of some kinds of biased behaviour toward certain groups
(regardless of one’s membership) – such as those defined by race, sex,
ethnicity, religion, and age – provides a non-psychological boundary
Psychologically, the small size of some ingroups is no doubt significant
because many people feel more obliged to help others when they’re one of
only a few people who can possibly be helpful, as may often be the case for
family members
o The situations in which people wish to be biased in favour of their smaller, important
ingroups – such as providing care for their own children – are often those for which
no question of possible discrimination arises
Nevertheless, a positive attitude toward any ingroup necessarily implies a
relative negativity toward a complementary outgroup. In some
circumstances, this relative favouring of the ingroup puts members of other
groups at a discriminatory disadvantage, as when one allows favouritism
toward a family member or friend to influence a hiring, job assignment,
rental, or admissions decision
Implicit Association Test (IAT)
- To measure a wide variety of the group-valence & group-trait associations that underlie
attitudes and stereotypes
o IAT = an implicit measure, because it infers group-valence and group-trait
associations from performances that are influenced by those associations in a
manner that’s not discerned by respondents
, - The most widely used IAT is the “Race IAT” which assesses implicit attitudes toward African
Americans (AA) relative to European Americans (EA)
o First, respondents practice distinguishing AA from EA faces by responding to faces
from one of these two categories with the press of a computer key on the left side of
the keyboard & to those of the other category with a key on the right side
o Next, respondents practice distinguishing pleasant-meaning from unpleasant-
meaning words in a similar manner
o The next two tasks, given in a random order, use all four categories: AA faces; EA
faces; pleasant-meaning words; and unpleasant-meaning words
In one of these two tasks, the IAT calls for one response (say, pressing a left-
side key) when the respondent sees AA faces or pleasant words, whereas EA
faces and unpleasant words call for the other response (right-side key)
In the remaining task, EA faces share a response with pleasant words and AA
faces with unpleasant words
o The implicit-attitude measure produced by this IAT is based on relative speeds of
responding in the two four-category tasks
This measure allows an inference about attitudes (category-valence
associations) because it’s easier to give the same response to items from two
categories when those two categories are cognitively associated
For American respondents taking the Race IAT, response speeds are often
faster when EA, rather than AA, is paired with pleasant words. This
frequently observed pattern supports the interpretation that EA-pleasant is a
stronger association than AA-pleasant
=> An implicit attitudinal preference for EA relative to AA
- There is systematic variation in the agreement between IAT (implicit) measures compared
with parallel survey-type self-report (explicit) measures:
o There’s substantially greater agreement between the two types of measures when
implicit & explicit attitudes have been shaped by the same experiences, which is likely
to be the case for attitudes toward consumer brands, sports teams, and politics
o When implicit & explicit measures of attitudes or stereotypes disagree, there’s said
to exist a dissociation between the two
For example, when a Race IAT shows preference for EA & a self-report
measure shows impartiality
- The IAT is extended into increasingly diverse domains, applying its general method to a wide
variety of groups and social categories. Findings from a meta-analysis on IAT & behaviour:
o Implicit (IAT) & the parallel explicit measures display predictive validity, meaning
that both types of measures are significantly correlated with measures of behaviour
o Implicit measures of bias have relatively greater predictive validity than explicit
measures in situations that are socially sensitive, like racial interactions, whereas
impression-management processes might inhibit people from expressing negative
attitudes or unattractive stereotypes
Additionally, implicit measures have relatively greater validity in predicting
spontaneous behaviours such as eye contact, seating distance, and other
such actions that communicate social warmth or discomfort
- Since 1998, IAT measures of implicit attitudes have been available on the internet for self-
administered use. These data came from voluntary visitors to the IAT website – a self-
selected sample, which is different from a representative sample that can be obtained
randomly from a defined population. These web-accessible demonstrations have
accumulated sufficient data to allow researchers to draw some conclusions:
o Explicit measures generally show much greater evidence for attitudinal impartiality
or neutrality. Averaged across the dozen topics, 42% expressed neutrality on explicit
