Kahneman & Tversky (1972) were among the firsts to propose that, when people make assumptions about the origin, categorization, or classification of a person, as well as of things, they usually follow a series of "rules of thumb" (heuristics) based on the superficial traits represented in the person or thing.
"Representativeness is an assessment of the degree of correspondence between a sample and a population, an instance and a category, an act and an actor or, more generally, between an outcome and a model." (Tversky and Kahneman (1984), In: Gilovich, Griffin and Kahneman (2002) )
To illustrate, think about how many people will judge a person's socioeconomic status based on the combination of the individual's possessions such as the car that they drive, their address, and the way that they dress. Based on those "representative" traits a consensus may be reached that someone who has a nice car, nice clothes, and a sophisticated personality is, in fact, someone from a high socioeconomic status. From that fact on, a plethora of other assumptions may be reached, whether they are correct or incorrect assumptions. Although this is typical human behavior, we know that looks can certainly be deceiving and this is where the cognitive pros and cons come into play.
Cognitively speaking, we must never rely on representative heuristic because, as was just discussed, it is easy to adopt or disguise ourselves as members of a specific social group in order just to fit in. However, looks are only skin-deep. Think about Ted Bundy and Jeffrey Dahmer; two middle-class, college-educated white males who were attractive, sophisticated, dressed well and seemed quite charismatic. Judging by the representative heuristic, they would have been thought to be college professors, or upper-class professionals. Instead, they were serial sociopaths, rapists, and murderers who socially-disguised themselves as clean-cut middle class guys precisely to lure their victims. Dahmer is even thought to have practiced cannibalism! Hence, the social tendency to judge by what we see in the surface does not allow us to cognitively use all of the data that we can gather from another individual to be able to know correctly judge them for what they really are. If we learn to rely less on representative heuristic, we may even learn to make better, more realistic, and more sincere human connections.
However, the use of representative heuristic may serve a good deed when addressing children and young adults on how to get to know people who may want to hurt them. This day and age, anybody can be "a stranger", from the clown at the circus, to the sociopath next door. Teaching the possible disguises that a sociopath, or any other type of criminal, may use to infiltrate a defined social group is imperative to activate our deductive and inductive cognitive skills so that we do not become victims of people merely "posing" as being part of a specific group.
Moreover, it is alright to be defined as a social group. It brings the group the identity and uniqueness that makes them feel that they belong somewhere. It is when we rely merely on looks, and not on other data, when representative heuristic may be a dangerous thing to rely on.