When you think about members of a group, you often evaluate how good a member of that group they are. And that evaluation matters.
For example, when people go to the shelter to adopt a puppy, they are trying to decide whether there is a dog there they want to take home with them. Some people go hoping to find a particular kind of dog, but many end up evaluating dogs based on whether they seem like good examples of the dog category. What kind of information goes into this evaluation?
Lots of research on categorization going back 50 years suggests that there are two different ways that people might determine how good something is as a member of a category.
Sometimes, people form a prototype of a category. A prototype is an average member of the category. For example, for a category like birds, you may have a prototype. The prototypical bird is small, has feathers, sings, and flies south in the winter. A good example of a bird is one that is similar to this prototype like a robin or a sparrow. Birds that are very different from this average member like ostriches, penguins, and emus, are bad examples of birds.
Other times, people form ideals. An ideal is a particularly good category member. For example, for a category like diet foods, you might evaluate examples based on how good they are compared to the ideal diet food, which tastes great and has very few calories. Interestingly, you may never have seen this ideal diet food, but you still use it to evaluate category members.
A paper by Tyler Davis and Brad Love from the University of Texas published in the February, 2010 issue of Psychological Science explores the circumstances that lead people to form prototypes or ideals.
They find that a crucial factor in the formation of categories is whether you are contrasting the category to others while you are creating it.
Sometimes, you learn categories by just focusing on the category itself. For example, when you see birds flying in the sky or sitting in a tree in your yard, you are looking at the bird and watching its behavior. You are probably not that interested in trying to contrast birds with squirrels or dogs. Because you are learning about birds without contrasting them with other categories, you will tend to form a prototype for that category.
When you learn a category by contrasting it with some other category, then you tend to create an ideal to make it easiest to form the contrast. For example, diet foods are contrasted with normal foods, so you tend to identify diet foods based on a particular ideal, which has very few calories.
These authors go further to say that you form an ideal only for those aspects of the category that you are trying to contrast with some other category. For all other aspects of that category, you have a prototype. Because you are contrasting diet foods with non-diet foods, you think of them as having very few calories. However, diet foods are not really distinguished from other foods based on the style of food (say Italian food or Mexican food), so for dimensions of food like style or spiciness you still have a prototype.
So, what happens to people who go to the shelter to adopt a dog? You tend to learn about dogs the same way you learn about birds. That is, you probably don't learn about dogs by contrasting them with other animals. So, you probably have a prototype representation of dogs. As a result, the dogs you think of as particularly good dogs are ones that are close to the average dog (like a lab or a golden retriever).