Cross-Cultural Methodologies: An Assessment In Four Key Steps

4m ago
23 Views
0 Downloads
319.46 KB
20 Pages
Transcription

Cross-Cultural Methodologies: An Assessment in Four Key Steps

Cross-Cultural Methodologies: An Assessment in Four Key StepsConducting research in the cross-cultural context brings forth many challenges that must beaddressed to ensure that any decision made regarding culture is accurate, meaningful, andgeneralizable to other cultures. This paper addresses these challenges by evaluating foursteps in the identification process of cross-cultural research. These steps are identification ofthe research objective, level of analysis, data collection methods, and data analysistechniques. Options for making decisions about cross-cultural methodology is suggestedfrom a review of sixty cross-cultural articles based on these four identification steps. Aprescribed selection for each step is offered such that the cross-cultural research conductedaddresses the challenges and provides an empirically based solution.Keywords: cross-cultural research; cross-cultural research methodology; methodology;research decisions; level of analysis; survey method; research methods1

Cross-Cultural Methodologies: An Assessment in Four Key StepsSo long as corporations continue to establish operations across national borders, it isimperative that researchers not only assess the validity of existing theories, but also developnew theories in these cross-cultural contexts (Roth & Kostova, 2003). Toward these efforts, itis essential for researchers to assure construct and measurement equivalence (McArthur,2007; Robert, Lee, & Chan, 2006) and to consider conducting research at multiple levels ofanalysis to study these complex organizational phenomena (Kostova, 1999). These crosscultural theories then need to be tested universally (Ember & Ember, 2000). Researchers areencouraged to acquaint themselves with the idiosyncrasies of conducting research in a crosscultural context before deciding to exert effort in such endeavors. Cross-cultural scholarsface numerous challenges when comparing cultural values, beliefs, and behaviors betweentwo or more countries and these challenges influence the decisions undertaken by crosscultural researchers.One question to answer at the onset of a study is the use of etic-emic approaches tocross-cultural research. A common tactic in cross-cultural studies is to take an existing theoryused in one country and extend the theory to another country without considering whether thetheory is relevant or applicable in the new context (Douglas & Craig, 2005); thus suggestingthat social phenomena is culture and context-free, i.e., universal or “etic” (Hantrais, 1999).Proponents of a culture-bound or “emic” approach uphold the view that cross-culturalresearch findings can only be properly understood within the context in which they occur andthat such findings are not amenable to generalization (Hantrais, 1999). A second challengethat must be faced is ecological fallacy or aggregation, which occurs when inferences aboutindividuals are drawn based upon aggregate level data (Robinson, 1950; Kramer, 1983).2

Aggregation may lead to inconsistent findings, as demonstrated by Hofstede, Bond, & Luk’s(1993) research conducted at the organizational level and later re-analyzed at the individuallevel. Their findings suggested that dimensions or factors identified at the organizationallevel were different from those found at the individual level. If measurement equivalence isnot established, cross-national empirical research results will lead to weak interpretations(Bensaou, Coyle, & Venkataraman, 1999). Finally, the challenge of identifying the constructsthat will be measured in the cross cultural context suggests the need for a basic understandingof the type of measures to be undertaken across cultures and how these measures can bemanipulated statistically to uncover the similarities and differences that exist betweencountries. Deciding upon the construct of measure will directly impact the data analysistechniques available to the researcher.The conflicting etic-emic approaches, the challenges of selecting the appropriatemeasures, as well as the challenges of establishing the many forms of equivalence whenconducting cross-cultural research give rise to the importance of decision-making with regardto cross-cultural methodologies. Toward the identification of a framework for cross-culturalresearch, Schaffer & Riordan (2003) identified a three-stage cross-cultural frameworkconsisting of: 1) research question development, 2) research context alignment, and 3)research instrument validation. They suggested four best-practices for the development of thecross-cultural research question. Specifically, they: 1) identified a combined emic and eticapproach, 2) suggested culture should be incorporated into the theoretical framework, 3)proposed the use of other delimiters besides country to operationalize culture, and 4)recommended measurement of Hofstede’s cultural value dimensions directly in the specificresearch context. Similarly, they emphasized the equivalence of samples and administration3

of surveys as best practices for research context alignment. Finally, they underscoredsemantic equivalence, as well as scaling and conceptual equivalence as best practices forresearch instrument validation. These issues of equivalence in the cross-cultural context mustbe addressed to ensure clarity in the conclusions made between cultures.While Schaffer & Riordan (2003) have developed a framework for cross-culturalmethodology, our framework provides prescription for deciding the most appropriateresearch objective, level of analysis, data collection method, and data analysis techniques incross-cultural research. We believe these prescriptions have a cumulative impact on howeffective a study will be in addressing the primary cross-cultural context of the researchquestion. Our research attempts to answer the following question. What is the mostappropriate choice at each step of the identification process (research objective, level ofanalysis, data collection method, and data analysis technique) to ensure the conclusionsresulting from cross-cultural research are generalizable across many cultures?In this study we offer a six-step framework (see Figure 1). We propose that thisframework has two distinct elements for conducting cross-cultural research: 1) researchidentification and 2) research execution. However, we assess only the four identificationsteps of the cross-cultural methodological framework in this study. In order to understandsome of the trends in cross-cultural research relative to these four steps we analyze sixtyarticles to provide us with a baseline. Second, we discuss the importance of understandingthe research objective. Our emphasis here is to highlight the importance of the researchobjective rather than the research question as proposed in Schaffer & Riordan’s (2003)framework. Third, we identify the important findings involving the level of analysis in cross-4

cultural research. Here, our objective is to develop a model that identifies a multi-levelanalytical approach to level of analysis in cross-cultural research.Identify ResearchObjectivetiIdentify Level ofAnalysisIdentificationIdentify DataCollectionMethodIdentify DataAnalysisTechniqueCollect Data &Evaluate ResultsExecutionShare BestPracticesFigure 1: Cross-Cultural Methodology FrameworkFourth, we review the extant literature on data collection methods to identify theimportant methodological issues related to the development of a cross-cultural researchsurvey. Our objective here is to identify past trends in the use of data collection instruments;thus, answering the call by Schaffer & Riordan (2003) to develop efficient cross-culturalresearch surveys. Fifth, we identify the dominant data analysis methods that have evolved incross-cultural research. Finally, we provide recommendations for cross-cultural researchapproaches for each of the four key identification steps in our framework.METHODWe selected sixty articles to get an understanding of the basic trends that are currentlyoccurring in cross-cultural research. We initially used the listing of 200 articles from Schaffer& Riordan (2003) as a source selecting articles for cross-cultural analysis in this research.5

However, in order to ensure the selection of articles beyond the period of study for the paperby Schaffer & Riordan (2003), the sixty articles used in this research for analysis wererandomly selected from a listing based on a search of Business Source Complete and JSTORusing the keywords: cross-cultural analysis and cross-cultural. Selected articles representedthe time period from 1980–2008. We settled on the definition of cross-cultural research asthose studies involving a minimum of two cultures to be compared. We excluded thosearticles that only evaluated one culture and generalized to others.RESEARCH OBJECTIVEWhile Schaffer & Riordan (2003) stress the importance of identifying the researchquestion, our framework identifies the identification of the research objective. Focusing onthe research objective addresses the overarching goal relative to the cross-cultural research.How the researcher aims to accomplish that objective is specified in the research question.The framework proposed offers a cumulative approach toward the identification of whatshould be done at each step to ensure credible cross-cultural research. van de Vijver &Leung’s (1997) taxonomy of cross-cultural studies is the basis of the first step in ourframework (see Table 1).The identification of the research objective consists of identifying whether theresearch will focus on the testing of a hypothesis or exploration of new theories and contextsbetween cultures. It is the consideration of the contextual factors (whether they exist or not)that creates the four types of cross-cultural studies. The first type of cross-cultural study hasthe objective of testing a hypothesis with no consideration of contextual factors. The secondtype of objective identified is hypothesis testing, but this objective does consider contextualfactors. An example of such research would be one in which a particular variable is6

Testing a HypothesisExploration of New TheoriesNo ContextGeneralizabilityPsychological DifferencesContextTheory-DrivenExternal ValidationTable 1: Cross-Cultural Research Taxonomy; Source: van de Vijver & Leung, 1997identified for comparison between cultures. The cultural variation in this case is used to testor validate a cultural theory. The third type of study, called psychological differences, is anexploratory study that has no consideration of contextual factors and does not seek to testhypotheses. The fourth study, external validation, focuses on the factors that cause crosscultural differences.In their discussion of the types of cross-cultural studies, van de Vijver & Leung(1997) continue to offer that each of these four types can have a level and/or structureorientation. The level orientation addresses the size of the difference between cultures, whilethe structure orientation addresses the similarities and differences between the relationshipsof the variables under study. This taxonomy from van de Vijver & Leung (1997) is the firststep in our identification process. We propose that the research objective should takeprecedence in cross-cultural research. While the taxonomy from van de Vijver & Leung’s(1997) is an adequate approach, we prescribe that the cross-cultural researcher selects onlythose objectives with context. Greater generalization leads to conclusions that are morecross-cultural and hence, more universal. Once the objective has been selected, the researchercan proceed to the second step of our framework, i.e., deciding upon the level of analysis.LEVEL OF ANALYSISOur model proposes that one must evaluate a minimum of two levels in cross-culturalresearch. In order to address the fallacy challenge and provide more thorough complexconclusions in the cross-cultural context, one level is insufficient. Lueng s (1989)7

comparison of individual vs. cultural level of analysis concluded that cross cultural researchcould only be done by aggregating at the ecological level. Hofstede, Bond, & Luk (1993)conclude that one has to select the correct level of analysis based on what is being compared.Comparing values can occur at the individual level (Leung, 1989) or using a parallelapproach compared at the individual and cultural level (Smith, 2008). A meta-analysis byTaras & Steel (2006) found that while individual and cultural levels should be considered,each should be considered separately. The relevance of time with regard to level of analysisis addressed by Dansearau, Yammarino, & Kobles (1999). If groups change fromhomogeneous to heterogeneous over time, then one would expect the level of analysis tochange over time as well. While there are many perspectives and approaches to level ofanalysis, we prescribe that a minimum of two levels should be identified for cross-culturalresearch. As such we offer a multidimensional, multilevel model that can be utilized todetermine and select the two levels of analysis.Kostova (1999) highlights the importance of conducting cross-cultural research usinga multilevel approach, emphasizing the individual, country, and organizational levels offocus for cross-cultural research. We agree with Kostova (1999) that not only do all threelevels need to be considered and addressed, but that at least two of the three levels should beselected and planned for in the data collection and data analysis steps. The nature of crosscultural research of course dictates that one of the two levels should be culture. A review ofthe literature shows that research conducted using multiple levels of analysis is still rare.Figure 2 shows the percentages of the sixty research articles categorized by level ofanalysis type. Country-level analysis is represented by 50% of the sixty articles analyzed inthis study; while analysis at the individual le