Some Basic Distinctions 96 THE NATURE OF EQUIVALENCE 97 Linguistic Equivalence 98 Conceptual Equivalence 98 Functional Equivalence 99 Metric Equivalence 99 Other Forms of Equivalence 100 THE NATURE OF BIAS 100 Construct Bias 101 Method Bias 101 Item Bias 102 THE TRANSLATION AND ADAPTATION OF TESTS 103 The Role of Language 103
Approaches to Test Adaptation and Translation 106 Rules for Adapting Test and Assessment Materials 108 Steps in the Translation and Adaptation Process 108 METHODS OF EVALUATING TEST EQUIVALENCE 109 Methods to Establish Equivalence of Scores 109 Methods to Establish Linkage of Scores 111 CONCLUSION 112 APPENDIX 113
Guidelines of the International Test Commission for Adapting Tests 113 REFERENCES 114
Some say that the world is shrinking. We know that worldwide cable news programs and networks, the Internet, and satellites are making communications across cultures and around the world much easier, less expensive, and remarkably faster. Cross-cultural psychology studies the psychological differences associated with cultural differences. (In a strictly experimental design sense, in much of cross-cultural research, cultures typically serve as independent variables and behaviors of interest as dependent variables.) At one time, such research implied crossing the borders of countries, and generally, it still may. However, as countries become more multicultural due to immigration (made much easier in Europe with the advent of the European Union, or EU), different cultures may exist within a country as well as in differing countries. Research in psychology, too, has been affected by these worldwide changes.
The world context has also shifted. The cultural makeup of the United States is certainly changing rapidly; recent U.S. Census Bureau analyses indicate the rapid increase in ethnic minorities, especially Hispanic Americans and Asian Americans, to the extent that the historical European American majority in American is likely to become a minority group within the next decade or so (see Geisinger, 2002, or Sandoval, 1998, for an elaboration of these data). While the United States is experiencing population changes with a considerable increase in groups traditionally identified as ethnic minorities, so have other parts of the world experienced these same population shifts. Many of these changes are occurring as "the direct consequence of cross-immigration and globalization of the economy" (Allen & Walsh, 2000, p. 63). Cultures change due to population changes caused by immigration and emigration, birth and death rates, and other factors, but they also change due to historical influences apart from population shifts. Countries that have suffered famine, aggression on the part of other nations, or other traumatic changes may experience significant cultural transformations as well as population changes. Most psychologists have studied human behavior within a single, broad culture, sometimes called Euro-American culture (Moreland, 1996; Padilla & Medina, 1996). In more recent years, psychologists have begun to recognize the importance of culture; in 1995 the American Psychological Association (APA) began publishing the journal Culture and Psychology. Such an event may be seen as an indication of the increased recognition of the importance of culture in psychology.
According to Berry (1980), cross-cultural psychology seeks to explore the relationships between cultural and behavioral variables. He included ecological and societal factors within the realm of cultural variables. Likewise, he included as part of behavioral variables those that must be inferred from behavior, such as personality, attitudes, interests, and so on. The variables that are studied in cross-cultural psychology, of course, must be measured and they have traditionally been examined using standardized tests, interviews, and a variety of formal and informal assessments. In fact, Triandis, Malpass, and Davidson (1971) describe cross-cultural psychology as follows: "Cross-cultural psychology includes studies of subjects from two or more cultures, using equivalent methods of measurement, to determine the limits within which general psychological theories do hold, and the kinds of modifications of these theories that are needed to make them universal" (p. 1; also cited in Berry, 1980, p. 4). The previous statement emphasizes the need for equivalent methods of measurement. If the findings of cross-cultural psychology are to have validity, then equivalent measurement is required. This point is the general theme of this chapter. It is argued that to note cross-cultural differences or similarities in terms of psychological variables and theories, one must have confidence that the measures used in research are equivalent measures in each culture.
When one is comparing two cultures with respect to a psychological or other variable, there are a number of factors that can invalidate the comparison. For example, if one selects well-educated individuals from one culture and less well-educated persons from the second culture, the comparison is likely to be flawed. (See van de Vijver & Leung, 1997, for a more complete listing and explanation of these confounding factors.) However, one of the primary sources of invalidity, one that is often not understood as easily as the previous sampling example, relates to measurement instruments. When the two cultures to be compared do not employ the same language to communicate and have other cultural differences, the measures that are used in the comparison must of necessity be somewhat different. A number of the possible options are discussed and evaluated in this chapter. Depending upon the use and other factors, cross-cultural psychologists and psychologists dealing with testing issues in applied use are provided with some strategies for solving the dilemmas that they face. Language is generally not the only disparity when making cross-cultural comparisons. Cultural differences in idioms, personal styles, experiences in test taking, and a plethora of other variables must also be considered in making cross-cultural or multicultural assessments. These factors are often more subtle than language differences.
Thus, there are theoretical reasons to examine testing within cross-cultural psychology. There are also applied reasons that testing is important in cross-cultural settings. Obviously, the applied uses of tests and assessment across cultures must rely on the theoretical findings from cross-cultural psychology. Tests and assessment devices have been found to have substantial validity in aiding in various kinds of decision making in some cultures. Thus, other cultures may wish to employ these measures, or adaptations of them, in what appear to be similar applied settings in cultures other than those where they were first developed and used.
Test development, test use, and other psychometric issues have long held an important role in cross-cultural psychology. Berry (1980) differentiated cross-cultural psychology from many other areas of psychology, and aligned it closely to measurement and methodology by reflecting that "most areas of psychological enquiry are defined by their content; however, cross-cultural psychology is defined primarily by its method" (p. 1; emphasis in the original). The words testing and assessment are used interchangeably by some psychologists, but differentially by others. When they are distinguished, testing involves the administration and scoring of a measurement instrument; assessment, on the other hand, is a broader term that includes score and behavioral interpretation in the context of the culture or the individual being evaluated.
Before beginning a formal discussion of testing in cross-cultural psychology, a few fundamental features must be differentiated. When one uses measures in two cultures, one engages in cross-cultural work. On the other hand, when one studies the various subcultures within a country, such as the United States of America, then one performs multicultural analyses. The distinction is somewhat more complex, however, and this demarcation in described in the next section. Next, one of the fundamental distinctions in cross-cultural psychology—the concepts of etic and emic—are described; these terms in some ways parallel the distinction between cross-cultural and multicultural analyses. In brief, etic studies compare a variable across cultures whereas emic studies are performed within a single culture. Finally, a distinction between the uses of tests in relatively pure research as opposed to the testing of individuals in real-life, often high-stakes decisions is described.
The differences between cross-cultural and multicultural are not entirely clear. Allen and Walsh (2000), for example, make the distinction that use of tests across cultural groups ("often internationally") is a cross-cultural application of tests whereas the use of tests with individuals of differing ethnic minority (or perhaps cultural) status within a nation is a multicultural application of tests (p. 64). They note that there is often overlap between culture and minority group status. Clearly, the distinction blurs.
Thus, the techniques used in cross-cultural psychology have great applicability to multicultural, psychological issues. The questions and concerns involved in adapting a test to make a comparison between U.S. and Mexican cultures, for example, are certainly applicable to the testing of Spanish-speaking Chicanos in the United States.
Linguists often use two important words in their work: phonetic and phonemic. According to Domino (2000), "Phonetic refers to the universal rules for all languages, while phonemic refers to the sounds of a particular language. From these terms, the word 'etic' and 'emic' were derived and used in the cross-cultural literature. Etic studies compare the same variable across cultures Emic studies focus on only one culture and do not attempt to compare cultures" (p. 296). These terms were originally coined by Pike (1967). The words etic and emic also refer to local and universal qualities, respectively. Thus, the terms have been used to describe both behaviors and investigations.
Emics are behaviors that apply only in a single society or culture and etics are those that are seen as universal, or without the restrictions of culture. A complaint made about traditional psychology is that it has presumed that certain findings in the field are etics, even though they have not been investigated in non-Western arenas (Berry, 1980; Moreland, 1996). Thus, some findings considered etics are only so-calledpseudo etics (Triandis et al., 1971). The emic-etic distinction is one that has broad applicability to the adaptation of tests developed in America to other cultural spheres.
The emic-etic distinction also applies to the goals of and approaches to cross-cultural research: The first goal is to document valid principles that describe behavior in any one culture by using constructs that the people themselves conceive as meaningful and important; this is an emic analysis. The second goal of cross-cultural research is to make generalizations across cultures that take into account all human behavior. The goal, then is theory building; that would be an etic analysis (Brislin, 1980, p. 391). In searching for emic findings, we are attempting to establish behavioral systems (or rules) that appear to hold across cultures. That is, we are endeavoring to verify that certain behavioral patterns exist universally. Etic studies look at the importance of a given behavior within a specific culture.
The Use of Tests and Assessments for Research and Applied Use
The goal of most cross-cultural psychologists and other researchers is the development of knowledge and the correlated development, expansion, and evaluation of theories of human behavior. Many applied psychologists, however, are more concerned with the use of tests with specific individuals, whether in clinical practice, school settings, industrial applications, or other environments in which tests are effectively used. The difference in the use of tests in these settings is significant; differences in the type or nature of the tests that they need for their work, however, may well be trivial. If we assume that the psychological variable or construct to be measured is the same, then differences required for such varied uses are likely to be minor. Both uses of tests, whether for research or application, require that the measure be assessed accurately and validly. Part of validity, it is argued, is that the measure is free from bias, including those biases that emerge from cultural and language differences. Some writers (e.g., Padilla & Medina, 1996) have accentuated the need for valid and fair assessments when the nature of the assessment is for high-stakes purposes such as admissions in higher education, placement into special education, employment, licensure, or psychodiagnosis.
Was this article helpful?