Quantitative research in the field of speech concerned with everything from vocal pitch and hearing loss to the impact of plays and speeches on audiences, has mushroomed in recent years. Frequently readers of articles reporting these studies are hard-put to understand what the researcher has done.

One of the comparatively new, and to some, esoteric, instruments currently being utilized in speech is the semantic differential. While originally designed to measure meaning, the semantic differential has also become the leading instrument used to measure attitudes. Those who developed this instrument believed that people tend to classify things on bipolar continua; from good to bad, from weak to strong, from fast to slow, etc. Osgood, Suci, and Tannenbaum in their book, The Measurement of Meaning, theorized that if we were able to find where on a variety of continua a person classified an object, idea, or other person, we would then have an objective measure of that thing's meaning to the rater.

Of course, the question was: How many continua do we have to measure to describe meaning adequately? Through extensive use of factor analysis (a statistical method for finding groups of continua which are highly correlated with one another but are not correlated with other continua), Osgood and his associates determined that meaning could be described by as few as three continua. These decisive continua were labeled "evaluate" (good-bad), "activity" (active-passive), and "potency" (strong-weak). Subsequent research indicated that these three dimensions of meaning were common even to people in vastly disparate cultures--American, Japanese, Indian, etc. What we commonly call "attitude" was found to be represented by the "evaluative" dimension of meaning. Thus the semantic differential emerged as a measure of meaning and attitude.

This remarkable tool of research is a very simple, straight-forward instrument. It usually consists of from four to twelve pairs of polarized adjectives e.g. good-bad) with each pair separated by seven spaces, thus:

The person being tested marks an "X" in the space between the adjectives that best represents his reaction to a given object, thus:

His responses are then scored. Each pair of polarized adjectives contributes a portion of the score. For the seven spaces above we may assign the numbers 1-7 from left to right. Our subject would then receive a score of "1" on DEMOCRACY.

If the researcher is concerned with attitude, he will include only pairs of adjectives that represent the evaluative dimension of meaning. Thus, if he includes six pairs of adjectives, a subject's attitude score is summed across all six pairs and may range from 6 to 42. If he is concerned with meaning, the researcher will include pairs of adjectives representing the three dimensions of meaning. Thus the subject responding to the semantic differential could receive three scores, one for each of the three dimensions of meaning. If attitude is to be measured, a subject score 42 on the evaluative dimension for a concept such as "Viet Cong," indicating that the subject considered the Viet Cong very bad and negative.

This score can then be compared with scores recorded on other concepts like "Communism" or "Gangster." The score on the concept "Viet Cong" can be compared with the scores of other subjects on the same concept. In this way, differences in attitude are determined quantitatively. We can, therefore, say that for given concept, people have similar (or very different as the case may be) attitudes. Finally, we can compare the scores recorded by one person on the same concept at different times. If a speech is presented to a listener between two measures of his attitude, the amount of attitude change produced by that speech can be recorded.

Basically, the semantic differential has been used by researchers in speech of four kinds of problems. First, it has been used to measure the credibility of speakers. Second, to measure listener attitudes. Third, it has received some use as a means of assessing speakers and speeches in classroom situations. Finally, the semantic differential has been used to assess the worth of speech courses to the students.

The measurement of credibility by the semantic differential has been accomplished in two ways. The initial credibility of the speaker has been measured before the actual speech was presented. In this procedure, either an introduction was given to the audience or it was assumed that the speaker was known to that audience. In either case, the experimenter asked the audience to rate the speaker on the semantic differential before they heard the speech. The effect of the message on the speaker's credibility can thus be determined. This procedure leads to the second credibility measurement taken after the presentation of some stimuli like a persuasive speech. Both initial and terminal measurements have been used by speech researchers.

As in the measurement of credibility, one semantic differential with six bipolar pairs of adjectives can be used to measure a number of different listener attitudes. For instance, the same semantic differential can be used to measure people's attitudes on dating, drinking, speaking, and draft dodging. At this writing, the semantic differential has proven useful in the measurement of over 200 different concepts. So far, it has proven to be valid and reliable on every concept reported in the literature to have been tested.

Much less use has been made of the semantic differential for critical purposes in the classroom. It has been used to assess a student's credibility as well as his effectiveness in speaking. Unfortunately the semantic differential is of little value for substantive criticism of the speaker for little can be learned or applied by a speaker when he is told that his speech was good or bad, honest or dishonest. The bipolar adjectives do not tell the speaker what he needs to improve or continue to use.

Finally, some research has been conducted with the semantic differential to compare student reactions to their speech course. Their reactions were recorded at the beginning of the course and then again at the end of the term. Changing attitudes toward the course can be reflected in a comparison of these scores. Several semantic differentials are used to evaluate instruction, course content, textbook, and the instructor.

In order to acquaint the reader with the mechanics of the semantic differential, several sample scales are presented. You are invited to try these scales.

Now that you have competed the scale, place a number value for each of the bipolar adjectives (from "1" on the left to a value of "7" on the right). Sum these six scores. A total score of twenty-one is the neutral point of the six-step scale. If your score is over 24, your attitude is declared favorable toward the concept. If your score is under 24, your attitude is assumed to be unfavorable. The lower or higher your score, the more extreme your attitude.

Try example two:

Follow the same procedure that you did for example one. You are now invited to make some of the comparisons suggested earlier in this article.

Readers are invited to submit--anonymously--total scores on each of the three concepts to the first author of this article: Dr. William Arnold, Speech Department, University of Connecticut, Storrs, Connecticut. Your responses will be compared with all others that are submitted, and the results will be published in Column One of the next issue.

Click Here To Go Back To PERIODICALS