Maya Angelou put it best in her fantastic 1973 conversation with Bill Moyers, considering the laziness of stereotypes:
All you have to do is put a label on somebody. And then you don’t have to deal with the physical fact. You don’t have to wonder if they are waiting for the Easter bunny or love Christmas, or, you know, love their parents and hate small kids and are fearful of dogs. If you say, oh, that’s a junkie, that’s a nigger, that’s a kike, that’s a Jew, that’s a honkie, that’s a — you just — that’s the end of it.
Participants rated their sexual orientation on a 10-point scale, ranging from gay to straight. Then they took a computer-administered test designed to measure their implicit sexual orientation. In the test, the participants were shown images and words indicative of hetero- and homosexuality (pictures of same-sex and straight couples, words like “homosexual” and “gay”) and were asked to sort them into the appropriate category, gay or straight, as quickly as possible. The computer measured their reaction times.
The twist was that before each word and image appeared, the word “me” or “other” was flashed on the screen for 35 milliseconds — long enough for participants to subliminally process the word but short enough that they could not consciously see it. The theory here, known as semantic association, is that when “me” precedes words or images that reflect your sexual orientation (for example, heterosexual images for a straight person), you will sort these images into the correct category faster than when “me” precedes words or images that are incongruent with your sexual orientation (for example, homosexual images for a straight person). This technique, adapted from similar tests used to assess attitudes like subconscious racial bias, reliably distinguishes between self-identified straight individuals and those who self-identify as lesbian, gay or bisexual.
Using this methodology we identified a subgroup of participants who, despite self-identifying as highly straight, indicated some level of same-sex attraction (that is, they associated “me” with gay-related words and pictures faster than they associated “me” with straight-related words and pictures). Over 20 percent of self-described highly straight individuals showed this discrepancy.
Notably, these “discrepant” individuals were also significantly more likely than other participants to favor anti-gay policies; to be willing to assign significantly harsher punishments to perpetrators of petty crimes if they were presumed to be homosexual; and to express greater implicit hostility toward gay subjects (also measured with the help of subliminal priming). Thus our research suggests that some who oppose homosexuality do tacitly harbor same-sex attraction.