11 Categorization and Cognition
Sahdra and Ross (2007, Study 1) asked Hindu and Sikh participants to recall instances of ingroup violence on the other group. Participants who identified strongly with their ingroup recalled fewer instances of ingroup violence on outgroup members than low ingroup identifiers. Both high and low identifiers recalled a similar number of instances when ingroup members were the victim. After reading about two instances when ingroup members harmed outgroupers and two instances when outgroupers harmed ingroupers, participants who identified strongly with the ingroup were more likely to agree with the statement that outgroup members should forget about the act of violence and move on. When ingroup members were on victims of outgroup violence, high identifiers were less likely to agree that the ingroup members should forget about the act and move on. In Study 2, Sahdra and Ross (2007), using a sample of Canadian undergraduate students, manipulated how strongly their participants identified with their ingroup (Canadians; high or low identifiers). Participants were then asked to recall instances of positive and negative acts that Canadians had committed against non-Canadians. The results show that participants in the high identity condition recalled significantly more positive acts than negative acts committed by Canadians towards non-Canadians. By contrast, participants in the low identity condition recalled an equal number of positive and negative acts committed by Canadians against non-Canadians.
Kohli, Sing, Lolliot, Meiser, and Hewstone (under review) had Indian participants read two profiles describing an ingroup member (of the same caste) and an outgroup member (from a different caste) respectively. Participants were then asked to recall the positive and negative traits that were used to describe the ingroup and outgroup members in the profiles. The results showed that participants correctly remembered significantly more positive traits for the ingroup member than for the outgroup member. Similarly, participants also recalled significantly more negative traits for the outgroup member compared to the ingroup member. Interestingly, Kohli and colleagues analysed the number of false positive and negative traits recalled and found that participants misremembered significantly more positive traits for ingroup members and misremembered significantly more negative traits for outgroup members. This results demonstrates the creation of false memories favouring the ingroup.
Van Bavel and Cunningham (2009) investigated the extent to which categorization affects processing fluency of positive and negative words. Using a sample of white students, the researchers showed participants the racial composition of two teams –the Tigers and Leopards. Both teams were made up of black and white individuals. Participants were then asked to memorize the faces. Next, participants were briefly flashed the faces from the members of both teams before being asked to categorize a word as either positive or negative (e.g., flash a white face –> see the word “hatred” –> categorize it as either positive or negative). In two conditions, Van Bavel and Cunningham had participants memorize the photographs and do the word-categorization task either without being assigned to a team themselves or after being assigned as a team member of either the Tigers or the Leopards. The results showed that when participants did the categorization task without being assigned to being a Tiger or a Leopard, they exhibited a race bias in that they were more accurate at correctly categorizing positive words compared to negative words when a white face was flashed compared to when a black face was flashed. When they were part of a team, however, participants categorized positive words (as positive) significantly more accurately for ingroup members compared to outgroup members regardless of their race. They were also more accurate at categorizing negative words for outgroup members (compared to ingroup members) regardless of their race.
These studies go to show that identification with an ingroup can bias memory and information processing in ways that favour the ingroup and can derogate the outgroup.
