Cross-cultural research is burgeoning. Behavioral and social sciences such as psychology, sociology, management, marketing, and political science witness a steady increase in cross-cultural studies. For example, during the last decades, there has been a consistently increasing number of psychological studies on cross-cultural similarities and differences (Boer, Hanke, & He, 2018; Smith, Harb, Lonner, & Van de Vijver, 2001; Van de Vijver & Lonner, 1995). The increased interest is undoubtedly inspired by various factors, such as the opening of previously sealed international borders, large migration streams, globalization of the economic market, international tourism, increased cross-cultural communications, and technological innovations such as new means of telecommunication.

In the previous chapters, typical problems and pitfalls of cross-cultural research were discussed and solutions proposed. The current chapter briefly integrates the major methodological issues into eight statements. Each statement is followed by an explanation. The last section is devoted to our view on the future of cross-cultural research.

This chapter contains a description of the sampling of cultures, design, data analysis, and major strengths and weaknesses of the eight types of cross-cultural studies described in Chapter 2: structure- and level-oriented psychological differences studies; structure- and level-oriented generalizability studies; structure- and level-oriented contextual linkage exploration studies; and structure- and level-oriented contextual linkage validation studies. The structure- and level-oriented studies differ primarily in the analyses employed, so their presentation is integrated. A schematic overview is given in Table 5.1.

This book addresses the methodological features of cross-cultural research. The common characteristic of such studies is their comparative nature, which involves the comparison of at least two cultural populations. Many studies involve different nation states, in sociology (e.g., Inglehart & Welzel, 2010; Van Deth, Montero, & Westholm, 2007), education (e.g., Arnove, Torres, & Franz, 2012; Van de Werfhorst & Mijs, 2010), political sciences (e.g., Coffé & Bolzendahl, 2010; Poguntke & Webb, 2007), management (e.g., House et al., 2004), and psychology (e.g., Schmitt, Allik, McCrae, & Benet-Martínez, 2007). However, comparative studies can also involve different ethnic groups from a single country such as the comparison of ethnic groups in the United States (e.g., Trinidad, Pérez-Stable, White, Emery, & Messer, 2011) and in Europe (Phalet & Kosic, 2006).

Two closely related concepts play an essential role in cross-cultural comparisons, namely equivalence and bias (Poortinga, 1989; Van de Vijver, 2015). There is no consensual definition of either concept in the cross-cultural literature. Johnson (1998) identified more than fifty types of definitions of equivalence, addressing dissimilar features, such as constructs, methodology, language, and context. All definitions refer to some aspect that is shared across cultures or to a qualitative or quantitative procedure to establish the shared features. A review of bias approaches would probably show a comparable variety.

Four procedures for sampling cultures can be discerned (cf. Boehnke, Lietz, Schreier, & Wilhelm, 2011). In convenience sampling, researchers select a culture simply because of considerations of convenience. These considerations can derive from various sources; researchers may be from that culture, are acquainted with collaborators from that culture, or happen to stay there for a period of time. The choice of culture is not related to the theoretical questions raised and is often haphazard. Studies adopting this sampling scheme often fall into the category of psychological differences studies.

Data analysis in cross-cultural research involves more than the preparation of the correct instructions to run a computer program of a statistical package. It is a link in the long chain of empirical research that starts with the specification of a theoretical framework and ends with drawing conclusions. Strategic decisions in the data analysis such as the choice of statistical techniques can only be made on the basis of a combination of substantive considerations such as the research questions or hypotheses involved and statistical considerations such as measurement level and sample size.

This chapter combines developments in the emotion literature with developments in cross-cultural methodology in order to formulate four recommendations that can bridge the gap between relativist and universalist views on cultural variation in emotion. We recommend that researchers (1) specify the emotions or facets of emotions they study, preferably using a multi-componential approach to assessing emotions; (2) check the equivalence across languages and cultures of the emotion vocabulary they use, either by existing data bases or by including the measurement of meaning in their design; (3) specify the level at which they compare emotions across cultures ranging from descriptions of culture-specific constructs to direct comparisons of mean scores, and apply adequate methods to demonstrate the level of comparability claimed; and (4) account for both similarities and differences when they formulate hypotheses, as well as when they interpret their data. These recommendations are illustrated with historical and contemporary cross-cultural emotion research.

The dichotomous conceptual framework that opposes relativism to universalism plays a key role in the debates on the culture–psychology relationships (e.g., Berry, Poortinga and Pandey, 1997; see also Eckensberger, this volume). The impact of this dichotomous framework is so profound that it has led to the emergence of the two fairly autonomously functioning sub-domains of cultural psychology and cross-cultural psychology, with the former advocating a relativist and the latter advocating a universalist perspective. The problem with this twofold framework, however, is that universalism is too broad a concept. Universalism can refer to a variety of concepts ranging from approaches that only assume universality of underlying psychological traits and processes to approaches that attribute quantitative cross-cultural differences in psychological test scores to genetic differences. The broadness of the concept causes confusion in theoretical discussions. This problem was recognised by Berry et al. (1992, 2002) in their landmark book Cross-cultural psychology: research and applications. They proposed a threefold conceptual framework in which the extreme form of universalism, which interprets quantitative cross-cultural differences in psychological test scores in terms of genetic differences, was treated as a separate, third perspective. The central thesis of the present contribution is that the recognition of an additional, distinct type of universalism is needed to clarify the confusion that is caused by the broadness of the universalism concept. This thesis is based on a critical analysis of both the twofold and the threefold conceptual framework.

It is standard practice in cross-cultural psychological research to apply a psychological instrument in different cultural groups and quantitatively compare them on the scale(s) of that instrument. Such cross-cultural comparisons can be misleading for three reasons: because of the cultural specificity of the construct, the distorting effects of methodological biases, and the lack of generalizability of individual-level constructs to the cultural level.

According to cultural relativists, psychological characteristics and processes are constructed by and derive their meaning from the cultural system in which they emerge (e.g., Miller, 1997). This implies that it is meaningless to compare cultural groups quantitatively, because it would amount to comparing apples and oranges. For instance, it has been claimed that amae is not only an untranslatable Japanese emotion term but also refers to a typical Japanese emotion process (e.g., Doi, 1971). Because the process is intimately tied to the Japanese culture, it makes no sense to compare cultural groups with respect to their salience of amae.