Stereotypes are always inaccurate Critically discuss
Does a person’s view of the qualities that members of particular groups have, reflect the reality of life, or are errors being made when groups and behaviours are observed? A definition of stereotypes with the fewest constraining assumptions is that they are qualities seen to be associated with particular groups or categories of people. Four key approaches, namely kernel of truth, illusory correlation, cognitive miser and self-categorisation theory will be closely looked into. I will attempt to answer a set of key questions that may help to address the situation. Firstly, How does each approach attempt to explain the origins of stereotypes? On the basis of their assumptions, where do these four positions stand on the issue of stereotype accuracy? Do they say stereotypes are always, sometimes or never inaccurate? This will allow me to relate the theories to each other and hence provide a better understanding on the question of accuracy.
Kernel of truth – KOT
This explanation of stereotypes assumes that there is some aspect of social reality distorted in the content of stereotypes. It assumes that a groups culturally distinctive behaviour or the socio-economic position it finds itself in, could give rise to certain stereotypical perceptions of the group being displayed. For example, a particular ethnic group may have an economically disadvantaged position in society. They may receive poor wages, have high unemployment, live in poor housing conditions and not achieve much in education. Because of these situations, it’s easy to see how people can perceive the group as, ‘poor’, ‘lazy’ and ‘stupid’. Once people start to categorise in this way, they knowingly make certain distinctions. They then exaggerate the attributes given, to the extent that they become stereotypes.
For KOT, stereotypes are partly accurate because they may be based upon real and important differences between groups. Stereotypes must have some sort of, ‘grain of descriptive truth’ about them or else we wouldn’t see the existence of so many differences between groups. There must be certain characteristics that are specific to particular groups based on culture, traditions or other aspects, which are true for that group only and reflect the social reality of being a member.
To see whether the kernel of truth idea truly existed, researchers observed whether the characteristics of a particular stereotyped group reflected how people saw that group. It was assumed that this would show the accuracy of stereotypes because they were comparing stereotypes, with information about personality characteristics of group members via self-report measures. Abate and Berrien (1967) assessed how Japanese and American students saw the stereotypes of their own nations (self-stereotypes) and that of the other nation (hetero-stereotypes). This was compared with the students’ responses on the Edward’s Personal Preference Schedule (EPPS). The accuracy of stereotypes was dependant upon the degree of correlation between the stereotypes and the EPPS score (the ‘vereotypes’).
Results showed agreement between self and hetero-stereotypes, but less agreement between the EPPS responses and both national self-stereotypes and national hetero-stereotypes. The respondents couldn’t even perceive their own national group accurately never mind a foreign group. This questions how useful these, ‘vereotypes’ are at reflecting stereotypes. Also, self-reports and ‘objective’ measures have come under many criticisms, which have lead to the conclusion that accuracy correlations such as these are hard to interpret.
Evaluation of the kernel of truth approach
It seems that the kernel of truth sees measuring individual-level characteristics as a more accurate way of assessing stereotype accuracy. How can you find the truth about people and hence stereotypes in their personality? We want to look at individual people as members of larger groups, not the stereotype of individual people. Katz and Braly emphasize the need to observe individuals and groups as separated levels,
“There are no racial or national groups which exist as entities and which determine the characteristics of the group members”
(1933, pg. 289, citation in Oakes and Reynolds, (1997) pg. 53)
They say that since this approaches view of how stereotypes can be measured is inaccurate, so must be their assumptions regarding its development.
A different approach taken to see whether there is a kernel of truth about stereotypes was by observing the convergence of the self-stereotype with that held by another group. Though some success has been found with this, it has been argued that the group being stereotyped simply takes in the stereotype that the other group holds of them, which is a general problem. Also, accuracy depends on whether this convergence can be accepted by the stereotyped group in terms of their beliefs. In addition, the factual validity becomes controversial when a given group assigns unfavourable traits to itself.
And so, do stereotypes have a ‘kernel of truth’? Evidence against it has by far outweighed evidence for it. However, they do say that stereotypes are partly accurate and that there is room for an individuals experiences to change their views, on top of the socialised fact. In many cases, increased contact has correlated with stereotype accuracy. However, convergence of self-stereotype and the stereotype of the group held by others should be understood by a shared understanding of social reality rather than outright stereotype accuracy
Illusory correlation – I-C
Research shows that people seem to perceive minority groups more negatively than majority groups, even though the two groups behave the same way. Illusory correlation explanation of this effect is that negative behaviours become associated with minority groups because of their shared distinctiveness. When people observe two rare events (minority group and infrequent act), they have the tendency to put the two together and see them as correlated. This results in perceivers assuming that the behaviours observed are diagnostic of a person or group and use this idea in future impressions.
To support this idea, Hamilton and Gifford (1976) carried out an experiment where subjects had to read 39 sentences. Each sentence described a behaviour of a person in either Group A or Group B. The experimental conditions ensured that both Group B members and undesirable behaviours were made distinctive. The participant’s task was to recall the statements by assigning the actions to one of the two groups. Results showed that the desirable actions were quite accurately assigned but undesirable ones were biased whereby more were assigned to the minority group. In the lab, it seems to work all the time and so it was concluded that,
“The illusory correlation in group perception is a quite reliable phenomenon”
(Hamilton and Sherman 1989, p.69)
I-C claims that stereotypes are always inaccurate. This is because stereotypes are formed on the basis of false combinations and so the likelihood of error is potentially high. People attempt to estimate difficult co-variations and so are prone to making errors when judging the relationship between two characteristics or events. This implies that their judgements are wayward, and so the whole formation of their stereotypes is inaccurate. As such, their associations lack any aspect of social reality within them.
Unlike KOT, this view can account for the many examples of stereotypes that are not statistically accurate. A typical example is that people have been accustomed to the stereotype of seeing AIDS as being a homosexual disease. This disease is infrequent and homosexuals are a minority group. When the two co-occur, the stereotype is formed. However, this isn’t true statistically as the incidence of AIDS is actually higher in heterosexual women. It’s hard to see how KOT can explain this using the idea of characteristics being specific to particular groups, based on cultures or traditions. Further support comes from the finding that people don’t even have to be presented with infrequent information about a target group for a stereotype to form. When both the majority and minority groups are described by positive information, participants still favour the majority group. For example, the illusory correlation of African-Americans and welfare recipients when most recipients are white.
For I-C, the extent to which stereotypes are accurate is dependant upon the truthfulness of the correlations. Considering that most, if not all, cases of correlations are not true, the accuracy of stereotypes must be non-existent. The human mind is not always an accurate perceiver of the external world and the illusory correlation effect is just another example of mistakes that we all make everyday.
Evaluation of the illusory correlation
Unlike KOT, this view seems to ignore the traits of other groups in shaping our beliefs, and the role of society, as important aspects of stereotype development. This questions the role of illusory correlation when people form stereotypes in the real world. Hamilton and Gifford (1976) used abstract, artificial groups in order to demonstrate a cognitive bias because they didn’t want any previously formed beliefs that people would’ve associated if actual social groups were used. As demonstrated by KOT, in the real world, people do carry views and beliefs about various groups and the extent to which they display their beliefs changes. Therefore, it’s not just a case of combining events and behaviours.
For I-C to claim that stereotypes are inaccurate would mean that they develop under conditions where the perceiver has no pre-existing beliefs about the particular group in question. For adults, these conditions are extremely rare, yet they continuously form new stereotypes of groups that they have never encountered. McArthur and Friedman (1980) showed, using real groups, that the illusory correlation effects were only found when the group acted in a way that was normative for them according to pre-existing beliefs.
Research has shown that members of minority groups associate themselves with more frequent behaviours, not less and with positive acts, as shown by Sanbonmatsu et al (1987) and Schaller and Maass (1989) respectively. For I-C, this is problematic since their argument is based on infrequent and negative acts. McGarty et al (1993) have shown that just by saying to participants that either there will be twice as many statements about Group A or that half the statement will feature desirable behaviours by Group A, yields the illusory correlation effect. Since I-C claim that it’s the way we filter and store information, a big problem arises because McGarty hasn’t even presented any information! Therefore, how can the I-C approach claim that stereotypes are inaccurate if contrary results, as well as alternative processes are being observed in studies?
However, the I-C may have a stronger role in the eyes of a child since they are yet to be exposed to the social world so have few, if any, pre-existing beliefs about groups. Also, if there were some, ‘kernel of truth’ on which groups were seen differently, the illusory correlation might contribute to overgeneralising how intergroup differences are perceived, thereby affecting future impressions. A weak initial correlation may be strengthened by one instance of the co-occurrence through recognition and overestimation. This approach may also play the part of reinforcing stereotypic beliefs acquired through social learning (such as through parents).