With a view to assessing the degree of attitudes possessed by persons and to be able to study a large number of people, the scaling technique was introduced into attitude measurement. Various scales of attitude measurement have been developed. Here we shall only broadly discuss the characteristics of some prevalent attitude scales so as to get acquainted with the general steps involved in their construction and use. It is likely that, in spite of numerous scales being available(*), you do not find one handy or suitable when you take up a particular study. Knowledge of how to develop an attitude scale will obviate such a crippling situation and help you have an instrument tailor-made for a given study. For a detailed discussion of how to construct an attitude scale, you may refer to Allen L. Edwards' (1957) Techniques of Attitude Scale Construction.

Thurstone's Scale

Louis L. Thurstone and E.J. Chave (1929) in their classic study of attitudes toward the Church developed an interval scale by using the method of equal-appearing intervals. To construct the Thurstone scale, a large number of statements are collected which express various possible opinions about the issue or object of study. These statements, after an editing for relevance and clarity, are given to judges, who are to independently sort them into eleven sets along a continuum that ranges from most unfavourable, through neutral, to most favourable. The eleven sets of statements are to occupy positions in the continuum in such a way that the positions are at equal intervals; that is, the difference between any two adjacent positions is the same as the one between any other two adjacent positions. For the final form of the scale, only those items are retained that have high interjudge agreement and fall at equal intervals.

The judges are to assign the statements to appropriate positions on the scale only on the logical basis of how favourable or unfavourable an opinion every statement expresses by itself and not how far the judges personally agree or disagree with the statements. The average judged position of a statement on the eleven-point continuum is the scale value for that statement. Thus, when a Thurstone scale is ready, every statement in it (there are usually about twenty statements) has a numerical value already determined. When administered, the respondent just checks the items s/he agrees with and her/his attitude score is the mean value of the items s/he checked.

Likert's Scale

For the Likert scale, various opinion statements are collected, edited and then given to a group of subjects to rate the statements on a five-point continuum: 1=strongly agree; 2=agree; 3=undecided; 4=disagree; and 5=strongly disagree. The subjects express the degree (one to five) of their personal agreement or disagreement with each of the statements. Only those items which in the analysis best differentiate the high scorers and the low scorers of the sample subjects are retained and the scale is ready for use. To measure the attitude of a given group of respondents, this scale is given to them and every respondent indicates whether s/he strongly agrees, agrees, is undecided, disagrees, or strongly disagrees with each statement. The respondent's attitude score is the sum of her/his ratings of all the statements. For this reason, the Likert scale is also known as the scale of Summated Ratings.

In the Thurstone scale, the respondent checks only those items with which s/he agrees, whereas in the Likert scale s/he indicates her/his degree of agreement or disagreement for all the items in the scale. Further, the development of a Likert scale does not require a panel of judges. It may also be noted that Likert did not assume equal intervals between the scale points. His scale is ordinal and, therefore, can only order respondents' attitudes on a continuum; it does not indicate the magnitude of difference between respondents.

By and large, a great majority of researchers prefer the Likert technique to Thurstone's. In many current research studies we come across seven-point scales being used, which bear the appearance of the Likert scale. It must be noted that the typical Likert technique requires an item analysis to establish that all the items in the scale measure the same attitude -- no matter whether the scale has five or more points.

Bogardus' Social-distance Scale

Back in 1933, E.S. Bogardus developed an attitude scale, called the social-distance scale, which became a classic instrument to measure attitudes toward ethnic groups. Different nationalities or racial groups are listed and various possible relationships with them are also given:

The respondent is asked to indicate the relationships to which s/he is willing to admit members of each group. Her/his attitude is measured by the closeness of the relationship s/he chooses.

The social-distance scale implied that a respondent who admitted a stimulus person to a particular relationship would also admit him/her to all other relationships which are related and are less close than the chosen relationship. Thus, if a person says that s/he would take a member of a given group as spouse, s/he would also have him/her as a friend or neighbour. But, such an assumption about the cumulative nature of the items in the scale would hold good only if all the items expressed one and the same aspect or dimension of the attitude in question. A person may be willing to marry a member of a particular group purely for the advantages such a marriage might entitle her/him to, but s/he may not be ready to have a member from that group as friend or neighbour.

Guttman's Scalogram

With a view to ensuring a cumulative measure of attitudes, Guttman developed a more refined Scalogram to measure unidimensional attitudes. The scalogram consists of a set of statements related to the attitude in question and arranged in increasing order of difficulty of acceptance. It is based on the same logic as the one in the Standford-Binet test of intelligence: if a person solves a difficult item, s/he should be able to solve all simpler items. The same way, if a person agrees with a statement that expresses a higher degree of a given attitude, s/he must be agreeable to all other statements which express lower degrees of the same attitude. Obviously, this logic holds good only if all the items in the scale are from one and the same universe, that is, the scale measures the same aspect of a particular attitude.

To construct the scalogram, opinion statements are collected and arranged in such a way that most people would accept the first statement and, going down the list, fewer and fewer persons would accept the subsequent statements. The list of statements, thus arranged, is given to sample subjects in order to test the increasing degree of acceptance. Based on the "accept" responses of respondents, the items are accordingly modified, arranged, and tested again on sample subjects. This process continues till a scalable (i.e., empirically tested for increasing degree of acceptance) set of items is developed. The final set of statements with their particular order is the scalogram. When this instrument is used for measuring a person's attitude, the person checks all the items s/he accepts. The person's score is the total number of successive or nearly successive items s/he has checked.

In practice, however, it has been observed that rarely respondents check items without skipping one or more items. This phenomenon confirms the difficulty involved in preparing a perfectly unidimensional scale. It may also point to the probable fact that people in real life respond not to a single dimension of reality, but to peculiar combinations of them.

The Semantic Differential

The now-classic research by Osgood and his colleagues, based on extensive factor-analytic studies across cultures, has shown that people understand, or give meaning to, words or concepts along three dominant dimensions--the evaluative (good-bad) dimension, the potency (strong-weak) dimension, and the activity (active-passive) dimension. It has also been found that scores on the evaluative dimension correlate highly with other measures of attitude toward a particular social object.

The Semantic Differential, developed by Osgood, Suci and Tannenbaum, can be used to measure attitudes from the meaning (semantic = meaning or psychological significance) which people give to a word or concept that is related to an attitude object. This instrument consists of a series of bipolar adjectives such as fair-unfair, pleasant-unpleasant, good-bad, clean-dirty, valuable-worthless, etc. Each pair constitutes a continuum of seven points, the endpoints being the opposites of the adjective pairs and the midpoint being the neutral position. A sample of the bipolar continuum is given below:

Suppose, by means of the Semantic Differential, you want to measure an individual's attitude towards legalised abortion. The respondent is given a set of bipolar adjectives (such as the ones sampled above) and s/he is asked to indicate as to where for her/him the given attitude object (legalised abortion) falls in each continuum. The numeral corresponding to the position checked by the subject is her/his score for that continuum. One's overall attitude score is the sum (or the mean) of the scores on all the continua.

Limitations of Self-report

The direct questioning and the various scaling methods we have discussed depend on respondents' own account of their reactions to a psychological object. The purpose of these instruments is also often apparent to the respondent. Responses to self-report techniques are, thus, under the conscious control and voluntary distortions of the respondent. If the respondent wants to preserve her/his self-image, or if s/he wants to impress the investigator favourably, s/he would respond in accordance with this purpose and not in accordance with her/his attitude that is being studied. In a study of attitude towards Harijans, for example, the respondent might state favourable attitudes in order to present a positive picture of her/him-self as being educated, mature, open-minded etc. But, in truth, s/he may harbour vehemently negative feelings against the group. When sensitive and controversial issues are studied, the respondent may feel apprehensive about expressing her/his true attitudes toward the issue. S/he may be afraid of the consequences, which s/he perceives will result from honest disclosure. In addition to being affected by such social desirability factors, responses to these instruments may also get distorted by the response set of the individual. A "response set" refers to the consistent tendency to agree or disagree with items, regardless of what the items are about.

These and other shortcomings of self-report techniques have long been recognised and steps to minimize them have also been suggested and followed. To make the purpose of the instrument less obvious to the respondent, a number of test items not relevant to the attitude object may be included. Giving assurances of anonymity and confidentiality can help de-inhibit the respondent. The importance of frank responses for the development of scientific knowledge, if properly pointed out, may win the cooperation of some respondents. Forced-choice formats (in which carefully edited alternative responses, either all positive or all negative, are listed and the respondent is asked to tick the one closest to her/his own view) are of help in breaking a person's response set. Another way to counteract possible response sets is wording the test items in such a way that for half the number of items (randomly spread over the whole test) agreement represents a favourable response and for the other half an unfavourable response.

Think up additional ways of making it easier for the respondent to overcome the influences of social desirability factors and of nullifying a response set.

Unobtrusive Measures

All the methods discussed thus far depend, in one way or another, on the cooperation or at least the physical presence of subjects, in whose attitudes the investigator is interested. Such dependence on or instrusion into a social setting is likely to contaminate the responses. To make up for situations where the measurement technique itself is reasonably suspected to produce what is to be measured, some ingenious approaches have been suggested under the name of nonreactive or unobtrusive methods. These methods include examination of the physical traces of erosion and accretion, perusal of private and public records, etc. For example, to get a measure of the popularity of a particular exhibit in a museum, the relative wear of the flooring below the exhibits has been studied. Such measures may not, however, be congenial to studies of current attitudes. Observation of natural behaviour, which we have discussed earlier, can be used unobtrusively if care is taken that such observation is not noticed by the persons observed. Exploring further nonreactive approaches to serve measurement of attitudes will gainfully widen the restricted horizons of today.

Multiple Measures

Typically, an investigator interested in measuring the attitude of a group towards a particular issue, selects (or develops) and uses just one instrument which s/he judges as suitable for the study. But, if one used more instruments to measure the same attitude of the same group with regard to the same issue, one would certainly obtain as many different measures as the number of instruments used. Which of them is the true measure? One of them true and others false? Maybe, none of them is true!

Well, measurement techniques at the time of their construction are subjected to a check for validity. That is, the instruments are tested to see if they in fact measure what they are intended to measure. On this ground, therefore, we know that results got by means of such instruments are not altogether false. But, we also know that the validity coefficients are always less than 1. This means the measures obtained in response to a particular instrument have been influenced by the main variable (the attitude being studied) as well as extraneous factors not related to the main variable. The kind of extraneous factors and the amount of their contaminating influence vary from instrument to instrument. Therefore, some social scientists argue that use of several independent instruments, in each of which different identifiable extraneous factors are operative, will give us a more reliable measure. Such a multiple-indicator approach to attitude measurement has been tried and advocated as an improvement over the commonplace single-indicator approach. Here a word of caution is in order. It, no doubt, appeals to reason that viewing an issue from various angles leads to a better grasp of or insight into the issue. But, increasing the number of instruments does not, by itself, guarantee better results. If the instruments used happen to be defective, any increase in their number will only inflate the errors.
------------------------------------------------------
(*) Pareek and Rao's (1974) Handbook of Psychological and Social Instruments gives fairly exhaustive information about such instruments available in the country.

References

Campbell, D.T. The indirect assessment of social attitudes. Psychological Bulletin, 1950, 47, 15-38.

Cook, S.W., and Selltiz, C. A multiple-indicator approach to attitude measurement. Psychological Bulletin, 1964, 62 (1), 36-55.

Edwards, A.L. Techniques of attitude scale construction. New York: Appleton-Century-Crofts, 1957.

Hess, E.H., and Polt, J.M. Pupil size as related to interest value of visual stimuli. Science, 1960, 132, 349-350.

Kiesler, C.A. Collins, B.E., and Miller, N. Attitude Change: A Critical Analysis of Theoretical Approaches. New York: John Wiley & Sons, 1969.

Leiderman, P.H., and Shapiro, D. (Eds.). Physiological Approaches to Social Behaviour. Standford, Calif: Standford University Press, 1964.

Osgood, C.E., Suci, G.J., and Tannenbaum, P.H. The measurement of meaning. Urbana: University of Illinois Press, 1957.

Pareek, U., and Rao, T.V. Handbook of Psychological and Social Instruments. Baroda: Samashti, 1974.

Scott, W.A. Attitude Measurement. In G. Lindzey and E. Aronson (Eds.), Handbook of Social Psychology (rev. ed), Vol. II, New Delhi: Amerind Publishing Co. Pvt. Ltd., 1975.

Selltiz, C., Jahoda, M., Deutsch, M., and Cook, S.W. Research Methods in Social Relations. New York: Holt, Rinehart and Winston, 1959.

Thurstone, L.L. and Chave, E.J. The Measurement of Attitude. Chicago, Ill. University of Chicago Press, 1929.

Webb, E.J., Campbell, D.T., Schwartz, R.D., and Sechrest, L. Unobtrusive Measures: Non-reactive Research in the Social Sciences. Chicago: Rand McNally & Company, 1966.

Westie, F.R., and DeFleur, M.L. Autonomic Responses and Their Relationship to Race Attitudes. Journal of Abnormal and Social Psychology, 1959, 58, 340-347.

Young, P.V. Scientific Social Surveys and Research. Bombay: Asia Publishing House, 1960.

Zimbardo, P. and Ebbesen, E.B. Influencing Attitudes and Changing Behaviour. Mass: Addison Wesley, 1970.