Comparison of Thurstone and Likert Scaling Methods:
It is obvious that the Thurstone and Likert scaling methods have much in common. They have been two of the most widely used methods; yet there have been relatively few tests of the comparative validity, reliability, and efficiency of the two. The evidence to date indicates the Likert scale is probably more reliable than the Thurstone method, but there is no sound evidence on the comparative validity.
Non-disguised, structured methods of measuring attitudes include various self rating scales of the graphic, ranking and multi-item types. In general they are simple in concept and use, yet seem to give good results. The most significant problem in using such scales has to do with their validity – that is, do they measure what they are presumed to measure? It is difficult to find standards against which to measure validity. The most common effort has been to compare opinions of products with their market shares or with purchases following the opinion measurement. These tests tend to support the validity of various attitude measures, but the correlation is far from perfect. Price is a confusing factor in these validity tests, as the most expensive brand is often considered the best even though many consumers may not choose to pay the high price.
As with questionnaires generally, there are efforts in attitude measurement to obtain the advantages of both structure and disguise. If respondents don’t know the specific topic of a survey, they will not be biased in their responses because of such knowledge.
Ingenious researchers have found ways to gain the advantage of both structure and disguise by trading on the fact the people tend to know more about things that they favor or, if forced to guess factual information, will guess in a direction that is favorable to items or ideas that they favor.
The basic premise underlying such tests is that respondents will reveal their attitudes by the extent to which their answers to objective questions vary from the correct answers. Respondents are provided with questions that they are not likely to be able to answer correctly. Thus, they are forced to guess at the answers. The direction and extent of these guessing errors is assumed to reveal their attitudes on the subject. For example, individuals tend to gather information that supports their attitudes and, therefore the extent and kind of information individuals possess on a given subject indicate something of their attitude.
In the hot cereal study mentioned previously, the disguised, structured technique was used by asking such questions as the following: How much do you think it costs for the hot cereal alone in an average bowl of cereal such as you’d serve at breakfast? (¼ cent or less, ½ cent, ¾ cent, 1 cent, 2 cents, 3 cents, 4 cents, 5 cents, 6 cents, plus, do not know?) Which of the following cereals do you think costs more per serving than a hot cereal, and which costs less? Do corn flakes cost more or less than hot cereal? Do Cheerios cost more or less than hot cereal?
Questions similar to these were asked regarding vitamins, protein content, and the fattening qualities of hot cereals as contrasted to selected brands of cold cereal. Data from these questions revealed, for example, that dry cereal users exaggerated the cost of hot cereal. Considerable differences were evident between regions and city size groups on perceived vitamin content; the larger the city size, the better the vitamin rating for dry cereals.
Disguised, structured techniques have many of the advantages of non-structured methods while eliminating much of the subjective element with which such methods are plagued. The authors have expected the disguised, structured methods to replace gradually the nonstructural methods as researchers become more imaginative in their use. This has not happened, but the potential advantages are large enough to warrant further efforts.