|
|
|
 |
||||||||
Sheth Foundation/Sudman Symposium |
Honoring the memory of
Seymour Sudman and commemorating the 40th anniversary of the Survey Research Laboratory |
|||||||
September 30 - October 2, 2004 |
||||||||
Illini Union, University of Illinois at Urbana-Champaign |
||||||||
| Held September 30 through October 2, 2004, the Sheth Foundation/Sudman Symposium on Cross-Cultural Survey Research featured leading scholars spanning multiple disciplines who discussed current research on global/cross-cultural survey measurement, all of whom are listed below. Clicking on a talk's title will take you to that talk's slides. |
|
Thank you to the SYMPOSIUM SPONSORS |
||||||
| Symposium Program (PDF) | ||||||||
| SPECIAL GUESTS | ||||||||
• Distinguished Introductory Speaker |
||||||||
| Jagdish Sheth (Marketing, Emory University) Dr. Sheth's bio | ||||||||
| Emerging Opportunities & Challenges in Cross-Cultural Survey Research | ||||||||
| Emerging economies such as China, India, Brazil, and Russia will be the growth engines of the world in the 21st century. Most of the market opportunities in these economies are nontraditional and at the bottom of the pyramid. However, most of cross-cultural survey research has been traditionally limited to educated people in metropolitan areas of these emerging economies. This gap between survey research opportunities and survey research methods will be the focus of this presentation. It will suggest several non- intuitive survey research techniques and processes appropriate for the rural and illiterate populations of the world. |
||||||||
| • Keynote Speaker | ||||||||
| Harry Triandis (Psychology, University of Illinois at Urbana-Champaign) | ||||||||
| Culture & Response Sets | ||||||||
| A review of three dimensions of cultural variation will be followed by arguments about the way each dimension is related to acquiescence and extreme checking style. The relevant empirical literature will then be examined to see where the theoretical argument is and is not supported, and suggestions will be made about needed research that will clarify which relationships between culture and response sets can be established. | ||||||||
| • Featured Speaker | ||||||||
| Fons van de Vijver (Social & Behavioural Sciences, Tilburg University) | ||||||||
| Toward an Understanding of Cross-Cultural Differences in Acquiescence & Extremity Scoring | ||||||||
| The prevalence of response styles in both single-country and cross-cultural surveys is widespread. Yet a comprehensive theoretical framework is missing. At least four partly related issues should be addressed in such a framework. First, it is not clear to what extent response styles such as social desirability and acquiescence reflect differences in style (“face management”) or substance (agreeableness). Second, it is not clear to what extent response styles differ across content domains (e.g., do they increase with the level of personal involvement?). Third, what is the relative contribution of individuals, countries, and domain of study to the overall score variation? Fourth, what are correlates of response styles at individual and country level? Re-analyses of data from the International Social Survey Programme (www.issp.org) are used to examine the patterning of acquiescence and extremity scoring and to address the above questions. Multilevel modeling as well as an examination of country-level correlates of indicators of these response styles are used to address their psychological meaning. | ||||||||
| SYMPOSIUM SPEAKERS | ||||||||
| Hans Baumgartner (Marketing, Penn State University) | ||||||||
Issues in Assessing Measurement Invariance in Cross-National Research |
||||||||
Before measures of psychological constructs can be compared across cultural groups, it is necessary to ascertain the equivalence of these measures. In this presentation, we will review some of the basic principles of assessing measurement invariance, using multi-sample structural equation modeling as a methodological tool, and discuss several issues that frequently pose problems in empirical research. These include choosing appropriate reference indicators for a given construct, deciding whether a factor loading or measurement intercept is sufficiently invariant, and dealing with non-invariant items depending on the comparison of interest. Special attention will be given to situations in which measurement invariance has to be assessed across many different groups simultaneously. To illustrate these ideas, we will report on a cross-national study about subjective well-being (life satisfaction, positive and negative affect) involving over 13,000 respondents from 28 different countries. |
||||||||
| Chi-Yue Chiu (Psychology, UIUC) | ||||||||
| An Intersubjective Consensus Approach to Measuring Cultural Values & Beliefs | ||||||||
| According to social representation theory, people in a culture
collaboratively construct an intersubjective reality, which often does
not correspond to the objective reality. Based on this theoretical
perspective, my colleagues and I have developed an intersubjective
consensus approach to assessing cultural values and beliefs, which
focuses on how people appropriate shared knowledge to guide their
responses to psychological surveys. To illustrate the utility of this
approach, I will present data from several cross-cultural studies of causal attribution, cultural identification, socially desirable responding, and regulatory focus. |
||||||||
| Paul Hanges (Psychology, University of Maryland) | ||||||||
| Measurement Equivalence with Multilevel Constructs | ||||||||
It is recommended practice in cross-cultural research that the measurement equivalence of an instrument be assessed before conducting any hypothesis tests. Measurement equivalence is established by conducting a multi-group confirmatory factor analysis in which the factor structure explains individual variability within each cultural group. While useful, this analytic approach is only applicable for scales measuring individual level constructs (e.g., personality, satisfaction, intelligence). The direct application of this technique to scales assessing group level constructs (e.g., team potency, organizational culture) is problematic. Further, no one has discussed how to establish the measurement equivalence of instruments assessing societal-level constructs (e.g., societal culture). These latter type of constructs are assumed to have little, if any, meaningful within cultural-group variation. This presentation will discuss the utility of multi-group, multi-level confirmatory factor analyses to establish the measurement equivalence of group-level and societal-level constructs. |
||||||||
| Janet Harkness (ZUMA, Mannheim) | ||||||||
The paper begins with questions in translation and presents an overview of what happens to questions in the course of translation, that is, how questions change (and in part why they change) when they are translated. Turning next to Questions in Translation, the paper examines a number of standard assumptions about question meaning and about questionnaire translation and assessment and addresses several of the issues raised by the need to translate questions. |
||||||||
| Steven Heine (Psychology, University of British Columbia) | ||||||||
| Considering the Validity of Cross-Cultural Comparisons of Self-Report Measures | ||||||||
| Cross-cultural psychological research is inherently challenging, and each method has its own shortcomings. The most commonly used method, comparisons of subjective self-report measures, is especially problematic. Considerations of various methodological obstacles that significantly compromise the validity of this approach, with particular focus on the reference-group effect, will be discussed. Alternative cross-cultural methods will be evaluated. | ||||||||
| Ying-yi Hong (Psychology, UIUC) | ||||||||
| Combining Nomothetic & Idiographic Approaches to Understand the Bicultural Self | ||||||||
| With rapid globalization in the 21st century, people nowadays are constantly
exposed to
cultures other than their own. How would individuals who have been
extensively
exposed to two cultures (bicultural individuals) represent cultures? In the
research
conducted in my laboratory, we have examined how Asian-American bicultural
individuals represent Asian and American cultures cognitively and emotionally.
Both
qualitative methodologies, such as structured interviews, and quantitative
methodologies, such as priming experiments, have been used in our studies.
Consistently, the findings revealed two processes: one related to cultural
knowledge
acquisition, and the other related to social identity affirmation. I will
discuss how this
work can illustrate the processes of bicultural self. |
||||||||
| Shinobu Kitayama (Psychology, University of Michigan) | ||||||||
| Cultural Shaping of Basic Psychological Processes | ||||||||
| Do people engaging in different cultures develop different psychological processes and, if so, how? In order to address these questions, I will distinguish between two divergent modes of self-regulation. In an independent mode of regulation, individuals use representations about their personal attributes (e.g., traits, goals, and abilities) in regulating their own actions, whereas in an interdependent mode of regulation, individuals use representations about social others (e.g., expectations and desires held by others in a relationship) in regulating their own actions. Although these two modes are likely to be available to all individuals in all societies, they are distributed unevenly across cultures such that the independent mode is more prevalent in European-American middleclass cultures and the interdependent mode is more prevalent in Asian cultures. I will illustrate these two modes of self-regulation by reviewing evidence from disparate domains including accessibility of self- vs. other-cognitions, motivational orientations of promotion vs. prevention, happiness and life satisfaction, and cognitive dissonance. I will further argue that early socialization of culturally appropriate styles of attention precedes the development of the two modes of self-regulation. Implications for survey research methods are discussed. | ||||||||
| Patrick Kulesa (International Survey Research, LLC, Chicago) | ||||||||
| Why Culture Trumps Strategy: The Influence of National Culture on Employee Attitudes | ||||||||
| Large-scale survey research from multinational organizations shows that national culture exerts a strong and consistent influence on employee attitudes. This influence often exceeds the effects of industry sector, organizational goals, and management level on employee sentiment. National culture is studied using the value scales developed by several researchers, including Hofstede and Trompenaars. Differences in value orientations underlie differences in employee attitudes. For instance, employees from high power distance and collectivist cultures tend to have more favorable attitudes toward the workplace and company leadership. Value orientations also impact survey response patterns, moderate the attitude gaps between managers and non-managers, and interact with the strategies used by multinationals to address culture. Although these results do not imply one best approach to managing culture, specific implications can be drawn for measuring employee attitudes across nations and addressing the challenges faced by multinational organizations as they globalize their operations. | ||||||||
| Angela Y. Lee (Marketing, Northwestern University) | ||||||||
| Roots of the Interdependent Self: Family Size & Self-Definition in a Changing China | ||||||||
| When the People's Republic of China instituted the "one-child policy," it embarked on a social experiment of grand scale. In addition to curbing population growth by modifying family structure, it seems that this policy may have unintended consequences for cultural identity development and world views. In one study, participants from China, Hong Kong, and the U.S. were asked to complete a brief self-definition task. In both Hong Kong and the U.S., participants born before and after 1979 did not differ in the expression of culturally consistent self-views. However, in the PRC, participants born before 1979 held self-views similar to those expressed by the Hong Kong Chinese, whereas those born after 1979 held self-views similar to those from expressed by North Americans. The role of family size in the formation of self-views was further examined in the world values survey data. Across 44 countries, parents' endorsement of interdependent vs. independent values appears to vary according to the number of children they have. Implications of self-views on subjective well-being will also be discussed. | ||||||||
| Durairaj Maheswaran (Marketing, New York University) | ||||||||
| Culture & Persuasion | ||||||||
| The presentation will review the current trends in cross-cultural persuasion from a marketing perspective. Several studies that examine advertising effectiveness across cultures will be briefly discussed, and a framework for understanding cross-cultural differences will be outlined. | ||||||||
| Peter Ph. Mohler (ZUMA, Mannheim) | ||||||||
| Recognizing the Obvious OR: Getting Your Scales Right in Cross-Cultural Surveys |
||||||||
| Item frugality is the guiding principle in constructing questionnaires for social surveys. Consequently, scales used in social surveys are short compared to those in psychological experiments. Cross-cultural psychology has established some standards of eliminating items showing nonequivalence in comparative analysis. However, in mult-national studies, such as the International Social Survey Programme (40 nations), eliminating all items that are not equivalent in each and every culture will result in few if any remaining items. Recognizing the obvious means accepting that translation, like all other cognitive procedures, cannot establish statistically sound equivalence. This paper will discuss such a best practice strategy that will (1) allow for a limited abundance of items in order to identify for each country the most salient short form, (2) take advantage of developments in machine translation, namely the interlingua or bridge language approach and, (3) use the logical rule of A=B and A=C thus B=C for comparisons. |
||||||||
| Daphna Oyserman (Psychology, University of Michigan) | ||||||||
| Rethinking Individualism, & Collectivism: Self-Construal, Cognition, & Communication | ||||||||
| Cross-national differences in individualism and collectivism and a general model for understanding the process by which they occur are reported. As a test of some of the assumptions of this model, priming experiments that temporarily induce independent vs. interdependent self-construals are summarized focused on influence on perception, memory, cognition& communication. Methodological implications for cross-cultural studies are highlighted and implications for the conceptualization of cultural differences addressed. | ||||||||
| Sharon Shavitt (Business Administration, Survey Research Laboratory, UIUC) and Timothy P. Johnson (Survey Research Laboratory, UIC) |
||||||||
| The Relation Between Cultural Orientation & Socially Desirable Responding | ||||||||
The cross-cultural literature suggests that collectivists are more likely to engage in socially desirable responding (SDR) in order to maintain good relations with others. In contrast, individualists are portrayed as candid and sincere because individualism encourages people to "be yourself." We propose that people with both types of cultural orientation engage in desirable responding, albeit in distinct ways. Across several studies, people with an individualistic cultural orientation appear more likely to engage in self-deceptive enhancement, the tendency to see oneself in a positive light and to give inflated assessment of one's skills and abilities. Collectivists are more likely to engage in impression management by misrepresenting their self-reported actions to appear more normatively appropriate. Similar results are observed using demographic indicators of culture (race, ethnicity). Additional analyses distinguish horizontal versus vertical forms of individualism-collectivism in order to gain further insight into the self-presentational goals served by SDR. This program of research contributes to survey methodology by (1) examining distinctions between types of SDR and (2) demonstrating that respondents with different cultural orientations use distinct strategies for self-presentation. |
||||||||
| Tom W. Smith (National Opinion Research Center, University of Chicago) | ||||||||
| Methods for Assessing & Calibrating Response Scales Across Countries & Languages | ||||||||
| To advance measurement equivalency, the International Social Survey Program (ISSP) has conducted detailed, national studies on the strength and comparability of response scales commonly used in the ISSP and other surveys. A method for assessing the position on a continuum of response options is demonstrated on two scales (agree/disagree and importance) in three countries (Germany, the US, and Japan). The technique shows that careful translation often yields highly comparable response options, but this is not always the case. The quantitative ranking of response options reveals cases in which supposed equivalent response option are not comparable and therefore the results are not comparable. These results can be used to make response scales more equivalent. | ||||||||
| Jan-Benedict E. M. Steenkamp (Marketing, Tilberg University) | ||||||||
| A Global Investigation into the Individual & Cultural Drivers of Socially Desirable Responding | ||||||||
| Socially desirable responding (SDR) has been recognized as an important source of method error in survey research. Consequently, SDR has attracted much attention in the social sciences. However, most of this important research has been conducted with students and in a single country only (i.e., the U.S,), and many studies have had a relatively narrow focus, studying a limited set of drivers of SDR. Consequently, the contribution of this study is threefold. First, we develop a conceptual model and set of hypotheses incorporating a systematic set of drivers of SDR, at both the individual and national-cultural levels. Our individual level variables consist of structural dispositions of individuals as well as sociodemographic variables. For the national-cultural dimensions, we draw on Hofstede's (2001) framework. Second, the hypotheses are tested in a cross-national context, using over 20 countries located in four different continents (Asia, Europe, North America, South America). Third, our data set, encompassing more than 10,000 respondents, provides a basis for deriving empirical generalizations concerning the drivers of SDR. | ||||||||
| Nancy Wong (Marketing, Georgia Tech) | ||||||||
| Linking ReligiousValues to Brand Behavior Across Cultures: An Exploratory Study | ||||||||
| Is there a connection between our commitment towards our religious beliefs to how we feel towards our brands? Can we measure that commitment reliably across cultures? Based on a review of the religion literature, we investigate two perspectives regarding the relationship between religiosity and brand commitment. We provide a comparative test of these two perspectives through an initial survey of 363 Americans and a follow-up survey of 300 Singaporeans. In this study, we tested different measures (values and behaviors) of religiosity and brand commitments across two cultures, three religions and four product categories. Issues related to measuring values across cultures, response patterns, validity of measures, and scale validations will be discussed. | ||||||||
| Robert Wyer (Marketing, Hong Kong University of Science and Technology) | ||||||||
| Normative & Motivational Influences of Cultural Identity | ||||||||
| People's cultural background can influence the concepts and knowledge they bring to bear on their interpretation of information they receive, and on the opinions and values they report. Furthermore, situational factors that increase the accessibility of cultural knowledge can increase the magnitude of this influence. When people are confronted with a behavioral decision that has positive or negative consequences, however, people's awareness of their cultural identity can have motivational effects. Specifically, by making them conscious of belonging to a group, it can induce a general sense of social responsibility that is manifested in a desire to minimize negative consequences of their decisions. This disposition, which influences choices in both interpersonal and individual choice situations, occurs regardless of the culture to which people belong. Research by Donnel Briley and his colleagues establishes these effects and, therefore, underscores the need to distinguish between the normative influences of cultural identity and its motivational influences. | ||||||||
|
||||||||
| s u r v e y r e s e a r c h l a b o r a t o r y •  •
s u r v e y r e s e a r c h l a b o r a t o r y |
||||||||