26.03.2013 Views

Chapter 9: Prejudice: Disliking Others (2947.0K) - Bad Request

Chapter 9: Prejudice: Disliking Others (2947.0K) - Bad Request

Chapter 9: Prejudice: Disliking Others (2947.0K) - Bad Request

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

correlations in people’s minds (Risen & others, 2007). This enables the mass media<br />

to feed illusory correlations. When a self-described homosexual person murders or<br />

sexually abuses someone, homosexuality is often mentioned. When a heterosexual<br />

does the same, the person’s sexual orientation is seldom mentioned. Likewise,<br />

when ex-mental patients Mark Chapman and John Hinckley, Jr., shot John Lennon<br />

and President Reagan, respectively, the assailants’ mental histories commanded<br />

attention. Assassins and mental hospitalization are both relatively infrequent, making<br />

the combination especially newsworthy. Such reporting adds to the illusion of<br />

a large correlation between (1) violent tendencies and (2) homosexuality or mental<br />

hospitalization.<br />

Unlike the students who judged Groups A and B, we often have preexisting<br />

biases. David Hamilton’s further research with Terrence Rose (1980) revealed that<br />

our preexisting stereotypes can lead us to “see” correlations that aren’t there. The<br />

researchers had University of California at Santa Barbara students read sentences in<br />

which various adjectives described the members of different occupational groups<br />

(“Juan, an accountant, is timid and thoughtful”). In actuality, each occupation was<br />

described equally often by each adjective; accountants, doctors, and salespeople<br />

were equally often timid, wealthy, and talkative. The students, however, thought<br />

they had more often read descriptions of timid accountants, wealthy doctors, and<br />

talkative salespeople. Their stereotyping led them to perceive correlations that<br />

weren’t there, thus helping to perpetuate the stereotypes.<br />

Likewise, guess what happened when Vaughn Becker and his colleagues (2010)<br />

invited university students to view a White and a Black face—one angry, one not—<br />

for one-tenth of a second. Then, as a brief distraction, they added two numbers that<br />

accompanied the faces (as in Figure 9.8 ). The participants’ subsequent recollections<br />

of what they had viewed revealed racial bias. “White anger flowed to neutral Black<br />

faces (34 percent likelihood) more readily than Black anger flowed to neutral White<br />

faces (19 percent likelihood).”<br />

Attribution: Is It a Just World?<br />

In explaining others’ actions, we frequently commit the fundamental attribution<br />

error that was discussed in <strong>Chapter</strong> 3: We attribute others’ behavior so much to<br />

their inner dispositions that we discount important situational forces. The error<br />

occurs partly because our attention focuses on the person, not on the situation. A<br />

person’s race or sex is vivid and gets attention; the situational forces working upon<br />

that person are usually less visible. Slavery was often overlooked as an explanation<br />

for slave behavior; the behavior was instead attributed to the slaves’ own nature.<br />

Until recently, the same was true of how we explained the perceived differences<br />

between women and men. Because gender-role constraints were hard to see, we<br />

–2 5<br />

<strong>Prejudice</strong> <strong>Chapter</strong> 9 339<br />

FIGURE :: 9.8<br />

Ingroup biases influence<br />

perceptions. When briefly<br />

shown two faces, one neutral,<br />

one angry, people more often<br />

misrecalled the Black rather than<br />

the White face as angry (Becker<br />

& others, 2010).

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!