Researchers within the field of animal behaviour has in general agreed in classifying a behaviour as "Stereotypic" if it can be described using the following three criteria's: The movements are relatively un-varied; is repeated regularly and is without an obvious function.

The definition is "loose" and harbours various problems.

First of all there is the practical problem by quantifying the first two criteria's. Which degree of variation is allowed and how regularly must the behaviour in question be repeated? The behaviour is supposed to show a high degree of predictability, but the individual behavioural elements will always be seen to vary upon closer analysis.

Second problem with this definition is that the first two criteria's can just as well be used to describe everyday phenomenon's e.g. digestive processes, drinking behaviour, cleaning the fur and various displays (e.g. courtship behaviour).

The last criteria can encompass what we call "habits" and such behaviour is sometimes observed during Operant Learning Procedures where a hungry rat, trained to deliver a certain response (e.g. pulling down a lever to receive a food pellet) persist in this response long after food has been provided. The last criterion is therefore probably more a reflection of our lacking knowledge, more than a proper "truth" about the nature of stereotypies.

In other words, this broadly recognized definition does contain subjective elements which in real life mean that what one researcher classifies as a stereotypic act, the other classify as something completely different.