An attitude is normally defined as a perceptual orientation and response readiness in relation to a particular object or class of objects.
Attitudes are most commonly measured through analysis of patterns of responses to questionnaires and other self-report techniques. These falls into two major groups: (a) Scales which present directional statements of opinion, to which responds with some amount of agreement or disagreement.
image source: imagesbuddy.com
ADVERTISEMENTS:
Among the best known scales of this kind are the F-scale (Adorno, Frenkel – Brunswik, Levinson and Sanford, 1950), the Dogmatism Scale (Rokeach, 1960), and the Social Attitude Inventory (Eysenck, 1954). (b) Scales which present non-directional concepts and require the respondent to evaluate them.
The Semantic Differential (Osgood, Suci and Tannen- baum, 1957) and the Conservatism Scale (Wilson and Patterson, 1968) are scales of this kind. When applicable, the latter item format would seem to be preferable on a number of counts.
It is economical, less susceptible to acquiescence response bias, eliminates certain sources of ambiguity (e.g. differences in the perceived point of emphasis in the opinion statement), and avoids algebraic addition problems for the subject (e.g. having to disagree with negatively worded statements)
Of the various techniques for selecting scale items, most of the traditional methods (discussed under attitude scales before) were concerned with the construction of unidimensional scales. With the advent of the computer, the mathematical models upon which they are based have been largely superseded by the technique of factor analysis (a general statistical method, involving correlation coefficients, that isolate a few common features from a large number of tests, ratings, or other measurements), which permits the development of multi-dimensional attitude scales.
ADVERTISEMENTS:
In addition to the self-report technique, various disguised measures of attitude have also been investigated, on the argument that they are less contaminated by various inhibitory factors (such as “social desirability”) which might lead to faking though this problem is not very insistent in the area of attitude measurement because of the evaluative nature of attitudes themselves.
Nevertheless, attitudes can be measured without the subject’s awareness in terms of their effects upon memory and perceptual processes (e.g. recall thresholds, binocular resolution), or performance on certain tasks involving materials relevant to the attitude object (e.g. classifying opinion statements as favourable or unfavourable towards an issue instead of indicating one is own agreement).
Several physiological measures of attitudes have also been used, especially autonomic measures such as the (GSR) Galvanic skin response (Cook and Selltiz, 1964). The major disadvantage of such measures is that they indicate only the intensity of an attitude response, not its direction.