International Marketing Research: Part 1
(KFK International Marketing)

Dr.Petra Riefler Chair of International Marketing

Comparison of different response styles based on cross-cultural differences

WS 2012/2013

Nikolina Saric (0409312)

Abstract
The variety of response styles based on national and cross-cultural differences and their impact on the interpretation of the research findings is of a significant concern for international marketing researchers (Diamantopoulos et al., 2006; Dolnicar & Grün, 2007; Rocereto et al., 2012). Whether the response styles remain stable across altering stimulus formats and whether ignoring the influence of different response styles would make research findings reliable or not was tested in Diamantopoulos et al.´s (2006) study. The Rocereto at al. (2012) investigated the impact of different stimuli on the degree of extreme response style (ERS) or “the respondents´ tendency to use the extreme categories on rating scales” (Harzin et al. 2011, p.3), by using semantic differential and Likert- type scales. Dolnicar & Grün (2007) on the other hand, extended the investigation by analyzing full response patterns instead only extreme values. The purpose of this paper is to highlight the new insights the above mentioned papers delivered, to find out on which points the authors agreed/disagreed and to emphasize the meaning of results for the future work.

Introduction and Findings
The major issue of international researches, according to Diamantopoulos et al. (2006), is that international researchers cannot rely on the stability of the response styles across alternative forms of scale types, data collection methods etc. Researchers should be aware of the consequences that routinely adapted stimulus formats have on their research findings and interpretation of data collected. Several types of response styles mostly displayed in the data are: extreme response styles (ERS) and acquiescence (ARS) or disacquiescence (DRS) response styles, which are “the tendency to agree or disagree with an item regardless of the content” (Harzin et al., 2011). Response styles have an enormous impact on research findings as they cause various problems; they affect the scale reliability, as they can inflate (ARS) or deflate (DRS) respondents´ scores on the measuring scale. They can also increase the response range (ERS) and affect the correlations between scales. (Diamantopoulos et al., 2006; Dolnicar & Grün 2007; Harzin et al., 2011; Rocereto at al., 2012). Diamantopoulos et al.´s (2006) study compared different response styles among different nationalities using different options of item wording (just positively worded items on one scale and both, positively and negatively worded items on second scale) measured on Likert-type scale. Results showed that there is no stability of response styles when altering the stimulus involved.

This occurred either due to instruments used (Likert-scale) or because it is not researched, whether different items have equivalent levels of meaningfulness in crosscultural environment. Most studies have investigated response patterns using Likerttype scale, even though this type of format is argued to be not the best stimulus format, especially not in cross-national studies (Rocereto et al., 2012). Rocereto et al.´s (2012) study showed, that using semantic differential scale as a stimulus format, leads to more stable results then using Likert-type scale. Likert-scale tends to increase levels of ERS, maybe because respondents are not as involved as when using other scale types. Agreeing or disagreeing with the statements does not involve much thinking, therefore, Likert-scales are not seen as reliable as SD scales where respondents has to pay closer attention when connoting the specific item with an attribute. The sample through which the Hypothesis1 of Rocereto at al.´s (2012) study were approved, included the USA as individualistic2 culture (self-centred cultures where everybody is mostly concentrated on their individual goals) and South Korea as collectivistic3 culture (cultures that think more in terms of „we „rather than „I“). Questions measured subject´s degree of ethnocentrism and attitude towards foreign advertisements. As USA is an individualistic culture, where individuals stand out of the group and value more the independent action and thoughts, the US respondents tend to choose more extreme ends of rating scales as compared to Asian, collectivistic culture (Confucian philosophy).

Conclusion
Authors of Diamantopoulos et al.´s 2006; Dolnicar & Grün´s, 2007; Rocereto et al.´s 2012 studies all agreed that there are different response styles affecting the research findings and analysis. All three papers include reasons why not to use Likert- type scales and problems that occur when using them.4 They also implement some opportunities for future work such as: how other Asian countries, others then South Korea respond to the two different scales (Likert and SD); need for further investigations into SD scales, as they proved to be more reliable then Likert-scales; whether different items have equivalent levels of meaningfulness in cross-cultural environment and whether the relationships between response styles can be generalized.
1

H1a. When using Likert-type scale formats, both U.S. and Korean respondents will exhibit greater levels of ERS bias than when using SD scale formats. H1b. When using Likert-type scale formats, U.S. respondents will exhibit greater levesof ERS bias than Korean respondents. H2. When using SD scale formats, there will beno differences in ERS bias between U.S. and Korean respondents. 2 International Business Cultures, XIII. Individualism versus collectivism from http://www.via-web.de/individualism-versuscollectivism/ on November 11, 2012 3 International Business Cultures, XIII. Individualism versus collectivism from http://www.via-web.de/individualism-versuscollectivism/ on November 11, 2012 4 Problems are stated in the „findings“ section

REFERENCES
Bearden William O, Netemeyer Richard G, Teel Jesse E. Measurement of consumer susceptibility to interpersonal influence. J Consum Res 1989;15 (4): 473-81. Diamantopoulos, A., Reynolds, N.L. and Simintrias, A.C. (2006). The Impact of Response Styles on the Stability of Cross‐National Comparisons. Journal of Business Research, 59(8), pp. 925‐935. Dolnicar, S., Grün, B. (2007). Cross‐Cultural Differences in Survey Response Patterns. International Marketing Review, 24(2), pp. 127‐143. Dolnicar, S., Grün, B., Leisch, F., Rossiter, J. (2011) Three good reasons NOT to use five and seven point Likert items. Downloaded on November 9, 2012 from http://ro.uow.edu.au/cgi/viewcontent.cgi?article=1821&context=commpapers Harzing, A., Brown, M., Köster, K., Zhao, S. (2011) Response styles differences in crossnational research: dispositional and situational determinants. Downloaded on November 12, 2012 from http://www.harzing.com/download/rsdispo.pdf Rocereto, J.F., Puzakova, R.E., Anderson, R.E., Kwak, H. (2012). The Role of Response Formats on Extreme Response Style: A Case of LIkert‐Type vs. Semantic Differential Scales. In Sarstedt et al. (eds): Measurement and Research Methods in International Marketing (Advances in International Marketing, Volume 22), pp. 53‐71.