In this section, the concept of two scale development approaches involving foreign measures and four translation methods are introduced.
Two scale development approaches involving foreign measures
In the four types of scale approaches in the Chinese management research proposed by Farh, Cannella, and Lee (2006), the de-contextualization approach and the contextualization approach are applied when developing a new scale (Farh et al., 2006, 307-310). However, while the foreign measure which is available for the target construct exists, the translation approach and the adaptation approach are applied and they are briefly discussed below.
1. The translation approach
It’s referred to a process in which a foreign existing scale is translated literally into a Chinese version (also called a target language scale). The translation approach is grounded on two major assumptions: (a) the meaning of the target construct is equivalent across cultures, and (b) a high quality unbiased source language scale is available for the target construct (Farh et al., 2006. p.303).
Although this approach doesn’t take lots of time and costs and it’s convenient for the researcher to compare the research results across cultures, the limitations of this approach do exist. One of the major limitations is that it’s difficult to achieve semantic equivalence by translating the source language scale word by word due to cultural distance between the West and Chinese. Another limitation is that some unknown cultural biases embedded in the Western scales are likely to keep hiding in the Chinese version through literal translation. To lessen the limitations of the translation approach, the adaptation approach is proposed.
2. The adaptation approach
In recent years, the term test translation has been frequently replaced with test adaptation because adaptations in contents and wordings in reference to culture are needed (Geisinger, 1994; Hambleton & Patsula, 1998). Adapting a scale has to do with item appropriateness. Altering the wording of items, dropping inappropriate item, or even adding new items to the scale is acceptable from the cultural perspective (Farh et al., 2006). According to Hambleton and Patsula (1998), test adaptation involves more activities such as:
(1) deciding whether or not a test can measure the same construct in a different language and culture, (2) selecting translators, (3) deciding on appropriate accommodations to be made in preparing a test use in a second language, and (4) adapting the test and checking its equivalence in the adapted form. (p. 155) Geisinger (1994) proposed ten steps for adapting a measure and he mentioned that not all steps would be required during a process of the measure adaptation.
Hambleton and Patsula (1998) suggested that steps and guideline of cross-cultural adaptation of the measure should include at least four steps: (a) translating the test from source to target language (forward translation); (b) translating the translated version back to the source language (back translation); (c) resolving the discrepancies between the original version and the adapted version; and examining cross-cultural equivalence; and (d) Pre-testing the original and adapted versions to check for equivalence. The comparison of two types of scale development approaches is shown in Table 2.4.
Table 2.4. Two Types of Scale Development Approaches in Chinese Management Research: Key Assumption, Major Strengths and Limitations Scale
Development Approaches
Key Assumptions Major Strengths Major Limitations
Translation approach
Target construct is equivalent across cultures in terms of overall
definition, content domain and empirical representations of the content domain
Availability of high quality culturally unbiased Western scales for target construct
Low developmental time and costs
Preserve the possibility of a high level of equivalence
Allow for direct cross-cultural comparison of research findings
Difficulty in achieving semantic equivalence between the Chinese and Western scales
Culturally unbiased Western scales are hard to come by
Adaptation approach
Target construct is equivalent between cultures in terms of overall definition and content domain
Availability of high quality Western scales for target construct
Low to moderate developmental time and costs
Ease of scholarly exchanges of research findings with the Western literature
Difficulty in conducting cross-cultural research
Drastic adaptation may create new scale that requires extensive validation in the Chinese context Source: Farh et al., 2006, p.304
Translation methods
Most measures used in Chinese contexts are borrowed form the foreign measures.
Since language is clearly regarded as one component of culture, and our thinking is influenced by the words and phrases existing in our language (Hofstede, 2001). Great care must be taken during the process of translating the measures developed in different languages. Brislin (1980) offered four translation methods which are elaborated in the following:
1. Back translation
It’s referred to a process in which the material in one language is forward translated into the target language and then back translated to the source language to examine the discrepancy between the original and back-translated one. If the two source language forms are not the same, follow-up modifications are made to enhance the translation equivalence. This technique is frequently mentioned by many researchers in the literature (Farh et al., 2006; Geisinger, 1994; Hofstede, 2001;
Hambleton & Patsula, 1998; Van de Vijver & Hambleton, 1996)
There are at least two bilingual translators needed during the process of back translation. Hofstede (2001) and Geisinger (1994) further noted that translators should meet a set of requirements. They have to be (a) fluent in both languages; (b) familiar with both cultures; and (c) knowledgeable on the characteristics and content of the instrument.
The advantages of back translation lie in that even if the researcher doesn’t understand the target language, he or she can still judge about the quality of the translation (Brislin, 1980). In addition, it avoids translation errors and prevents the translators from filtering the meanings according to their value systems in their countries (Hofstede, 2001). However, one of the criticisms of back translation is that once the translators realize their work was going to undergo back translation, they would use wordings which were correctly translated into original wordings instead of a translation in reference to culture differences (Hambelton, 1993, cited in Geisinger, 1994; Van de Vijver & Hambleton, 1996). The possible solution is to invite a whole
Alternatively, one group take the test in the source language, and the other group take the same test in the target language. After running a statistical analysis, items which yield discrepant responses or different frequency of responses can be easily found by the researcher. The criteria of selecting samples of respondent should include not only their language proficiency but also demographic variable (e.g., gender, age, occupation) which affect the meaning of answers (Hofstede, 2001).
The advantage of this technique is sophisticated statistical methods and concepts which assist in judging the quality of the translation. The disadvantage is that the test, or the research instrument, is developed on the basis of the bilinguals’ responses which may not represent the population of interest.
3. The committee approach
A group of bilinguals works together to translate the instrument from the source to the target language. Because they are not independent workers, any mistake made by one member can be found by others on the committee. However, it’s possible that the committee members may be unwilling to criticize one another.
The committee approach is suggested to be combined with back translation (Geisinger, 1994; Van de Vijver & Hambleton, 1996). A group of individual who meet the same requirements as the translators reviews the translated or adapted version of the instrument.
4. Pretest procedures
Once a translation is completed, it’s suggested that it should be field tested to ensure that the translated version can capture the same construct underlying the original instrument. Thus, a small sample of individuals similar to the target population should be identified (Geisinger, 1994).
A combination of the above-mentioned translation methods should depend on the needs of the research project (Brislin, 1980). The disadvantages of one translation method are offset by the advantages of another.
Summary
In translation approach and adaptation approach, the term adaptation has been frequently used in the literature because adaptations in contents and wordings in reference to culture are needed instead of translating the instrument literally.
Four translation methods proposed by Brislin (1980) include back translation, the bilingual technique, and the committee approach, and pretest procedures. These four
translation methods can be combined or used individually. However, the quality of the translation lies in a careful selection of the translators (Geisinger, 1994; Hofstede, 2001)
CHPATER III. METHODOLOGY
This chapter outlines research methods as well as strategies for data analysis.