Attitude

Research in the social science tradition relies heavily on the concept of attitude. We are taught that it is an internal disposition composed of a cognitive (knowledge about an object), emotive (subjective valuation of the object) and dynamic (willingness to act in relation to the object) component - or to put it dialectically: moment. It is most frequently (that is almost exclusively) measured by relying on introspection. Technically this means administering a questionnaire.

Let us look at an earlier psychological tradition that puts this method in question: behaviourism. This questioning is pertinent since attitudes, meant to explain behaviour, are notoriously useless in doing just that: predicting behaviour. George Herbert Mead had a different understanding of attitudes (if it had not degenerated to the sorry state of psychophysics, behaviourism might have been the missing link freeing psychology from solipsism). He understood them as an inclination towards a specific response towards a specific stimulus in specific circumstances. The third element of his definition - circumstances - explains the failure of explaining behaviour via attitudes. An item in a likert scale is a stimulus that elicits a certain verbal response in the situation of anonymously filling out a questionnaire. Generalising this response to other verbal responses in anonymous situations is valid (that is why opinion polling is so effective in predicting outcomes of elections), generalising it to incommensurable situations that involve different responses is the proton pseudos of attitude research.