A stereotype is defined as a social perception of an individual in terms of group membership or physical attributes. It is a generalization that is made about a group and then attributed to members of that group. Such a generalization may be either positive or negative. For example, women are talented speakers or women are bad drivers. Stereotyping is a form of social categorization that affects the behaviour of those who hold the stereotype, and those who are labeled by a stereotype. Researchers explain stereotyping as a result of schema processing.

The formation of stereotypes

How is it that stereotypes develop? Although Tajfel argues that this is a natural cognitive process of social categorization, this theory on its own does not fully explain how the stereotype is actually formed. Tajfel argues that we see our out-groups as all having similar traits, what is known as out-group homogeneity.  But even this does not really explain why we can come to believe that members of a group all share a positive or negative trait.  Where does the connection between groups and certain traits come from?

Schneider (2004) argues that there are two ways that stereotypes develop: indirectly, as a product of our culture or society or directly, as a result of our own experience with other people.  Campbell (1967) maintains that there are two key sources of stereotypes: personal experience with individuals and groups, and gatekeepers - the media, parents, and other members of our culture. He goes on to argue that stereotypes thus have a basis in some reality. His grain of truth hypothesis argues that an experience with an individual from a group will then be generalized to the group.

Hamilton and Gifford (1976) argue, instead, that stereotypes are the result of an illusory correlation - that is, people see a relationship between two variables even when there is none. An example of this is when people form false associations between membership of a social group and specific behaviours such as women’s inferior ability in mathematics. The illusory correlation phenomenon causes people to overestimate a link between the two variables, here “women” and “ability in mathematics”. Illusory correlations can come in many forms and culturally-based prejudice about social groups can to some extent be classified as illusory correlations. Illusory correlation is an example of what researchers call “cognitive bias”, that is, a person’s tendency to make errors in judgment based on cognitive factors. Attribution errors are also examples of cognitive bias.

Once illusory correlations are made, people tend to seek out or remember information that supports this relationship. This is an example of confirmation bias. Generally, this means that people tend to overlook information that contradicts what they already believe. In a social context, they pay attention to behaviours that confirm what they believe about a group and ignore those behaviours contrary to their beliefs. Confirmation bias makes stereotypical thinking resistant to change. Snyder and Swann (1978) conducted a study in which they told female college students that they would meet a person who was either introverted (reserved, cool) or extroverted (outgoing, warm). The participants were then asked to prepare a set of questions for the person they were going to meet. In general, participants came up with questions that confirmed their perceptions of introverts and extroverts. Those who thought they were going to meet an introvert asked, “What do you dislike about parties?” or “Are there times you wish you could be more outgoing?” and extroverts were asked, “What do you do to liven up a party?” The researchers concluded that the questions asked confirmed participants’ stereotypes of each personality type so that it became a self-fulfilling prophecy - for example, because they believed he was an introvert they asked him questions that made him appear to be one.

Stereotypes can also be formed as a means of taking on the in-group’s social representation of the out-group. In other words, individuals may conform to the group norms with regard to the “other”. Rogers and Frantz (1962) found that white immigrants to Rhodesia (today’s Zimbabwe) developed more stereotypes and prejudice against the local people the longer they stayed there. They argue that this is because they adopted the social norms that were dominant in the group they were joining in order to fit in. The study, however, was cross-sectional.  In other words, the behaviour of the participants was not measured over time, but instead a "snapshot in time" was taken and the data were compared. We cannot know if there was a significant change in the stereotyping and prejudice of the participants over time as their original attitudes toward the local people were never measured.

Effects of stereotyping

In the cognitive approach, we learned that schema have an effect on behaviour.  As stereotyping is a form of schema processing, it should be no surprise to you that stereotyping can affect behaviour. Two ways in which stereotyping may affect behaviour are stereotype threat and memory distortion.

Stereotype threat occurs when one is in a situation where there is a threat of being judged or treated stereotypically or a fear of doing something that would inadvertently confirm that stereotype. Steele and Aronson (1995) carried out an experiment to see the effect of stereotype threat on performance. They gave a 30-minute verbal test, made up of very difficult multiple-choice questions. When one group was told it was a genuine test of their verbal abilities, African American participants scored significantly lower than European American participants. In another group that was presented with the same test but told that it was not diagnostic of their ability, African American students scored higher than the first group, and their performance was not significantly different from the European American students. However, it was not possible to argue that awareness of racial stereotyping was the actual cause of the difference, so they did one more version of the experiment where the participants had to fill in a personal information questionnaire.  Half of the questionnaires asked the participants to identify their race. African Americans who had to identify their race did poorly; those that did not did just as well as their European American peers.

Steele & Aronson argue that you don’t need to believe in a stereotype for it to affect your behaviour. Stereotype threat turns on spotlight anxiety, which causes emotional distress and pressure that may undermine performance. Students under the stereotype threat often underperform and this can naturally limit their educational prospects. Spencer et al (1999) tested the effect of the stereotype threat on intellectual performance. The researchers gave a difficult mathematics test to students who were strong in mathematics. They predicted that women under the stereotype threat would underperform compared to the men taking the test. The stereotype threat that women experience in mathematics-performance settings originates from a negative stereotype about women’s mathematics ability, which is quite common in society. For women who are good at mathematics and see mathematics as an important part of their self-definition, such a stereotype threat may result in an interfering pressure in test situations. Spencer et al. found that this was true: women in the experiment significantly underperformed compared with equally qualified men on the difficult mathematics tests.

Another effect of stereotyping is memory distortion. In a study carried out by Martin & Halverson (1983), the researchers wanted to see if gender stereotyping would influence recall in 5 and 6-year-old children. Each child was shown 16 pictures, half of which depicted a child performing gender-consistent activities (for example, a boy playing with a truck) and half showing children displaying gender-inconsistent behaviours (for example, a girl chopping wood). One week later, they tested the recall of the children to see how many of the photos that they could recall accurately.

The results showed that children easily recalled the sex of the actor for scenes in which actors had performed gender-consistent activities. But when the actor’s behaviour was gender inconsistent, the children often distorted the scene by saying that the actor’s sex was consistent with the activity they recalled - that is, they would remember that it was the boy playing with a truck, when in fact they had been shown a photo of a girl playing with a truck.

In a classic experiment done in 1947, Allport & Postman wanted to see the effect of stereotyping on recall.  To carry out the experiment, the researchers showed a participant a drawing of a white man holding a razor and threatening a black man on the subway. The participant was then asked to describe what he had seen to another participant – a process called serial reproduction, one of the same strategies used by Bartlett in his classic War of the Ghosts study.  The process was repeated for up to seven participants.  The researchers found that in over half of the experiments, at some stage the story was retold so that the black man was threatening the white man. It appears that stereotyping had affected the retelling of the story. The study is very old, so is it still relevant today? Unfortunately, the answer is yes.

In a study by Payne (2001), participants were first given a test of racial prejudice to eliminate strong racial prejudice as a confounding variable.  32 non-black university students were asked to sit at a computer screen.  They would be shown a face - either a white or a black face -  followed by an image.  They had only a second to choose whether the image was either a tool or a gun. The study showed that priming the participants with black faces led to more errors in identifying the image as a weapon rather than a tool. Thus, stereotyping affected perception, which then had implications for memory.  Several studies have shown that the race of an alleged criminal influences memory of a weapon. The effect of stereotypes can be very serious indeed.

Checking for understanding