Stereotypes sit at the heart of psychological research on social perception. They are typically described as beliefs about the traits, behaviors, or other attributes that characterize a social group. A persistent puzzle, however, is why these “characteristic” attributes so often skew negative. A recent paper by Christian Unkelbach, Anne Irena Weitzel, and Hans Alves proposes an explanation that does not rely on prejudice-driven motives or the idea that people are simply wired to focus on bad news.
The authors advance what they call a cognitive-ecological theory of how people form impressions of groups. It rests on two core claims—one about how minds summarize social information, and another about the informational environment those minds operate in.
On the cognitive side, the theory argues that people tend to describe groups using features that help set one group apart from others. In other words, what becomes “stereotypical” is not necessarily what is most common within a group; it is what seems most diagnostic for telling groups apart.
On the ecological side, the authors draw on two patterns in social information. First, positive information tends to be more prevalent overall: kind acts, ordinary cooperation, and neutral-to-pleasant interactions are frequent in daily life. Second, negative information tends to show greater diversity: harmful, unethical, or socially undesirable behaviors come in many forms. Put together, this creates a statistical asymmetry. If positive attributes are common across many groups, they do less to distinguish one group from another. Negative attributes, by contrast, may be rarer but spread across a wider variety of specific behaviors—making them more likely to uniquely “tag” a particular group in people’s minds.
The theory therefore predicts that the attributes selected as “characteristic”—the raw material of stereotypes—will disproportionately be negative, even in a world where positive behaviors are more frequent. Importantly, it does so without assuming people must be motivated to derogate outgroups, without invoking essential, deep differences between groups, and without requiring a blanket “negativity bias” in attention or memory.
The paper supports this account with evidence from two new studies and two pre-existing datasets, showing the expected tendency toward negative stereotype content. The authors then formalize their argument and use simulations to highlight a key implication: an attribute can be highly informative about group membership even if it applies to only a minority of that group’s members. This helps resolve an apparent contradiction in stereotype research. Stereotypes can feel revealing and may even carry some predictive signal (they can be diagnostic), while still being poor descriptions of what most group members are actually like (they can be descriptively inaccurate).
Beyond explaining negativity, the framework also specifies conditions under which positive stereotypes should be more likely—namely when positive attributes become unusually distinctive between groups. The authors also connect their account to broader “dimensional” models of stereotype content and consider what the theory implies for efforts to reduce harmful stereotyping. If negativity emerges from how distinctive information is sampled and emphasized, then interventions may need to change what information is treated as group-diagnostic, not merely tell people to “think positively.”
Why it matters
Negative stereotypes shape hiring, policing, healthcare, and everyday social relationships, often in ways that disadvantage targeted groups. This theory suggests that negativity can arise from ordinary cognitive shortcuts interacting with the structure of social information—not only from hostile intent—so reducing harm may require changing which group differences are made salient and repeatedly reinforced. The next test is whether interventions that alter perceived “diagnostic” traits can measurably shift stereotype content in the real world.