Transcription

An Overview of the Challenges faced duringCross-Cultural ResearchMoon Halder, Jens Binder, James Stiller and Mick GregsonInstitution: Nottingham Trent UniversityEmail: [email protected] diversity in cross-cultural research is something which academic researchersneed to recognise. This paper is an overview of the challenges faced during a crosscultural research project in UK and in India. It identifies challenges which academicresearchers can face in relation to data collection, cultural obligation and peerpressure, ethical considerations and awareness, the experiences working with across-cultural team and issues faced at personal level. This paper recommends waysfor how such challenges could be addressed without compromising on the quality ofthe research. Personal experiences along with a review of literature in interculturalpsychology suggest that an understanding of cultural norms, approaches andbehaviours along with flexible and adaptable methodological and high ethicalawareness are vital. Translation, instrumentation and data collection, culturalobligation and peer pressure are some of the crucial factors discussed.KeywordsCulture, Individualist, Collectivist, Individualism, Collectivism.1

Culture defines to a large extent who we are, how we perceive our environmentaround us and how we behave (Nisbett & Miyamoto, 2005; Oyserman & Lee, 2008).Understanding behaviours requires an understanding of cultural specific norms,values and behaviours. While Western cultures (typically labelled Individualist) areconsidered more independent, self-focussed and autonomous, Eastern cultures(typically labelled Collectivist) are considered more interdependent, group orientatedand focussed on maintaining harmony (Markus & Kitayama, 1991; Triandis et al. 1988).As researchers in social sciences and related disciplines, it is important thatconsideration of such cultural differences are acknowledged when formulating andimplementing research strategies and interpreting findings. With increasedglobalisation, the need for cross-cultural research is on the rise which raises thequestion of one’s understanding of cultural variability and differences while conductingcross-cultural research.This paper is an overview of the challenges encountered during a cross-culturalresearch study conducted in a city in the UK (Western culture) and in one of the largestcities in India (Eastern culture). The UK as an individualist country and India as acollectivist country are clearly separated on the dimension of individualism andcollectivism (see Hall’s, 1976) and Hofstede’s (1980) Dimensions of Cultural Variabilityfor Selected Countries). One of the main aims of the project was to understand howculture influences our attitudes and the decisions we make in specific situations, andthe study was embedded in a larger research plan. An attempt was made in this paperto encompass some of the major challenges experienced during the conduct of thestudy in both countries. The main challenges experienced were in the areas oftranslation, instrumentation and data collection, cultural obligation, peer pressure,ethical consideration and awareness of the ethical requirements.ChallengesSurvey designThe study was in the form of a survey adapted from previous studies into culturespecific attitudes and behaviours in different social situations (Gardner et al.1999;Oyserman & Lee, 2008). The scales used in the survey were validated scales,however they had been developed keeping in mind Western populations who scorehigher on levels of individualism, unlike Eastern populations who score higher on2

levels of collectivism (Hofstede, 1980; Markus & Kitayama, 1991), and had beenmostly used on Western samples. Therefore, it was important that item equivalencewas maintained in both cultures in order to identify any “true” cultural differences. Suchequivalence in instrument design can only be achieved when researchers are mindfulof the various idioms, phrases and grammatical details to be found in a particularlocation and, more generally, how respondents make inferences in different cultures(Sekaran, 1983). For example, “Feeling guilty for my brother’s/sister’s failure” was oneof the items in the survey. Such feelings might be “strongly relevant” to respondentsin Collectivists countries, where members believe in being part of their social structure,and as such, they might hold themselves responsible for not being able to guide orsupport their brother/sister, which could have prevented their failure. However, suchfeelings might be less agreeable in Individualist countries, where members believe inbeing responsible for their own actions and behaviours. The inferences we make arehighly influenced by our cultural background, which influences how we think, perceiveand react to situations around us (Cunningham et al.1995; Dake, 1991; Kühnen &Oyserman, 2002; Oyserman & Lee, 2008). While acknowledging cultural influenceson participant responses, a culturally fit instrument and an understanding of culturaldemands can help interpret results correctly.TranslationCultural variability limits the development of a single instrument which can beconsistently used in all cultures (Sechrest et al. 1972; Sekaran, 1983). Administeringa culturally viable research instrument can only be achieved when it is used in theparticipant’s native language. This means that a translation of the research instrumentneeds to ensure the identification of such cultural variance adequately. It is importantto note that concentrating too much on maintaining methodological equivalence canlead to researchers overlooking other, important individual cultural differences.However, maintaining such equivalence helps to minimise variance in a data set moregenerally and is desirable (Sekaran, 1983).This particular study was in the format of a survey and very much text-based. Theimmediate question was that of translation. One of the primary goals of translation isto obtain an instrument that could be used consistently between cultures, i.e., withsame literal and cultural meaning (Brislin, 1970; McGorry, 2000). It was not only3

important to get at the correct “linguistic” translation, but it was also important toachieve an equivalence in cultural meaning, thereby giving the researcher confidencethat any intercultural differences in the findings are not due to translation error. Thereare different ways of conducting a translation in a systematic way, e.g., the one-waytranslation, which is without any back translation in the original language, andtranslation by committee, which involves two or more individuals who are familiar withboth languages and who help in translating the instrument from the original version.The researcher then uses the independent translators to arrive at a consensus orrecruits an additional independent party to choose the version that fits best with theoriginal version. Decentring is a method that involves re-designing survey instrumentsto fit in with the target culture and involves constant revision of the original surveyinstrument. This method can also alter the items and survey length (McGorry, 2000).Using bilinguals, who can read and write fluently in both languages, is crucial during atranslation process (Marin & Marin, 1991). However, cultural phraseology should alsobe considered, where regional or class differences matter. So, it is first important todecide which groups of culturally different people are under investigation.Being able to develop a culturally and literally viable instrument was crucial for theresearch design and for maintaining validity and reliability of the data. There were twoindependent groups of translators, who were fluent in reading and writing in bothEnglish and Bengali (the native language most relevant among individuals living in theIndian city). Each group consisted of three members who were recruited in order toconduct the back translation process adequately (see Figure 1). This was in line withstandard recommendations (Marin & Marin, 1991) and was also cost effective andquick, unlike, for example, the decentring process of translation (McGorry, 2000) whichwould require a significant and time consuming departure from the original version ofthe survey instrument. It was also different from the translation by a committee method,where the three members in each group translated the survey items after a groupdiscussion and not as independent translators.The translation method selected for this particular study had the advantage of quicklyremoving any discrepancies in the translated version as survey items were translatedafter coming to a group consensus. Although it is recommended that applications ofseveral translation processes could help to achieve a more accurate and culturally fitinstrument (McGorry, 2000), such a procedure is not always possible for researchers4

due to restrictions in funding and time constraints. Below is the step by step guide ofthe back translation process used.Back Translation Procedure for research instrument in India1)A focus group of three Bengali translators were selected who had studied andlived in the Indian city. They could read, write and speak fluently in Bengali and inEnglish.2)The focus group was asked to read through the questionnaire which was in theoriginal English version and then after discussion with each other came up with aBengali version of each item on the survey.3)Another focus group of three Bengali translators were selected for the backtranslation procedure. They could also read, write and speak fluently in Bengali andEnglish.4)The second focus group was asked to read every item in the Bengaliquestionnaire and come up with an English version of each item on the questionnaire.This was also achieved with discussion between the three members in the focus group.5)The original English version of the questionnaire was then compared with thesecond English version of the questionnaire by all parties including the researcher.Although both the original English version and the second English version of thequestionnaire were not exactly the same, the core meaning of each survey item wasfound to be well preserved. Hence there were no further changes made to thetranslated version of the questionnaire for the study. On completion of the translationprocess the survey were administered to participants in India.5

Figure 1. An Illustration of the Back Translation Process followed in IndiaData collectionThe level of familiarity with the general research process and participation in researchstudies among the Indian sample was certainly a concern as not all universities inIndia indulge in similar research activities. This called for developing creative ways toadminister the surveys in both cultures so that all participants could fully understandthe participation process and their rights. All the respondents were from HigherEducation institutions and had good levels of English both written and verbal, but itwas difficult to find out what their actual level of understanding was as they had notthemselves been involved in similar research projects previously. Additionally, it wasdifficult to have a one-to-one chat with the student participants due to time constraints.It was observed that in spite of their familiarity with English language, they still haddifficulty in following the overall study participation process. This could be due to theirlack of familiarity with the research process and the ethical demands or due to otheraspects of the survey design. This further highlights the importance of culturally fit6

instruments to help attain valid participant responses. Here, this issue was resolvedas the researcher was present in person along with one of the bilingual translators toassist participants while taking the survey. However, in other cases such as onlinesurveys, participants’ understanding of the survey questions might be restricted andparticipants might end up responding incorrectly to the questions, when they fail tounderstand task requirements, further highlighting the importance of clear andappropriate instructions. Unintended responses will certainly have an adverse impacton the research outcome and may produce an intercultural effect when actually thereis none. The global demand and use of the internet has made researchers changeand adapt to newer ways of conducting their research and in particular for crosscultural research online methods are attractive (Kraut et al. 2004) as they save timeand are also cost effective. However, the absence of personal cues and support mightalso have a negative impact on the quality of the data collected.Conducting cross-cultural research also requires planning ahead. While particulardates and times might be useful and convenient for the researcher in one country, itmight not be the same in another country. The data collection process in the UK wasconducted without any hindrances, whereas a different picture emerged in India asnational holidays were suddenly called for by the government due to local elections,which also resulted in social unrest in some parts of the country. This had an adverseeffect on the sample size as it resulted in a reduced number of student attendance.Therefore, it is suggested that although planning ahead is always useful, researchersshould always plan for sudden changes during cross-cultural research as it is notalways possible to predict cultural or political changes in other countries. Planningahead can include aspects like allocating additional meeting times, checking universityopening and closing times, and also identifying the most promising time for datacollection.Informal meetings with Indian students after their participation in the surveysuggested that although they were anxious about the social situation in the city, theyfelt that as students, they were obligated to participate in the survey as it has beenrequested by their lecturers. Such