Pattern Recognition and Cognitive Categories: It is on the basis of pattern recognition that we form cognitive categories (in linguistics, these are generally what we call "nouns," which are persons, places or things).  Categories are essential for understanding and navigating in the world.

Stereotype: A stereotype is something used to characterize, or describe a category.  Sometimes it is seen as a template, like a prototype, or archetype.  They can be either positive, or negative, or somewhere in between.

Labeling: This involves describing someone or something in a word or short phrase.  All nouns are labels of a sort, a name that is given to a person, place, or thing.  Labeling is an outgrowth of this sort of pattern recognition.  An example is "dog" for the upper left image above.  Another could be describing someone who has broken a law as a "criminal."  Another would be calling someone who has some sort of physical incapacity a "cripple."  The problem that can arise involves the fact that, in addition to denotations, such labels can have connotations, which can be either positive, or negative, or in some cases some of both.

Using labels (and stereotypes) can help simplify the patterns folks recognize.  Neither are automatically detrimental.

The problem arises when it results in over-simplification (i.e., reification) of things that are more complex and/or nuanced.  Unfortunately, there are those who rely on stereotyping and labeling as a way to avoid complexity and/or avoid recognizing their inability to deal with complexity.

Prejudice: The thing about a prejudice is that it involves evaluation (judging something) in advance.  It can and often is based on a stereotype, or label (often contained in the connotation).  It can be either positive, or negative, and in some cases a little of both.

Preference: When looking at bias, it is important to factor in the evaluative portion.  Thus, when looking at a pattern being recognizes, is there a positive, or negative value attached (like a connation attached to the definition of a word). 

When it comes to the flavor or ice cream, some prefer vanilla, while some prefer chocolate.  This can be viewed as a bias in favor of one versus the other.  Similarly, people may prefer not to see certain things, or engage in certain activities.

Cognitive categories Positive, or Negative? These cognitive categories (e.g., stereotypes and prejudices) are not inherently positive, or negative.  There can be positive ones and negative ones (e.g., "war veterans" can be seen positively, while "welfare recipient" can often be seen negatively).

Biases and Pattern Recognition: From this perspective, a bias is simply a pattern (either spatial, or temporal) that we use at times when they (one, or the other, or both) aren't applicable.  They are erroneous patterns that we see that prevent use from actually perceiving reality. 

As with cognitive categories discussed above, they can be either positive, or negative, but what is most important about them is their potential to obscure our perceptions of reality.  This can be especially prevalent in the case of a bias based on a negative stereotype.

So, for example, a person who tends to have a spatial bias for the perception of dirt might see the picture of the dog at the top of this page as actually just a bunch of dirt blotches.  Biases and prejudices can prevent us from seeing reality and they can lead us into a variety of logical fallacies.

Inkblot Test: By the way, it is this ability to recognize patterns, even when the visual material is quite vague, that form the basis of the inkblot test.  So, what do you see?  The "biases" revealed when people state what they see in the inkblot gives some information that can be used to assess personality.

Biases Positive, or Negative?  As with cognitive categories discussed above, they can be either positive, or negative, but what is most important about them is their potential to obscure our perceptions of reality.  This can be especially prevalent in the case of a bias based on a negative stereotype.