Ingroup and Outgroup Favoritism in Implicit Attitudes

Individuals hold two kinds of attitudes: explicit versus implicit. Explicit attitudes are relatively more controllable, intentional, and consciously endorsed, and can be measured by simply asking whether someone prefers White to Black people. Implicit attitudes are more unconscious and automatic, and thus require a subtler form of measurement.

Implicit and explicit attitudes can often differ in strength or direction. For example, while 62% of people in a recent sample reported no explicit preference for White versus Black people, 59% showed evidence of favoring White over Black people in implicit attitudes. One ongoing issue in attitude research concerns what types of information are more likely to shape implicit versus explicit attitudes a topic that may then shed light on why the two forms of attitudes often diverge. Recently, my colleagues and I explored this issue by using a relatively novel measure towards understanding the implicit attitudes of majority and minority group members.

Often, psychologists measure implicit attitudes using the Implicit Association Test (IAT). In a typical IAT, participants categorize positive and negative words as well as stimuli related to specific categories (e.g., images of Black and White people) as quickly as possible using a computer keyboard. Participants must categorize words and images using just two keys. For example, participants might use one key for either White people and positive words and another key either Black people and negative words. Later on, the key assignments are switched; participants categorize Black people and positive words with one key, White people and negative words with the other key. Categorizing items faster when Black people and negative words (and White people and positive words) are paired together compared to when Black people and positive words (and White people and negative words) are paired together is believed to indicate more positive implicit attitudes towards White versus Black people.

Since implicit attitudes are more automatic, they may be susceptible to subtle forms of information. Psychologists often use the IAT to examine the impact of cultural messages or values on individuals’ attitudes. For example, when racial majority group members (e.g., White Americans) vs. racial minority group members (e.g., Black Americans) take an IAT, majority group members typically show clear implicit preferences for their own group, whereas members of minority groups show either much weaker preferences for their own group or no preferences at all. Similar findings have occurred for straight versus gay people as well as Christian versus Jewish people.

These data suggest that implicit attitudes are shaped by two factors: 1) one’s own ingroup identity and 2) cultural messages about group status or value. For members of majority groups (e.g., White people), these influences point in the same direction. For members of minority groups (e.g., Black, Asian or Hispanic people), these influences point in opposing directions. Minority group members may derive positive associations from their own group based on their group identity, yet at the same time live in an environment where they are exposed to messages that their group is less liked or has less status. The end result is a “tug of war” in implicit attitudes among members of minority groups.

In our own work, we investigate whether this same pattern would emerge when using a more nuanced measure of implicit associations that separated positive from negative associations. Recall that in a typical race IAT, participants are told to keep in mind whether the word is Good or Bad as well as whether the face is Black or White. In our modified IAT, participants were told sort the same stimuli, but instructions emphasized to press one button for words from one attribute (e.g., Good words) and one category (e.g., White faces), while pressing the other button for “everything else” (e.g., Bad words and Black faces). The end result is that certain information is made more salient as participants complete the task. Prior work using this measure found that these instructions matter; for instance, when participants are told to focus on only the Good words, responses are indeed faster to those words and slower to Bad words.

We then compared the implicit attitudes of people who completed measures focusing only on the Good words, which captured positive associations, versus those who completed measures focusing only on the Bad words, which captured negative associations. We did so for implicit attitudes about race, religion, and sexual orientation.

Our studies found that for members of majority groups, such as White, straight or Christian people, the results were very similar with those from past work. Measures of both positive and negative associations each revealed clear preferences for one’s group. However, for minority group members, results noticeably diverged from past research.

In general, minority group members showed evidence of favoring their own group when assessing positive associations, often even to the same degree as members of majority groups. However, minority group members actually showed favoritism towards the majority group when assessing negative associations. For example, Asian participants implicitly preferred Asian versus White people on measures of positive associations, but showed implicit preferences in favor of White people on measures of negative associations.

One potential explanation for our findings is that positive implicit associations may be more sensitive to information about ingroup identity, whereas negative implicit associations may be more sensitive to cultural messages about group status or value.

We tested this hypothesis by comparing the implicit attitudes of participants who had the same group identity (Judaism) but lived in cultures where they were the minority versus majority group (the United States versus Israel). In our study, American and Israeli Jews showed similar preferences for their own group when measuring positive associations, but only American Jews showed preferences towards Christians (their majority group) when measuring negative associations. This provides more evidence that being a cultural minority has a specific impact on forming negative associations.

Our studies illustrate how implicit attitudes may be more complex than previously believed, and provide an intriguing initial investigation into thinking about the implicit attitudes of minority group members. In the future, we hope to follow-up these studies with longitudinal studies that track minority and majority group members’ positive and negative implicit associations over the course of several years to see if there are particular moments, such as when first enrolling in school, that lead minority group members to show a preference for the majority group when measuring negative associations. Such work would advance our understanding of when and how cultural messages are most likely to impact implicit attitudes.

Author Bio: Jordan Axt is a post-doctoral researcher at Duke University’s Center for Advanced Hindsight as well as Director of Data and Methodology for Project Implicit. His research focuses on explore how people form and express intergroup bias in attitudes and behavior.