Scales Make a Difference

There is an article in Quark's marketing Research Review in January written by Kristin Cavllaro. The article is titled "Are global scales as easy as 1-2-3 or A-B-C? The point of her article is that scales make a difference when we are dealing with cultures outside the US.

We normally assume that when we use a survey in a global market, the respondents in the various markets will respond the same as those respondents in the US. According to Ms Cavallaro, that is not true. The article is based on a seven country study which tested the five-point Likert scales in two different manners. The study was performed by Survey Sampling International. The first set of questions used a verbal scale in which all points were described as words. The second set used the same questions but presented the answers as a numerical scale where only the first and last options were anchored or defined. The questionnaire included six different sections; namely,
1. self-comparison to set reference groups
2. women's rights
3. importance of cultures and traditions
4. dependent versus independent thought processes
5. public appearance
6. acceptance/likelihood to purchase a new product concept.

The countries included in the study were the United States, UK, Italy, Japan, China, Brazil and Mexico.

The conclusion drawn by the study are too many to include in this short blog; however, many of the study results are noted in the article. The key result of the study is that the scale does make a difference. Culture has a significant influence on the way people respond to scales. Teh results from Brazil indicate that numeric scales appear stronger than verbal scales. On the other hand Italy, Japan, China and the UK perfromed better with the verbal scale than the numberic. In this case better means a higher score than the data from the numeric scales.

The second major finding was that the difference between the results using the two scales were statistically different. There is a "however" and that is that it is not clear that there is one "best" scale. The validity of the results are not limited ot one scale. In other words, neither scale can be shown to be more correct or true.

The bottom line is the way we use the Likert scale may make a diffrence in the responses on a survey when we are taking measures in multiple cultures. There is no scale that will always out-perform the others. Cultural variations are important and we must continue to accommodate or, at the very least, explain them in any survey that includes other cultures.

The use of scales seems like an automatic and something we should not have to worry about. As it is turning out, the scale may be a new source of error that lead to incorrect solutions. The Customer Institute will continue to follow research in this area.