Order Effects on Social Attitude in an Eastern Culture: Implications For Cross-Cultural Consumer Research

Yung-Chien Lou, National Chengchi University
Shih-Fen S. Chen, Brandeis University
EXTENDED ABSTRACT - Market researchers have long been aware that responses to an earlier measurement may influence the answers to subsequent items. Answers to a question can be affected by the context activated by the previous questions and thus do not provide a reliable and valid indication of true attitudes and beliefs. The findings of order effects on the part-whole combination of questions are consistent in that specific questions are free from order effects while general questions are the focus of the effects. Nevertheless, the order effects on the general question are somewhat inconsistent. In some studies, the responses to the general questions are assimilated to the responses given to the specific ones, whereas in others they are contrasted to the preceding responses. Researchers have suggested that time pressure, communicative context, and the number of the specific questions preceding the general question may determine whether or not a piece of information activated by the prior specific question will be used to evaluate a subsequent general question.
[ to cite ]:
Yung-Chien Lou and Shih-Fen S. Chen (2002) ,"Order Effects on Social Attitude in an Eastern Culture: Implications For Cross-Cultural Consumer Research", in NA - Advances in Consumer Research Volume 29, eds. Susan M. Broniarczyk and Kent Nakamoto, Valdosta, GA : Association for Consumer Research, Pages: 386.

Advances in Consumer Research Volume 29, 2002     Page 386

ORDER EFFECTS ON SOCIAL ATTITUDE IN AN EASTERN CULTURE: IMPLICATIONS FOR CROSS-CULTURAL CONSUMER RESEARCH

Yung-Chien Lou, National Chengchi University

Shih-Fen S. Chen, Brandeis University

EXTENDED ABSTRACT -

Market researchers have long been aware that responses to an earlier measurement may influence the answers to subsequent items. Answers to a question can be affected by the context activated by the previous questions and thus do not provide a reliable and valid indication of true attitudes and beliefs. The findings of order effects on the part-whole combination of questions are consistent in that specific questions are free from order effects while general questions are the focus of the effects. Nevertheless, the order effects on the general question are somewhat inconsistent. In some studies, the responses to the general questions are assimilated to the responses given to the specific ones, whereas in others they are contrasted to the preceding responses. Researchers have suggested that time pressure, communicative context, and the number of the specific questions preceding the general question may determine whether or not a piece of information activated by the prior specific question will be used to evaluate a subsequent general question.

However, past literature of the order effects on part-whole combinations of questions has been exclusively conducted in the western culture and thus the generalizability of the result to a multi-culture context, especially to an eastern culture, remains unknown. In addition, Researchers in cross-cultural psychology have also found a systematic variation in communication styles across cultures. Therefore, the underlying process of answering part-whole questions may be culture specific. People from different culture may have different interpretation, and/or processes of the information at hand and hence influence their attitude and purchase behavior.

Therefore, results from previous research on order effects in the western culture may not be applicable to the members of the eastern culture. Specifically, in a field-independent culture, respondents tend to use an abstractive and factual-inductive reasoning approach and thus apply conversation norm to interpret information. This processing mechanism results in the contrast effect when question order is varied. However, in a field-dependent culture, such as in Taiwan, individuals react to a judgment task in a holistic way and consider all the information at hand. Thereby, individuals in Taiwan are less likely to disregard information activated by the previous specific question in responding to the subsequent general question.

In this article, we systematically examine how individuals in an eastern Asian culture respond to specific questions en route to forming judgment on a later general question. Two split-ballot experimental self-administered surveys were conducted to test the general hypothesis.

Study I shows that the order effect on the general abortion item is totally reversed. That is, when it was asked after the specific abortion item, respondents in Taiwan are more likely to say yes to the general abortion item (81.13 percent) than when it was asked before the specific abortion item (67.86 percent). It is clearly an assimilation effect on the general abortion item. Study II shows that the correlation between general and specific questions increased from .138 in the general-specific condition, to .622 when the general question followed the specific question. It is comparable to a nearly identical correlation of .615 when respondents were instructed to consider specific situation in making their judgment.

The results obtained in both Experiments I and II suggest that members of East Asian cultures differ from members of Western cultures in forming opinions. In a conversation context, members of Western cultures tend to subtract relevant information activated by the previous question in attitude formation. On the other hand, members of East Asian cultures adopt addition strategy and consider all the information at hand to form their judgments.

The present research has important implications for models of consumer information processing and provides a basis for a cross-cultural extension of comprehensive models of information processing in consumer judgment and in survey as well. In addition, our results of cross-cultural difference in the response process hold important implications for questionnaire design and survey procedures.

----------------------------------------