Response Styles

Along with faithfully translating questionnaires and other communications, a challenge of multi-cultural research is handling response styles that vary by country. Response style refers to a respondent’s tendency to respond systematically to questionnaire items regardless of their content. The most common styles are:

1. Acquiescence Response Style (ARS): The tendency to agree with a question regardless of content.



2. Extreme Response Style (ERS): The tendency to select the end points of a response scale, e.g., for a 10-point scale, to provide ratings of 1-3 or 9-10.



3. Middle Response Style (MRS): The tendency to use the middle response category, e.g., to provide ratings of 4-8 on a 10-point scale.

Why do these response styles differ between countries? The explanation lies in each country’s distinct cultural differences. A framework developed by Geert Hofstede, cited in Anne-Wil Harzing’s article, "Response Styles in Cross-National Survey Research A 26-country Study, in the International Journal of Cross Cultural Management", offers the most cited description of these cultural differences. Using a study of 116,000 employees of IBM Corporation operating in more than 40 countries, he identified four dimensions of national culture that impact response styles:

1. Power distance: The degree to which people in a country accept a hierarchical or unequal distribution of power in organizations. High power-distance cultures are more authoritarian and thus expect conformity, leading to a greater acquiescent behavior style (ABS).



2. Uncertainty avoidance: The degree to which people prefer structured vs. unstructured situations. Cultures high in uncertainty avoidance prefer unambiguous situations and are therefore more likely to use the endpoints of the scale as opposed to the middle, thus exhibiting an extreme response style (ERS).



3. Individualism: The degree to which people in a country focus on working as individuals vs. working together. Cultures high in individualism are less likely to exhibit an acquiescent response style (ARS) because they would emphasize their individual opinion as opposed to their perception of the group opinion. Among all the response styles, individualistic cultures may exhibit extreme response styles (ERS). On the other hand, low individualistic cultures or collectivist cultures emphasize harmony and thus exhibit a middle response style (MRS).



4. Assertiveness: The degree to which people in a country emphasize traits such as assertiveness and insensitivity to feelings. One could hypothesize that individuals in these cultures would favor more extreme response styles and that “softer,” more “sensitive” cultures, exhibit more modesty or middle response styles. However, research is inconclusive with regards to the impact of this dimension on response styles.

Variations by Country and Region

Based upon these cultural dimensions, general observations can be formed regarding different regions of the world.

Anglo, Nordic, Western and Latin Europe: Anglo (USA, Canada, Australia), Nordic and Western European cultures tend to be high in individualism and low in power distance and thus tend to use all points on the scale, as they are less concerned with any social consequences of strong opinions (Table I). Latin European countries on the other hand have low individualism and high uncertainty avoidance. These attributes will results in higher levels of ARS and possibly ERS, given the fact that the power distance dimension is not "low." The middle response style (MRS) is not noted for any of these groups.

Far East and Japan: In countries such as China, Hong Kong and Japan, which test low in individualism and where conformity is stressed, a middle response style is more common. Another observation is that such cultures tend to think dialectically in that they see both good and bad attributes in their interactions, i.e., they don’t see everything in black and white, hence the use of the middle of the scale. Interestingly, China and the Philippines share the same levels of Hofstede’s cultural dimensions but show differing response styles, with China having a higher MRS and the Philippines showing higher ERS and ARS. In contrast, cultures such as Singapore, Malaysia and Hong Kong are all high in power distance and therefore exhibit more acquiescent response styles (ARS) as opposed to middle response styles (MRS). India is also a quite distinct region, exhibiting both ERS and ARS.

Latin America. Latin cultures are high in power distance and thus are more likely to exhibit ARS, as their culture emphasizes conformity via deferential or acquiescent behavior. These cultures are also high in uncertainty avoidance, which emphasizes rules and intolerance for ambiguity. Research suggests that cultures that are high in uncertainty avoidance prefer the endpoints of a scale because they are more definite and clear. Of note is Brazil, which demonstrates a mix of Hispanic and Northern/Western European response styles owing to its distinctive subcultures. As a result, it is difficult to categorize as typically "Latin" or "European."

Cultural Differences or Real Differences in Products, Services, and Delivery?

When analyzing global customer feedback, the researcher has to judge whether the Asian responses vs. the European or US responses are based on different levels of performance, or are merely a result of differences in scale usage. If it determined that these differences cannot be explained by differences in products, services, or delivery of same, how can we best handle these cultural differences in customer satisfaction and loyalty research? While some researchers have come up with complex algorithms to adjust scores across countries, there is no universally accepted approach.

One suggestion is to avoid cross-country comparisons. Rather than directly compare results across countries, compare your results for:

• Sub-segments of the country relative to each other
• The country relative to the same country’s results for a baseline time period (baseline gap)
• The country relative to industry leaders for the same period (benchmark gap)

If it is determined that cross-country comparisons are needed or that data across all countries will be incorporated in an enterprise-wide index, then the following two approaches help to eliminate cultural differences from the comparisons.

1. Create a global performance index, weighting each country by revenue contribution, business potential, or strategic importance. If desired, normalize the index so its value is 100 for the baseline period, for example, the second quarter of calendar 2007. Then, in each subsequent quarter, calculate the index using the latest quarterly scores from each country but use the same weights for each country. In the third and fourth quarters of 2007, for example, the values might be 103 and 105, respectively. This would indicate improvement in overall global performance, independent of cultural differences. From time to time, but infrequently, country weights may be revised to reflect changes in revenue mix, business potential, and strategy.



2. Relate respondents' scaled satisfaction responses (e.g., 1-10 on a 10-point scale) to a "yes/no" question such as willingness to recommend or willingness to repurchase. For example, we may find that 100% of respondents who gave a '10' also are willing to recommend the product, 90% of respondents who gave a '9' also are willing to recommend the product, 85% who gave an '8' are willing to recommend the product, and so on. We then create a graph for each scale point, fit a line to the points, and the resulting graph will be the relationship between satisfaction and the probability of recommending.

The percentages are then compared across countries and new rating scores are calibrated. For example, let's say that 85% of respondents in Country A who are willing to recommend the product gave a satisfaction rating of 8. This can be compared to Country B where 85% of respondents who are willing to recommend gave a 6. In this way of relating a binary response (yes/no question) to our satisfaction scale, we are able to compare cross-cultural responses.

Sources:

Harzing, Anne-Wil Response Styles in Cross-national Survey Research A 26-country Study MInternational Journal of Cross Cultural Management, 2006 Vol 6(2): 243–266.

Helgeson, Neil Research isn’t linear when done globally. Marketing News, July 1999.

Johnson, Timothy et al. THE RELATION BETWEEN CULTURE AND RESPONSE STYLES Evidence From 19 Countries, Journal of Cross-cultural Psychology Vol. 36 No. 2, March 2005 264-277.

Kitayama, Shinobu Cultural Shaping of Basic Psychological Processes: Implications for Survey Research ShethFoundation/Sudman Symposium, The University of Illinois, Urbana-Champaign, October 1, 2004.

Schermerhorn, John R Managing Organizational Behavior, John Wiley and Sons, 1994 5th Ed.