Facets and Dimensions of Cross-Cultural Adaptation - Ashridge

Facets and dimensions ofcross-cultural adaptation:

refining the toolsArno Haslberger

PEF Private University for Management, Vienna, Austria

Abstract

Purpose – The management literature on cross-cultural adaptation has used a conceptualisation andmeasurement approach developed by Black and Stephens. Their work has led to significantdevelopment in the field. Now it is time to move beyond and use a more refined tool. This paperproposes such a tool and compares its characteristics with those of the older instrument.

Design/methodology/approach – The paper is based on a sample of 204 expatriates, who weresurveyed using the older and the proposed instrument. It uses confirmatory factor analysis to comparethe two instruments. Independent variables include cultural difference and language skills.

Findings – This study demonstrates that a more refined measurement of adaptation outcomes,which distinguishes cognitive and affective factors and four non-work facets, is superior tomeasurement based on the older instrument. The new scale also provides a case for the improvementof cross-cultural adaptation theory.

Research limitations/implications – The paper is based on a cross-sectional sample and samplesize is relatively small for confirmatory factor analysis. Additional research is necessary to corroboratethe evidence presented here about the superiority of the proposed measure. The paper providesresearchers with a new tool for use in cross-cultural adaptation studies.

Originality/value – The paper describes a new, empirically developed measurement tool forcross-cultural adaptation.

Keywords Expatriates, Cross-cultural management, Management effectiveness, Measurement

Paper type Research paper

For over a decade, many cross-cultural adaptation studies (e.g. Black, 1990a, b, Blackand Gregersen, 1991a, Black, 1994, Taylor and Napier, 1996a, Robie and Ryan, 1996,Aycan, 1997, Kraimer et al., 2001, Selmer, 2001, 2002, Takeuchi et al., 2002) in themanagement literature have looked at a three-fold split initially proposed by Black(1988): general adjustment, interaction adjustment, and work adjustment. Asubsequent measure developed by Black and Stephens (1989) has fairly consistentlyreproduced these three facets in these studies.

The literature in the wake of the Black and Stephens questionnaire has contributedgreatly to our understanding of cross-cultural adaptation of expatriates. Yet, there are afew problems associated with the measure: it was not developed in a systematic,theory-driven way; it measures adaptation one-dimensionally on an adjusted- unadjustedrange; the three facets of adjustment may be an artefact of the unsystematically chosenitems going into the original measure.

The Emerald Research Register for this journal is available at The current issue and full text archive of this journal is available at

www.emeraldinsight.com/researchregister www.emeraldinsight.com/0048-3486.htm

This paper makes reference to some data analyses without including detailed statistics. Pleasecontact the author, if you are interested in the details.

Dimensions ofcross-cultural

adaptation

85

Personnel ReviewVol. 34 No. 1, 2005

pp. 85-109q Emerald Group Publishing Limited

0048-3486DOI 10.1108/00483480510571897

Ashridge Business School UK - http://www.ashridge.org.uk

The purpose of this article is to deal with the second shortcoming. First, it willreiterate the principle that cross-cultural adaptation is a multi-dimensionalphenomenon. Second, it will suggest a way to measure at least two separatedimensions of adaptation. Third, it will demonstrate that there are significantdifferences in the association of independent variables with two different dimensions ofcross-cultural adaptation on a sample of 204 expatriates, thus showing the superiorityof multi-dimensional over one-dimensional measurement of adaptation outcomes. Thearticle will close with a discussion of the practical relevance of these results for themanagement of expatriate assignments.

Multiple dimensions of cross-cultural adaptationCross-cultural adaptation is a complex process in which a person becomes capable offunctioning effectively in a culture other than the one he or she was originally socialisedin. As people become immersed in foreign cultures their whole being gets affected. Oneway to analyse this is to look at the whole person only, and this is what managementresearchers in the wake of the Black and Stephens (1989) instrument have done. Anotherway is to disaggregate the holistic view into its constituent parts to gain a betterunderstanding. This has a long-standing tradition going back to Plato (1989, p. 436a)who asked whether we “learn with one part of ourselves, feel anger with another, andwith yet a third desire the pleasures of nutrition and generation and their kind, orwhether it is with the entire soul that we function in each case when we once begin” anddecided in favour of the former. Modern neuroscience also distinguishes a “tripartiteamalgam that includes cognition, emotion, and motivation” (LeDoux, 2002, p. 174). In thecross-cultural adaptation literature several authors have used differentiated models.

Grove and Torbiorn (1985) developed a model that integrates behaviour andcognition. Gudykunst and Hammer (1988) focus on anxiety, an emotion, anduncertainty, an element of cognition. They theorised that different variables impact onuncertainty and anxiety, and that some are associated with both. Kim (1988) has foryears stressed the need to integrate behaviour, cognition and emotion. Roth’s (1997)neuroscience model of consciousness supports the split into behaviour, cognition andemotion.

Cross-cultural adaptation is measurable from an external point of view bybehavioural observation either by members of the local culture or by researchers. Thistype of measurement is resource intensive.

Alternatively, there is the “internal” measurement of the person’s cognitive andaffective states. The concept of uncertainty avoidance (Berger and Calabrese, 1975;Berger, 1979, 1987) lends itself to studying the cognitive side of adaptation. Gudykunstand Hammer (1988) and others (Grove and Torbiorn, 1985; Gudykunst and Nishida,1986; Gudykunst, 1988; Berger and Gudykunst, 1991; Black and Gregersen, 1991b;Black et al., 1992a) regard uncertainty about predictions of others’ behaviours and thedriving forces behind those behaviours as well as about the likely outcomes of one’sown behaviours as central to adaptation. Low uncertainty or, alternatively, highconfidence is an indicator for a state of high adaptation.

Emotions are difficult to measure. LeDoux (2002) emphasises that directmeasurement of emotions in humans encounters ethical and practical barriers.Rather than feelings researchers measure behaviour or verbal reports of emotions.Cross-cultural adaptation studies have to rely on summary reports of emotions[1].

PR34,1

86


Gudykunst and Hammer (1988) have combined uncertainty reduction with an affectivedimension. They focus on the negative emotion of anxiety. Other research into theemotional aspects cross-cultural moves regards satisfaction, a positive emotion, indifferent areas of life as part of adaptation (Naumann, 1993a, b; Bhuian and Islam1996). Rarely do adaptation studies include the whole range of positive and negativeemotions[2]. The literature mentions four basic emotions rooted in human biology:anxiety, satisfaction, anger, and depression (Kemper, 1987; Mattenklott, 1996), i.e. onepositive and three negative. Roth (1997, p. 209) states that the limbic system, whichplays a major role in the experience of emotions, evaluates based on the basic criteria of“pleasure”, meaning a range from positive to negative.

This study has used the distinction between cognitions and emotions to exploreadaptation outcomes in a cross-sectional sample of expatriates. In this it goes beyondresearch done with the Black and Stephens (1989) scale, which combines cognitionsand emotions into an adjusted-unadjusted range. In order to establish the usefulness ofa separate measurement of cognitions and emotions, the following independentconstructs will be used: novelty, discretion, self-efficacy, and social networks. Thechoice of independent constructs is guided by prior research. Variousoperationalisations, particularly of novelty, have been researched extensively in thepast. Therefore, they are useful for the demonstration of the advantages ofmulti-dimensional measurement of adaptation outcomes.

NoveltyCultural distance and previous expatriate experience are two aspects of the construct ofnovelty that will be discussed here.

Cultural distanceIn the cross-cultural adaptation literature cultural novelty has played an important rolefor a long time (e.g. Simmel, 1987; Schuetz, 1944; Nash, 1967; David, 1972; Taft, 1977;Triandis, 1980; Church, 1982; Grove and Torbiorn, 1985; Earley, 1987; Kim andGudykunst, 1988; Kim, 1988; Gregersen and Black, 1990; Black et al., 1991; Taylor andNapier, 1996a; Aryee et al., 1996). Hofstede’s research on differences in values hasinfluenced the field for a quarter of a century (Hofstede, 1980, 1984, 1992; Hofstede andBond, 1984, 1988). In its wake other researchers have contributed their own versions ofcultural difference measures (Trompenaars, 1993; Javidan and House, 2001).

Research has indicated a negative influence of cultural novelty on adaptationoutcomes (Black and Stephens, 1989; Black and Gregersen, 1991c). However thisrelationship is not always consistent. In a repatriation study, Black and Gregersen(1991a, p. 689) found no association between the novelty of the culture in the foreignlocation and any of three facets of adaptation for expatriate employees or their spouses:

This may suggest that “net culture novelty” or the difference between the absolute novelty ofone culture relative to another and a given individual’s familiarity with the target cultureshould be included as a variable in future expatriation and repatriation adjustment studies.

These differing results for expatriates and repatriates are interesting. Yet, themeasurement of cultural novelty in these studies is problem-laden. For example, in therevised version of an existing instrument, respondents “were asked to indicate on [a]5-point Likert scale how similar or different the following were compared to America


adaptation

87


. . . 1. Everyday customs that must be followed. 2. General living conditions. 3. Usinghealth care facilities. 4. Transportation systems used in the country. 5. General livingcosts. 6. Available quality and types of food. 7. Climate. 8. General housing conditions”(Black and Stephens, 1989, p. 542). A comparison with Black and Stephens’ (1989)adaptation scale, which is described below, shows that the cultural novelty measureand the adjustment measure match closely. Under such circumstances, culture noveltyis bound to correlate significantly with adaptation outcomes. Black and Stephens’(1989) results, although in line with theory, have to be read with caution.

The most important and most obvious aspect of novelty in cross-cultural adaptationis the distance between home and host cultures. It is the essence of content novelty.

Over the years, a number of conceptualisations of culture and cultural distance havebeen introduced. A widely used conceptualisation of cultural distance stems fromDutch researcher Geert Hofstede, who studied IBM employees in over 50 differentcountries in the 1970s and 1980s (Hofstede, 1980, 1984, 1992; Hofstede and Bond, 1984,1988). Hofstede distinguished originally four dimensions of national culture: powerdistance, individualism, masculinity, and uncertainty avoidance. In co-operation withresearchers from the People’s Republic of China, he and a colleague found a fifthdimension: long-term orientation (The Chinese Culture Connection, 1987).

Hofstede published scores on each of the first four dimensions for most of thecountries covered in this study.

Previous expatriate experienceBesides cultural distance, previous expatriate experience relates to the noveltyexperienced by expatriates.

A first time expatriate is uninformed about his or her reactions to and ability to livein another country. Depending on the length of stay, the expatriate will or will not beable to reach a relatively stable state in the foreign location. After a few months mostpeople will have learned enough about their own reactions to cross-cultural moves tomake a subsequent move significantly less novel as they can anticipate their reactionsand feelings. For previous experience to become an adaptation facilitating force,although, it must have been a positive one.

Theoretical discussions have, for a long time, included previous expatriate experienceas relevant for the adaptation process (Nash, 1967; Church, 1982; Black and Stephens,1989; Black and Gregersen, 1991c; Nicholson and Imaizumi, 1993; Brewster and Pickard,1994; Taylor and Napier, 1996a; Aycan, 1997). Yet empirical evidence, particularly whenfocused on the length of exposure, has been weak. Church (1982, p. 549) suggested that:“the nature and quality – for example, the depth, intimacy . . . , and similarity . . . – ofthe previous cultural experience or host culture exposure may be more important”. Theline of research focusing on the distinction of work, general, and interaction adjustmentsupports the relevance of previous expatriation experience on adaptation outcomes onlyin some instances (Black, 1988). Most of the time, it does not. Black and Stephens (1989)as well as Black and Gregersen (1991c) found no significant relationship betweenspouse’s general and interaction adaptation and previous experience. Black andGregersen (1991b) found no significant relationship for either of the three facets ofadaptation. Taylor and Napier (1996a, p. 68) reported that “previous experience abroaddid not strongly affect work adjustment of the foreign women professionals”. Nicholsonand Imaizumi (1993) even found an inverse relationship between previous overseas

PR34,1

88


experience and the adaptation of Japanese expatriates in Britain. They suggest that “lackof experience may actually be beneficial, presumably because it is more likely to conveyto the assignment the benefits of being a completely new and exciting experience”(Nicholson and Imaizumi, 1993, p. 128).

Theoretically and intuitively, it makes sense to include previous overseas experiencein cross-cultural adaptation models. As Church (1982) pointed out, with Black andStephens (1989) and Black and Gregersen (1991b, c) concurring, the problem may lie inthe simple quantitative measurement of the variable rather than its lack of relevance.

Cultural novelty influences both cognitive uncertainty and emotional well-being ofexpatriates. Prior experience is a moderating factor that helps people cope better with theemotional strain of cross-cultural adaptation. This leads to the following hypotheses:

H1a. Cultural distance is negatively related to both cognitive and emotionaladaptation.

H1b. Positive prior expatriate experience is positively related to emotional adaptationand unrelated to cognitive adaptation.

DiscretionNovelty makes adaptation harder. Discretion granted to the individual by the hostenvironment, allows him or her to influence surroundings and to offload some of theburden of adaptation[3]. Discretion comes from two sources: the first is the level oftolerance within the host society (Berry et al., 1988), and the second is thesocio-economic status the stranger enjoys in the new location.

The level of tolerance within society refers to two facets (Kim, 1988): one is thepressure to conform, in Kim’s (1988, p. 66) words: “the extent to which the environmentchallenges strangers to adopt the normative patterns of the host culture and itscommunication system” or “the degree to which the environment overtly or covertlyexpects or demands that strangers follow its normative cultural and communicationpatterns” (Kim, 1988, p. 128). In other words, conformity pressure is the deviationallowed from the average culture such as values, norms, behaviours, etc. It may also beexpressed as the homogeneity or heterogeneity of lifestyles in a particular socialenvironment. Major cities are more complex and, therefore, possess more diversity andshow a greater number and more heterogeneous life styles than rural areas.

The other facet is the host society’s receptivity for strangers. Kim (1988, p. 66)defines it as “the opportunities offered to strangers to participate in on-going socialactivities” or “the degree to which a given host environment shows openness andacceptance toward strangers” (Kim, 1988, p. 128).

This study included the first facet of tolerance only. The host society’s receptivityfor strangers and the respondents’ socio-economic status were not observed. Sincediscretion lowers the impact of novelty by allowing strangers to only loosely adhere tolocal norms, the following hypotheses are advanced:

H2a. Freedom to adhere or not adhere to local rules is positively related to cognitiveand emotional adaptation.

H2b. The size of the city of residence is positively related to cognitive and emotionaladaptation.


adaptation

89


Self-efficacyA person moving internationally arrives in the new location with a set of beliefs andattitudes that will determine his or her resistance to the temptation to give up andwithdraw that develops when problems arise. The conviction of the person that he orshe can overcome the obstacles encountered is his or her self-efficacy.

Black and Mendenhall’s (1991) conceptualisation of the cross-cultural adaptationprocess based on social learning theory describes self-efficacy as an important variableinfluencing the person’s motivation to retain and to reproduce a learned behaviour(also Black et al., 1992b):

Theoretically, one of the most powerful individual variables that determines persistence ineffort is the individual’s belief in control or self-efficacy (Black et al., 1992a, p. 749).

A person with high self-efficacy will be less likely to give up when he or she encountersadaptation problems. Therefore, withdrawal will become less likely and positiveadaptation outcomes will become more likely.

This study included one aspect related to self-efficacy that is particularly relevantbut often neglected in cross-cultural adaptation studies: language skills. In a rare studyincluding language variables on work adaptation of expatriate women in Japan, Taylorand Napier (1996b, p. 78, emphasis added): “found language skills and age to be themost important personal attributes for successful adjustment”. Language skills willdetermine the person’s ability to interact with host nationals and learn frominteractions, refining his or her cognitive maps and behaviours. The accurateassessment of skill levels requires the use of standardised tests such as the Test ofEnglish as a Foreign Language (TOEFL). Alternatively, self-ratings will provide someindication of language ability. For the measurement of self-efficacy the perceivedsufficiency of language skills independent of their absolute level is also important.

Language skills are key for an adapted life, which leads to the following hypotheses:

H3a. Language skills are positively related to cognitive and emotional adaptation.

H3b. The perception of possessing sufficient language skills is positively related tocognitive and emotional adaptation.

Social networksAn expatriate does not exist in a social vacuum. Many expatriates move with theirspouses and children. They establish contacts with neighbours, eventually developinga social network of acquaintances and friends. In the case of expatriate employees,there is the company support system or lack thereof. Taken together, these form thesocial networks that influence adaptation outcomes directly as well as withdrawaltendencies.

Like many of the other variables discussed here social support or social networkshave long featured in cross-cultural research (Nash, 1967; David, 1972; Berry et al.,1988; Black and Gregersen, 1991c; Aycan, 1997). As Church’s (1982) review articlepoints out, results sometimes differ according to how networks or support aremeasured.

Oberg’s (1960) conceptualisation of culture shock included the sense of losing one’sfriends back home. The replacement of social networks is important in cross-cultural

PR34,1

90


adaptation. Expatriates who are able to develop friendly relationships with localnationals will be more adapted than those who do not:

H4a.The friendly character of contacts with the local population is positively related tocognitive and emotional adaptation

Measurement of variablesDependent variablesRefined outcome measure. Adaptation outcomes relate to behaviour, cognition andemotion. In addition, they fall into two broad spheres: societal processes, which includeday-to-day interactions and the establishment and maintenance of relationships; andsocietal structures, comprised of general conditions, institutions, values and norms.The self-report questionnaire, therefore, included 12 items related to processes andstructures which were measured with questions regarding cognition and emotion[4,5].

Behaviours were not included because self-reports on behaviours are notoriouslyinaccurate:

(1) Interactions (A):. Reprimanding a local person of lower status than you – telling off someone

for something that they have done wrong (based on Furnham and Bochner,1982).

. Dealing with a local person who is upset/cross and aggressive/abusive(based on Furnham and Bochner, 1982).

(2) Relationships (R):. Establishing friendships with local people.. Getting to know your local neighbours.. Finding social contact with locals (based on: Spradley and Phillips, 1972;

Earley, 1987).. Maintaining interpersonal relationships with local persons (based on

Hammer et al., 1978).. Approaching others (locals) – making the first move in starting up a

friendship (based on Furnham and Bochner, 1982).

(3) Conditions (C):. Living conditions in general.. Dealing with the housing conditions, for instance reliability of electricity or

telephones, or amount of available space.. Dealing with the environment, for instance noise levels, pollution, litter

(based on De Leon and McPartlin, 1995).

(4) Values (V):. Prevailing political values of local nationals.. Religious values of local nationals.

The response scales for each of the items are as shown below. Respondents were askedto answer the following three questions for each of the items above:


adaptation

91


(1) Cognition (C). People who live in a foreign country sometimes are uncertainabout various aspects of life in the foreign culture. How confident are you aboutyour knowledge regarding the following: 7 ¼ completely confident; 1 ¼ notconfident; N/A (move to next item)?

(2) Emotion (E):. Which one of the following four groups of emotions most closely resembles

what you generally feel? Satisfaction/happiness; anxiety/fear; impatience/anger; sadness/depression.

. How strong is this emotion (are these emotions) generally: 7 ¼ extremelystrong; 1 ¼ barely noticeable?

Emotions were coded along a positive-negative range (þ7 to 27) with anxiety, angerand depression forming the negative branch.

Black and Stephens (1989) scale. The nine interaction and general adaptation itemsfrom Black and Stephens’ (1989) scale formed part of the questionnaire. Their scale alsoincludes items related to work adaptation which were omitted here. This approach isthe same as Robie and Ryan’s (1996) who also excluded work-related items from theirstructural equivalence study of the Black and Stephens scale.

The Black and Stephens scale looks at adaptation one-dimensionally along anadjusted-unadjusted range. By not separating cognitions and emotions, their scalemeasures either both concurrently or one of the two. It is impossible to tell which it isand measurement may differ from one respondent to the next. The following questionswere included:

How unadjusted or adjusted are you regarding the following: 1 ¼ unadjusted 7 ¼ adjusted?1. Socialising with host nationals2. Interacting with host nationals on a day-to-day basis3. Living conditions in general4. Housing conditions5. Food6. Shopping7. Cost of living8. Entertainment/recreation facilities and opportunities9. Health care facilities.

Exploratory factor analysis of the Black and Stephens (1989) items showed poor resultsfor the entertainment/recreation and the health care facilities items. They wereexcluded from further analysis. Socialising and interacting items formed acommunication adjustment factor. The other items make up an adjustment togeneral conditions factor.

Independent variablesNovelty (H1 and H2)[6]. Two indicators served to determine cultural novelty. By askingopen-ended questions about the country of citizenship and the country the respondentcurrently lives in an approximate assessment of cultural distance was made. Hofstede’s(1980 1984 1992) four original dimensions of power distance, masculinity/femininity,individualism/collectivism, and uncertainty avoidance provided the first indicator. Hisfifth dimension of long-term orientation is unavailable for most countries covered inthis study. Most of the data refer to home and host countries of Western heritage, while

PR34,1

92


long-term orientation is particularly important when looking at Asian cultures.Therefore, long-term orientation was excluded. Based on Hofstede’s scores for eachcountry on the four dimensions, the absolute difference for each was entered for everyrespondent. In the model below, the simple sum of absolute differences was used[7].

Besides cultural distance the second indicator of novelty used in this study isquality of the respondent’s previous expatriate experience. As discussed earlier, lengthof prior experience has shown very little relation to adaptation outcomes. Thefollowing question was used:

How satisfied were you in your previous overseas stays (if you had differing experiences indifferent stays, please answer for the one that most influences your attitude towards overseasstays)? 1 ¼ very satisfied, 7 ¼ very dissatisfied.

The question was coded in reverse for data processing.Discretion (H3a and H3b). Two questions gauged the discretion expatriates felt in

the foreign environment. One was a more objective, the other a purely subjectivemeasure tapping into respondents’ perceptions. The more objective measure makes adistinction between simple and complex or diverse environments in terms of size andpopulation density. The following question assessed the size of the environment:

In what type of area do you live?1 ¼ rural, 2 ¼ small town, 3 ¼ medium-size town, 4 ¼ major metropolitan area (includingsuburbs), 5 ¼ major city.

A weakness of the question is that it can provide relative information within onecountry but is not necessarily accurate across countries. It is likely that a small town ina pluralistic society allows more freedom than a major city in a repressive country.Since most respondents lived in pluralistic countries, the consequences of thequestion’s weakness should not be too severe.

The other measure of discretion was purely subjective:

How much freedom do local people give you to adhere or not adhere to local norms andvalues? 1 ¼ total freedom; 7 ¼ no freedom at all.

This question refers to what Kim (1988) calls the pressure to conform. As a subjectivemeasure included in the same questionnaire with adaptation outcome scales it maysuffer from common method variance. The question was coded in reverse for dataprocessing.

Self-efficacy ( H4a and H4b). The literature contains well-established self-efficacyscales for general self-efficacy, social self-efficacy (Woodruff and Cashman, 1993;Harrison et al., 1996), and job self-efficacy (Jones, 1986). In addition, plenty of researchexists on measurement and effects of self-efficacy in various settings (e.g. Lust et al.,1993; Mitchell et al., 1994; Mone, 1994; Parker, 1994; Saks, 1995). Space restrictions inthe questionnaire – general and social self-efficacy scales taken together contain 23items – led to the exclusion of the scales. Rather, attention was focused on one aspectof self-efficacy that is only relevant in cross-cultural contact and not included inestablished scales: language skills.

The following questions were used:

How well do you speak/understand/read/write the language?1 ¼ not at all; 7 ¼ like a native speaker


adaptation

93


– Speak– Understand– Write– ReadIs your level of knowledge of the local language:1 ¼ much more than sufficient; 7 ¼ completely insufficient?

The second question was coded in reverse for data analysis.Social networks (H5). As with self-efficacy, extensive literature exists on the

influence of the social environment on adaptation, coping and satisfaction in variouscontexts, including expatriation (Tardy, 1985; Maynard, 1986; Black, 1990a, b; Taalet al., 1993; van der Poel, 1993). Tardy (1985) comprehensively covers all facets of socialsupport in a discussion of its measurement. The measures used in empirical researchare diverse and sometimes lengthy, including as many as 41 separate items (Maynard,1986). They deal with aspects as varied as emotional support, instrumental support(Taal et al., 1993; van der Poel, 1993), social companionship (van der Poel, 1993), andnetwork support satisfaction regarding family, friendship, social and community,profession, and work (Maynard, 1986). Again, space restrictions in the questionnaireprevented the inclusion of these established scales and required focusing on issuesunique to expatriate adaptation.

Only one simple question was used to look at the effects of networks:

How would you characterise your contacts with local nationals?1 ¼ very friendly, 7 ¼ very unfriendly.

The question was coded in reverse for data analysis.

Analysis methodThis study used the Amos structural equation modelling package of SPSS. It wasused for confirmatory factor analysis, an advanced statistical technique thatcombines factor with regression analysis. Groups of dependent and independentvariables are factor analysed. In Figure 1, the independent variables of speak,understand, write and read load on a common factor language skills. Similarly, thedependent variables on the right load on the factors of cognition-related relationshipformation (RC) and emotion-related relationship formation (RE). There are severalother observed (all rectangular boxes represent observed variables) independentvariables such as “satprev” (satisfaction with previous experience abroad). Thearrows in the middle show the hypothesised relationships between independentvariables/language factor and dependent factors of adaptation. Regression weightswere calculated for these arrows. Curved arrows represent covariances amongindependent variables. Confirmatory factor analysis produces model fit statistics,which indicate how well the proposed theoretical model fits the data analysed. Forbetter comparison, fit statistics for the perfectly-fitting saturated model and theterribly-fitting independence model are also calculated. Statistics are in theAppendix.

Figure 1 only exemplifies the underlying model. The left side shows all independentvariables. On the right side is a sample of the dependent variables: relationship factorfor cognitions and for emotions. The complete model includes the same arrows for thedependent factors of day-to-day interactions, conditions, and values. For the Black and

PR34,1

94


Stephens (1989) scale the independent variables represented in Figure 1 correlate withinteraction and general conditions factors. Error variables were omitted in the drawing,but not in the analysis.

ResultsSample demographicsThis study used a convenience sample (n ¼ 204). It consists mainly of members ofAmerican and British women’s clubs in Madrid, Spain and the Frankfurt metropolitanarea in Germany and their spouses (61.9 per cent of sample population). In addition,expatriate employees of a multinational corporation residing in and outside of Europeparticipated in the study. Finally, a small number of expatriate faculty members at anAmerican university and members of the expatriate community in Madrid respondedas well.

Two-thirds of the respondents were female. Their age ranged from 20 to 73 witha mean of 40 years and a median of 38. Almost all respondents (97.5 per cent ofresponses; two missing answers) were under 60 years of age. Roughlythree-quarters were married, about 15 per cent were single, and 5 per cent werecohabitating. Respondents were highly educated. A total of 80 per cent had

Figure 1.Model used in analysis


adaptation

95


obtained at least a Bachelor’s degree, with a further 15 per cent having completedsome course work towards a college degree. Half of the respondents had childrenresiding with them.

Exactly half of the respondents stated that they were the expatriate, that is theperson initiating the overseas stay; 42 per cent were the accompanying spouse, and 8per cent declined to answer the question.

Almost two-thirds of respondents were currently employed (64 per cent). Their jobsfell into three main categories: general business such as information technology,finance, sales and marketing (58 per cent); journalism (33 per cent); and public sectorsuch as positions in embassies and teaching (9 per cent). Most of those employedindicated the level of their employment. Almost half (46 per cent) worked innon-supervisory professional or technical positions. A further 27 per cent worked inmid- and upper-level management positions. A total of 10 per cent were in executivepositions and 6 per cent reported to be self-employed. The remaining 11 per centworked in administrative and lower management positions.

The length of the current stay ranged from a minimum of three months to amaximum of 42 years. The median length of stay was just under 2.5 years, the meanwas 5.5 years. Table I summarises length of stay.

Of the respondents, 51 per cent reported that the current was not their first overseasstay. Of those on multiple assignments, 39 per cent were on their second assignment,23 per cent on their third, and a further 16 per cent on their fourth. The remainder hadhad up to a maximum of 11 overseas stays. For multiple expatriates the total length ofoverseas experience ranged from 16 months to 44 years.

Respondents resided in 22 different countries. Almost half of them (48 per cent)lived in Germany, 28 per cent were in Spain, and 6.5 per cent in the USA.

About three-quarters of respondents were citizens of either the USA (39 per cent) orthe UK (35 per cent). Only three respondents reported a dual citizenship, all of themAmerican and Spanish.

A plurality lived in a major city (37 per cent), followed by a major metropolitan area(26 per cent). A total of 12 per cent lived in a medium-size town, 21 per cent in a smalltown and only 5 per cent lived in a rural area.

Value label Value Frequency Per cent Valid per cent Cum per cent

Up to 3 months 1 2 1.0 1.0 1.04 to 6 months 2 27 13.2 13.4 14.47 to 9 months 3 16 7.8 7.9 22.310 to 12 months 4 11 5.4 5.4 27.71 to 1.5 years (13-18 months) 5 17 8.3 8.4 36.11.5 to 2 years (19-24) 6 23 11.3 11.4 47.52 to 2.5 years (25-30) 7 12 5.9 5.9 53.52.5 to 3 years (31-36) 8 16 7.8 7.9 61.43 to 4 years (37-48) 9 9 4.4 4.5 65.84 to 5 years (49-60) 10 5 2.5 2.5 68.35 to 10 years (61-120) 11 31 15.2 15.3 83.7Over 10 years (121 þ ) 12 33 16.2 16.3 100.0

2 1.0 MissingTotal 204 100.0 100.0

Table I.Length of stay

PR34,1

96


Tables AI-AVI in the Appendix provide the parameter estimates for evaluation of thehypotheses based on the new outcome measure described above and the Black andStephens (1989) scale.

Model fitBoth models, the one for the refined outcome measure and the one for the Black andStephens (1989) scale, show a borderline fit. Arbuckle and Wothke (1999) cite a rule ofthumb for CMIN/DF based on the literature that calls for a value of ,2 to ,5 asreasonable. This study is mostly concerned with demonstrating that the separatemeasurement of cognitive and emotional adaptation outcomes is superior to aone-dimensional measurement. It does not attempt to provide a complete model ofadaptation. Therefore, the model fit indicated seems sufficient. Of course, better fittingmodels would be preferable to more firmly establish the superiority of the proposedmeasure. More research is required to provide further evidence.

Explained varianceThe refined outcome measure model shows an explained variance range of 0.152 to0.704, better than the Black and Stephens (1989) model except for two instances. TheBlack and Stephens (1989) model has an explained variance of 0.314 for communicationand only 0.099 for general adjustment.

Results related to proposed hypotheses

H1a.Cultural distance is negatively related to both cognitive and emotional adaptation.

H2b. Positive prior expatriate experience is positively related to emotional adaptationand unrelated to cognitive adaptation.

When tested with the Black and Stephens (1989) scale both H1a and H1b have to berejected[8].

The refined outcome measure shows significant negative relationships of culturaldistance with three of the four emotion factors, but none with the cognitive factors,lending partial support to H1a (see Tables II and III). This indicates that a largecultural distance may be an issue for expatriate well-being, but not for their confidencein knowing how to behave. As this is a cross-sectional study with a median length ofstay of just under 2.5 years, this is understandable. The “average” person in the samplehas had enough time to learn the rules, but perhaps not enough to feel entirelycomfortable with them.

Positive prior expatriate experience has highly significant correlations with twocognition and three emotion factors, lending partial support to H1b. The significantrelationships with cognition factors are surprising. Both factors relate to interactionswith people, either on a day-to-day basis or regarding the formation of relationships.Previous expatriate experience helps people become more confident in how to conductinteractions with the local population perhaps because they have developed their ownway of interacting with people from cultures foreign to them. One seasoned expatriatementioned to the author that she had developed, over the course of several moves, aportable expatriate lifestyle that requires only relatively little adaptation. This may


adaptation

97


affect the confidence with which expatriates approach interactions withrepresentatives of the host culture.

The different results for the refined and the Black and Stephens (1989) measureprovide evidence that the one-dimensional measurement may lose importantinformation about the association of cultural distance and prior expatriateexperience and adaptation outcomes.

Perhaps the critical comments of some scholars about the measurement of culturaldistance and prior expatriate experience should be re-evaluated in light of this studyand at least part of the criticism directed to the way the dependent variables weremeasured.

H2a. Freedom to adhere or not adhere to local rules is positively related to cognitiveand emotional adaptation.

H2b. The size of the city of residence is positively related to cognitive and emotionaladaptation.

Estimate of standardised regression weights

H1a RE ˆ hof_sum 20.185AE ˆ hof_sum 20.277CE ˆ hof_sum 20.271

H1b RC ˆ satprev 0.297AC ˆ satprev 0.252RE ˆ satprev 0.399AE ˆ satprev 0.414CE ˆ satprev 0.258

H2a VC ˆ freedom 0.173CC ˆ freedom 0.195AE ˆ freedom 0.192VE ˆ freedom 0.243

H2b CC ˆ area 20.199AE ˆ area 20.195CE ˆ area 20.416

H3a RC ˆ SKILL 0.463AC ˆ SKILL 0.782VC ˆ SKILL 0.552RE ˆ SKILL 0.510AE ˆ SKILL 0.878

H3b AE ˆ sufficnt 20.458H4 RC ˆ charlocc 0.221

VC ˆ charlock 0.140RE ˆ charlocc 0.189

Table II.Refined outcome measure(significant relationships)

Estimate of standardised regression weights

H2a GEN ˆ freedom 0.233COMM ˆ freedom 0.191

H3a COMM ˆ SKILL 0.416H4 COMM ˆ charlocc 0.156

Table III.Black and Stephens(1989) measure(significant relationships)

PR34,1

98


H2a is confirmed by the Black and Stephens (1989) scale, while H2b is not confirmed.The refined outcome measure shows two significant relationships each of freedom

to adhere to local rules with cognition and emotion factors, lending partial support toH2a.

The freedom expatriates experience is associated with the confidence they have inrelation to values and general living conditions. The respondents’ confidence regardingvalues can be understood as stemming from a reduced pressure to know exactly thelocal value system. The relationship with living conditions is more difficult to explainwithout further information. Perhaps expatriates who are confident had good supportfrom their employer in dealing with conditions. But this is pure speculation that has tobe tested in future studies.

Freedom also relates to more positive feelings regarding values and day-to-dayinteractions. They probably feel better because they do not have to adhere tovalues and to daily interaction practices they do not share or support.

One cognition and two emotion factors are significantly, but negatively, related tothe size of the area of residence. H2b is rejected. Expatriates who live in smallercommunities are more confident and feel better, particularly when it comes to livingconditions. It is harder to get orientated in a big city and understand all the specificsof living conditions, and conditions may be experienced as worse by big citydwellers, making them feel worse. In the same vein, life in big cities may be morestressful thus causing more tension in day-to-day interactions resulting in worsefeelings.

H3a. Language skills are positively related to cognitive and emotional adaptation.

H3b. The perception of possessing sufficient language skills is positively related tocognitive and emotional adaptation.

The Black and Stephens (1989) scale supports H3a in relation to communication, butnot to general adjustment. This makes perfect sense, since language skills areimportant in interaction, but hardly in adjustment to general conditions. H3ba was notconfirmed.

The refined outcome measure shows similar results to the one-dimensional one.Relationship and day-to-day interaction factors for both, cognitions and emotions,are highly correlated with language skills. The cognitive side of the values factor isalso strongly associated with language skills, while the emotional side is not.Language skills may help to learn about local values, but not to make them morecomfortable. Thus, the refined measure partially confirms H3a and shows itspotential to add more depth to the understanding of adaptation. It does so bydistinguishing additional factors of adaptation and by separating cognitive andemotional dimension.

H3b is rejected in its entirety. It shows only one significant relationship, which isquite strong but opposite to expectations. The more sufficient people feel theirlanguage skills are, the worse they feel in day-to-day interactions. It is difficult toexplain this. This relationship needs to be tested again in future studies.

H4. The friendly character of contacts with the local population is positively related tocognitive and emotional adaptation


adaptation

99


The friendly character of contacts with locals is significantly related with Black andStephens’ (1989) communication factor, lending support to H4. As with H3a, it isunderstandable that the association is with communication and not with generalconditions.

The refined measure shows significant relationships with two cognition factors andone emotion factor. This partially confirms H4. The distinction between relationshipand day-to-day interaction factors as opposed to Black and Stephens’ (1989)communication factor shows that relationships are particularly associated withfriendly contacts. Expatriates experience greater ease in the formation of relationshipswith representatives of the local culture, if contacts are friendly. Yet, friendly contactsseem unrelated to day-to-day interactions.

In addition, the understanding of local values, but not the attitude towards them, isrelated to local contacts. Friendly contacts may enable expatriates to learn about localvalues. How they feel about those values is unrelated to the character of contacts.

DiscussionThe results of this study demonstrate that a more refined measurement of adaptationoutcomes, which distinguishes cognition and emotion factors, is superior to aone-dimensional measurement. The association of cultural distance with adaptationoutcomes is a case in point. Theoretically and for every expatriate it is clear that alarger cultural difference is related to a more difficult adaptation process. Yet, theBlack and Stephens (1989) measure, which does not distinguish between cognitive andaffective outcomes, was unable to establish any significant relationships.

The study also demonstrates that the distinction of general and communicationadjustment (besides work adjustment, which was excluded here) in theone-dimensional measure may be to simplistic. Several results showed that there isa benefit to distinguishing at least the formation of relationships and day-to-dayinteractions. The separate value-related factor also produced some unique andinteresting results.

The discussion of the results in relation to the hypotheses shows also that theorycould be enhanced significantly. The current theory of expatriation does notdistinguish between the relationships of independent variables with cognitive and withaffective outcomes. Based on the current state of theory it is sometimes difficult toexplain the presented results. As always, some of the less explicable results can, and insome cases should, be attributed to shortcomings of sample selection, data collection,and data quality. But the point is, the state of theory based on one-dimensionalmeasures does not allow us to do more than speculate on whether a result is surprisingbecause the underlying data are poor or because we simply do not sufficientlyunderstand the dynamics of cross-cultural adaptation. Research based on refinedoutcome measures will provide a challenge to theorists to come up with better, moredetailed models.

Practical relevanceIt is nice if a new tool leads to improved theory and an advancement of the discipline.But what is the ultimate benefit to expatriates and practitioners?

The most immediate benefit may follow from an application of the instrument toregularly monitor adaptation levels of expatriates. Depending on whether cognitive

PR34,1

100


or affective adaptation is lacking, support measures can be tailored. If cognitivecertainty is less than expected, additional cognitive input such as training, books,etc. may be provided. If expatriates do not feel overly good, psychological supportmay be the answer. The better the diagnosis, the better the resulting interventionwill be.

Once a sufficient number of studies has led to a further refinement of ourunderstanding of the adaptation process, preparatory as well as follow-up training andsupport abroad may be designed to provide the most pertinent services as apreventative measure to the most common adaptation problems. Expatriatemanagement has made great strides over the last two decades, but it is still farfrom perfect.

To cite just one small example from this study: basic language skills training is acommon feature in expatriate preparation. However, once abroad, most expatriatesdrop classes under the pressures of work and companies happily oblige to save money.Trailing family members often receive less training to begin with. The strongrelationship of language skills with cognitive and affective adaptation warrantsheightened attention to continued language training once abroad.

Limitations and outlookThis study, like many before, suffers from a number of shortcomings:

. It is based on a cross-sectional convenience sample[9].

. The sample size of just over 200, while about average for expatriate studies, issmall for confirmatory factor analysis.

. The independent variables do not represent a systematic selection from acomprehensive theoretical model. They were included because prior research hasdealt with them, allowing the development of grounded hypotheses.

On the positive side, the study has demonstrated that a two-dimensional measurementof adaptation outcomes is more sensitive and thus superior to the one-dimensionalmeasurement still prevalent in expatriate studies.

Future research has to show whether these results can be reproduced to confirm theview about the superiority of the proposed two-dimensional measure. Ideally, thesestudies would be based on a longitudinal sample and a fully-elaborated theory ofcross-cultural adaptation.

Notes

1. This adds an interesting twist to adaptation research since people tend to remember selectivelywhat emotions they felt during the course of an interaction (see LeDoux, 2002, p. 202, for anexample). This supports Roth’s (1997) contention that emotions are concentrated experiences,and is particularly the case when people try to remember them. Therefore, reports on emotionsare cognitively filtered to fit with a person’s construct of reality.

2. An exception is Wong-Rieger (1984) who included nine emotions to measure responses toadaptation problem scenarios.

3. Nicholson (1984) has analysed the role of novelty and discretion in work role transitions.With appropriate alterations his analysis can be usefully applied to cross-culturaladaptation.


adaptation

101


4. A different part of the study included 67 items, most used in prior research andsupplemented by additional items related to relationships, day-to-day interactions, generalconditions, institutions, values and norms. Factor analysis resulted in the selection of the 12items used here. Day-to-day interactions and values dimensions include only two items each.Exploratory factor analysis showed them as one dimension with four items. Butconfirmatory factor analysis indicators improved when a model with the two separatedimensions was used. Items related to institutions and to norms did not form significantfactors.

5. Dependent variable names are combinations such as “RC” or “RE”, the first indicating thecognitive relationship factor and the second indicating the emotion-based relationship factor.

For the Black and Stephens (1989) scale “GEN” denotes general adjustment and “COMM”denotes communication adjustment. Items in the refined outcome measure taken from orinspired by Black and Stephens (1989) are not separately identified.

6. Independent variable names in the statistics: cultural difference is called “hof_sum” in thestatistics. Satisfaction with previous experience is called “satprev.” The type of area is called“area” and the freedom from local norms and values is called “freedom.” The four variables“speak,” “understand,” “write,” and “read” were combined into a language skills factor fordata analysis called “SKILL.” Perceived knowledge of the local language is called “sufficnt.”The characterisation of contacts is called “charlocc.”

7. No corrections such as those suggested by Kogut and Singh (1988) were made.

8. For the evaluation of H1a-H4 with the Black and Stephens (1989) scale any hypothesisedrelationship, for cognitions or emotions together or separately, is treated asone-dimensional.

9. The discussion of results sometimes uses a language of causality. The reader should beaware that this cross-sectional research has only established associations between variables.A more elaborate research design involving longitudinal data is required to show causalrelationships.

References

Arbuckle, J.L. and Wothke, W. (1999), Amos 4.0 User’s Guide, SmallWaters Corporation,Chicago, IL.

Aryee, S., Chay, Y.W. and Chew, J. (1996), “An investigation of the willingness of managerialemployees to accept an expatriate assignment”, Journal of Organizational Behavior, Vol. 17No. 3, pp. 267-83.

Aycan, Z. (1997), “Expatriate adjustment as a multifaceted phenomenon: individual andorganizational predictors”, The International Journal of Human Resource Management,Vol. 8 No. 4, pp. 434-56.

Berger, C.R. (1979), “Beyond initial interaction: uncertainty, understanding, and the developmentof interpersonal relationships”, in Giles, H. and St Clair, R.N. (Eds), Language and SocialPsychology, University Park Press, Baltimore, MD.

Berger, C.R. (1987), “Communicating under uncertainty”, in Roloff, M.E. and Miller, G.R. (Eds),Interpersonal Processes: New Directions in Communication Research, Sage Publications,Newbury Park, CA.

Berger, C.R. and Calabrese, R.J. (1975), “Some explorations in initial interaction and beyond:toward a developmental theory of interpersonal communication”, Human CommunicationResearch, Vol. 1 No. 2, pp. 99-112.

PR34,1

102


Berger, C.R. and Gudykunst, W.B. (1991), “Uncertainty and communication”, in Dervin, B. andVoigt, M.J. (Eds), Progress in Communication Sciences, Vol. 10, Ablex Publishing,Norwood, NJ.

Berry, J.W., Kim, U. and Boski, P. (1988), “Psychological acculturation of immigrants”, in Kim, Y.Y.and Gudykunst, W.B. (Eds), Cross-cultural Adaptation – Current Approaches, SagePublications, Newbury Park, CA.

Bhuian, S.N. and Islam, M.S. (1996), “Continuance commitment and extrinsic job satisfactionamong a novel multicultural expatriate workforce”, The Mid-Atlantic Journal of Business,March, pp. 35-46.

Black, J.S. (1988), “Work role transitions: a study of American expatriate managers in Japan”,Journal of International Business Studies, Summer, pp. 277-94.

Black, J.S. (1990a), “Locus of control, social support, stress, and adjustment in internationaltransfers”, Asia Pacific Journal of Management, Vol. 7 No. 1, pp. 1-29.

Black, J.S. (1990b), “The relationship of personal characteristics with the adjustment of Japaneseexpatriate managers”, Management International Review, Vol. 30 No. 2, pp. 119-34.

Black, J.S. (1994), “O Kaerinasai: factors related to japanese repatriation adjustment”, HumanRelations, Vol. 47 No. 12, pp. 1489-508.

Black, J.S. and Gregersen, H.B. (1991a), “When Yankee comes home: factors related to expatriateand spouse repatriation adjustment”, Journal of International Business Studies, 4thQuarter, pp. 671-94.

Black, J.S. and Gregersen, H.B. (1991b), “Antecedents to cross-cultural adjustment for expatriatesin Pacific Rim assignments”, Human Relations, Vol. 44 No. 5, pp. 497-515.

Black, J.S. and Gregersen, H.B. (1991c), “The other half of the picture: antecedents of spousecross-cultural adjustment”, Journal of International Business Studies, 3rd Quarter,pp. 461-77.

Black, J.S. and Mendenhall, M.E. (1991), “The U-curve adjustment hypothesis revisited: a reviewand theoretical framework”, Journal of International Business Studies, 2nd Quarter,pp. 225-47.

Black, J.S. and Stephens, G.K. (1989), “The influence of the spouse on American expatriateadjustment and intent to stay in Pacific Rim overseas assignments”, Journal ofManagement, Vol. 15 No. 4, pp. 529-44.

Black, J.S., Gregersen, H.B. and Mendenhall, M.E. (1992a), “Toward a theoretical framework ofrepatriation adjustment”, Journal of International Business Studies, 4th Quarter, pp. 737-60.

Black, J.S., Gregersen, H.B. and Mendenhall, M.E. (1992b), Global Assignments: SuccessfullyExpatriating and Repatriating International Managers, Jossey-Bass Publishers, SanFrancisco, CA.

Black, J.S., Mendenhall, M.E. and Oddou, G. (1991), “Toward a comprehensive model ofinternational adjustment: an integration of multiple theoretical perspectives”, Academy ofManagement Review, Vol. 16 No. 2, pp. 291-317.

Brewster, C. and Pickard, J. (1994), “Evaluating expatriate training”, International Studies ofManagement and Organization, Vol. 24 No. 3, pp. 18-35.

(The) Chinese Culture Connection (1987), “Chinese values and the search for culture-freedimensions of culture”, Journal of Cross-Cultural Psychology, Vol. 18 No. 2, pp. 143-64.

Church, A.T. (1982), “Sojourner adjustment”, Psychological Bulletin, Vol. 91 No. 3, pp. 540-72.

David, K.H. (1972), “Intercultural adjustment and applications of reinforcement theory toproblems of ‘culture shock’”, Trends, Vol. 4 No. 3, pp. 1-64.


adaptation

103


de Leon, C.T. and McPartlin, D. (1995), “Adjustment of expatriate children”, in Selmer, J. (Ed.),Expatriate Management – New Ideas for International Business, Quorum Books, Westport,CT and London.

Earley, P.C. (1987), “Intercultural training for managers: a comparison of documentary andinterpersonal methods”, Academy of Management Journal, Vol. 30 No. 4, pp. 685-98.

Furnham, A. and Bochner, S. (1982), “Social difficulty in a foreign culture: an empirical analysisof culture shock”, in Bochner, S. (Ed.), Cultures in Contact: Studies in Cross-culturalInteraction, Pergamon Press, Oxford.

Gregersen, H.B. and Black, J.S. (1990), “A multifaceted approach to expatriate retention ininternational assignments”, Group & Organization Studies, Vol. 15 No. 4, pp. 461-85.

Grove, C.L. and Torbiorn, I. (1985), “A new conceptualization of intercultural adjustment andthe goals of training”, International Journal of Intercultural Relations, Vol. 9 No. 2,pp. 205-33.

Gudykunst, W.B. (1988), “Uncertainty and anxiety”, in Kim, Y.Y. and Gudykunst, W.B. (Eds),Theories in Intercultural Communication, Sage Publications, Newbury Park, CA.

Gudykunst, W.B. and Hammer, M.R. (1988), “Strangers and hosts – an uncertainty reductionbased theory of intercultural adaptation”, in Kim, Y.Y. and Gudykunst, W.B. (Eds),Cross-Cultural Adaptation – Current Approaches, Sage Publications, Newbury Park,CA.

Gudykunst, W.B. and Nishida, T. (1986), “Attributional confidence in low- and high-contextcultures”, Human Communication Research, Vol. 12 No. 4, pp. 525-49.

Hammer, M.R., Gudykunst, W.B. and Wiseman, R.L. (1978), “Dimensions of interculturaleffectiveness: an exploratory study”, International Journal of Intercultural Relations,Winter, pp. 382-92.

Harrison, J.K., Chadwick, M. and Scales, M. (1996), “The relationship between cross-culturaladjustment and the personality variables of self-efficacy and self-monitoring”,International Journal of Intercultural Relations, Vol. 20 No. 2, pp. 167-88.

Hofstede, G. (1980), “Motivation, leadership, and organization: do American theories applyabroad?”, Organizational Dynamics, Summer, pp. 42-63.

Hofstede, G. (1984), Culture’s Consequences: International Differences in Work-related Values,abridged ed., Sage Publications, Newbury Park, CA.

Hofstede, G. (1992), “Cultural dimensions in people management – the socialization perspective”,in Pucik, V., Tichy, N.M. and Barnett, C.K. (Eds), Globalizing Management – Creating andLeading the Competitive Organization, Wiley & Sons, New York, NY.

Hofstede, G. and Bond, M.H. (1984), “Hofstede‘s culture dimensions: an independentvalidation using Rokeach’s value survey”, Journal of Cross-cultural Psychology, Vol. 15,pp. 417-33.

Hofstede, G. and Bond, M.H. (1988), “The Confucius connection: from cultural roots to economicgrowth”, Organizational Dynamics, Vol. 16 No. 4, pp. 4-21.

Javidan, M. and House, R.J. (2001), “Cultural acumen for the global manager: lessons from ProjectGlobe”, Organizational Dynamics, Vol. 29 No. 4, pp. 289-305.

Jones, G.R. (1986), “Socialization tactics, self-efficacy, and newcomers’ adjustments toorganizations”, Academy of Management Journal, Vol. 29 No. 2, pp. 262-79.

Kemper, T.D. (1987), “How many emotions are there? Wedding the social and the autonomiccomponents”, American Journal of Sociology, Vol. 93 No. 2, pp. 263-89.

Kim, Y.Y. (1988), Communication and Cross-cultural Adaptation, Multilingual Matters, Clevedonand Philadelphia, PA.

PR34,1

104


Kogut, B. and Singh, H. (1988), “The effect of national culture on the choice of entry mode”,Journal of International Business Studies, Fall, pp. 411-32.

Kraimer, M.L., Wayne, S.J. and Jaworski, R.A. (2001), “Sources of support and expatriateperformance: the mediating role of expatriate adjustment”, Personnel Psychology, Vol. 54No. 1, pp. 71-99.

LeDoux, J. (2002), Synaptic Self – How Our Brains Become Who We Are, Viking Penguin, NewYork, NY.

Lust, J.A., Celuch, K.G. and Showers, L.S. (1993), “A note on issues concerning the measurementof self-efficacy”, Journal of Applied Social Psychology, Vol. 23 No. 17, pp. 1426-34.

Mattenklott, A. (1996), “On the methodology of empirical research on suspense”, in Vorderer, P.,Wulff, H.J. and Friedrichsen, M. (Eds), Suspense: Conceptualizations, Theoretical Analyses,and Empirical Explorations, Laurence Erlbaum Associates, Mahwah, NJ.

Maynard, M. (1986), “Measuring work and support network satisfaction”, Journal of EmploymentCounseling, March, pp. 9-19.

Mitchell, T.R., Hopper, H., Daniels, D., George-Falvy, J. and James, L.R. (1994), “Predictingself-efficacy and performance during skill acquisition”, Journal of Applied Psychology,Vol. 79 No. 4, pp. 506-17.

Mone, M.A. (1994), “Comparative validity of two measures of self-efficacy in predicting academicgoals and performance”, Educational and Psychological Measurement, Vol. 54 No. 2,pp. 516-29.

Nash, D. (1967), “The fate of Americans in a Spanish setting: a study in adaptation”, HumanOrganization, Vol. 26 No. 3, pp. 157-63.

Naumann, E. (1993a), “Organizational predictors of expatriate job satisfaction”, Journal ofInternational Business Studies, 1st Quarter, pp. 61-80.

Naumann, E. (1993b), “Antecedents and consequences of satisfaction and commitment amongexpatriate managers”, Group & Organization Management, Vol. 18 No. 2, pp. 153-87.

Nicholson, N. (1984), “A theory of work role transitions”, Administrative Science Quarterly,Vol. 29, June, pp. 172-91.

Nicholson, N. and Imaizumi, A. (1993), “The adjustment of Japanese expatriates to living andworking in Britain”, British Journal of Management, Vol. 4 No. 2, pp. 119-34.

Oberg, K. (1960), “Cultural shock: adjustment to new cultural environments”, PracticalAnthropology, July-August, pp. 177-82.

Parker, L.E. (1994), “Working together: perceived self- and collective-efficacy at the workplace”,Journal of Applied Social Psychology, Vol. 24 No. 1, pp. 43-59.

Plato (1989), Republic, in Hamilton, E. and Cairns, H. (Eds), The Collected Dialogues, PrincetonUniversity Press, Princeton, NJ.

Robie, C. and Ryan, A.M. (1996), “Structural equivalence of a measure of cross-culturaladjustment”, Educational and Psychological Measurement, Vol. 56 No. 3, pp. 514-21.

Roth, G. (1997), Das Gehirn und seine Wirklichkeit – Kognitive Neurobiologie und ihrephilosophischen Konsequenzen, Suhrkamp, Frankfurt am Main.

Saks, A.M. (1995), “Longitudinal field investigation of the moderating and mediating effects ofself-efficacy on the relationship between training and newcomer adjustment”, Journal ofApplied Psychology, Vol. 80 No. 2, pp. 211-25.

Schuetz, A. (1944), “The stranger: an essay in social psychology”, The American Journal ofSociology, May, pp. 499-507.


adaptation

105


Selmer, J. (2001), “Adjustment of Western European vs North American expatriate managers inChina”, Personnel Review, Vol. 30 No. 1, pp. 6-21.

Selmer, J. (2002), “Practice makes perfect? International adjustment experience and expatriateadjustment”, Management International Review, Vol. 42 No. 1, pp. 71-87.

Simmel, G. (1987), “Der Fremde”, in Simmel, G. (Ed.), Das individuelle Gesetz – PhilosophischeExkurse, Suhrkamp Taschenbuch Wissenschaft, Frankfurt am Main (originally publishedin 1908).

Spradley, J.P. and Phillips, M. (1972), “Culture and stress: a quantitative analysis”, AmericanAnthropologist, Vol. 74 No. 6, pp. 518-29.

Taal, E., Rasker, J.J., Seydel, E.R. and Wiegman, O. (1993), “Health status, adherence with healthrecommendations, self-efficacy and social support in patients with rheumatoid arthritis”,Patient Education and Counseling, Vol. 20 No. 2-3, pp. 63-76.

Taft, R. (1977), “Coping with unfamiliar cultures”, in Warren, N. (Ed.), Studies in Cross-culturalPsychology, Academic Press, New York, NY.

Takeuchi, R., Yun, S. and Russell, J.E.A. (2002), “Antecedents and consequences of the perceivedadjustment of Japanese expatriates in the USA”, International Journal of Human ResourceManagement, Vol. 13 No. 8, pp. 1224-44.

Tardy, C.H. (1985), “Social support measurement”, American Journal of Community Psychology,Vol. 13 No. 2, pp. 187-202.

Taylor, S. and Napier, N.K. (1996a), “Successful women expatriates: the case of Japan”, Journal ofInternational Management, Vol. 2 No. 1, pp. 51-78.

Taylor, S. and Napier, N. (1996b), “Working in Japan: lessons from women expatriates”, SloanManagement Review, Spring, pp. 76-84.

Triandis, H.C. (1980), “A theoretical framework for the study of bilingual-bicultural adaptation”,International Review of Applied Psychology, Vol. 29, pp. 7-16.

Trompenaars, F. (1993), Riding the Waves of Culture: Understanding Cultural Diversity inBusiness, Nicholas Brealey Publishing, London.

Van der Poel, M.G.M. (1993), “Delineating personal support networks”, Social Networks, Vol. 15No. 1, pp. 49-70.

Wong-Rieger, D. (1984), “Testing a model of emotional and coping responses to problems inadaptation: foreign students at a Canadian university”, International Journal ofIntercultural Relations, Vol. 8 No. 2, pp. 153-84.

Woodruff, S.L. and Cashman, J.F. (1993), “Task, domain, and general efficacy: a re-examinationof the self-efficacy scale”, Psychological Reports, Vol. 72, pp. 423-32.

Appendix

Refined outcome measure modelAnalysis summary (see Tables AI-AIII):

. The model is recursive.

. Sample size ¼ 204

Black and Stephens (1989) model

Analysis summary (see Tables AIV-AVI):. The model is recursive.. Sample size ¼ 204.

PR34,1

106


Est

imat

eof

reg

ress

ion

wei

gh

tsS

.E.

C.R

.P

Est

imat

eof

stan

dar

dis

edre

gre

ssio

nw

eig

hts

RC

ˆh

of_

sum

20.

003

0.00

22

1.51

30.

130

20.

089

AC

ˆh

of_

sum

20.

001

0.00

22

0.49

50.

620

20.

033

VC

ˆh

of_

sum

0.00

40.

003

1.49

30.

136

0.08

9C

Cˆ

hof

_su

m2

0.00

20.

002

21.

339

0.18

02

0.10

0RE

ˆhof_sum

20.017

0.006

22.784

0.005

20.185

AE

ˆhof_sum

20.

021

0.007

22.951

0.003

20.277

VE

ˆh

of_

sum

20.

011

0.00

62

1.81

10.

070

20.

174

CE

ˆhof_sum

20.025

0.008

23.215

0.001

20.271

RC

ˆsatprev

0.277

0.065

4.261

***

0.297

AC

ˆsatprev

0.242

0.076

3.183

0.001

0.252

VC

ˆsa

tpre

v0.

115

0.08

81.

307

0.19

10.

091

CC

ˆsa

tpre

v0.

075

0.05

51.

363

0.17

30.

120

RE

ˆsatprev

10.048

0.208

5.041

***

0.399

AE

ˆsatprev

0.874

0.236

3.700

***

0.414

VE

ˆsa

tpre

v0.

145

0.18

30.

790

0.42

90.

082

CE

ˆsatprev

0.678

0.257

2.633

0.008

0.258

RC

ˆfr

eed

om0.

099

0.05

51.

789

0.07

40.

107

AC

ˆfr

eed

om0.

118

0.06

41.

831

0.06

70.

124

VC

ˆfreedom

0.217

0.075

2.872

0.004

0.173

CC

ˆfreedom

0.121

0.048

2.528

0.011

0.195

RE

ˆfr

eed

om0.

043

0.17

20.

252

0.80

10.

017

AE

ˆfreedom

0.401

0.195

2.057

0.040

0.192

VE

ˆfreedom

0.427

0.185

2.303

0.021

0.243

CE

ˆfr

eed

om0.

177

0.21

70.

816

0.41

50.

068

RC

ˆar

ea2

0.01

80.

060

20.

296

0.76

72

0.01

8A

Cˆ

area

20.

042

0.06

92

0.60

00.

548

20.

041

VC

ˆar

ea0.

078

0.08

20.

959

0.33

70.

059 (continued

)

Table AI.Regression weights


adaptation

107


Est

imat

eof

reg

ress

ion

wei

gh

tsS

.E.

C.R

.P

Est

imat

eof

stan

dar

dis

edre

gre

ssio

nw

eig

hts

CC

ârea

20.131

0.052

22.525

0.012

20.199

RE

âr

ea2

0.10

10.

186

20.

543

0.58

72

0.03

7AE

ârea

20.432

0.211

22.044

0.041

20.195

VE

âr

ea2

0.35

10.

186

21.

880

0.06

02

0.18

8CE

ârea

210.147

0.247

24.644

***

20.416

RC

ˆSKILL

0.356

0.078

4.546

***

0.463

AC

ˆSKILL

0.616

0.099

6.248

***

0.782

VC

ˆSKILL

0.574

0.105

5.445

***

0.552

CC

ˆS

KIL

L0.

097

0.06

41.

507

0.13

20.

188

RE

ˆSKILL

10.102

0.249

4.427

***

0.510

AE

ˆSKILL

10.526

0.302

5.060

***

0.878

VE

ˆS

KIL

L0.

381

0.22

91.

662

0.09

70.

261

CE

ˆS

KIL

L0.

259

0.29

80.

868

0.38

60.

120

RC

ˆsu

fficn

t0.

069

0.06

41.

084

0.27

80.

106

AC

ˆsu

fficn

t2

0.10

70.

074

21.

442

0.14

92

0.16

0V

Cˆ

suffi

cnt

0.01

50.

087

0.17

50.

861

0.01

7C

Cˆ

suffi

cnt

0.05

50.

054

1.01

50.

310

0.12

5R

Eˆ

suffi

cnt

20.

187

0.19

92

0.93

90.

348

20.

102

AE

ˆsufficnt

20.677

0.230

22.948

0.003

20.458

VE

ˆsu

fficn

t2

0.22

00.

185

21.

191

0.23

32

0.17

7C

Eˆ

suffi

cnt

20.

204

0.25

12

0.81

30.

416

20.

111

RC

ˆcharlocc

0.220

0.062

3.528

***

0.221

AC

ˆch

arlo

cc0.

037

0.07

20.

520

0.60

30.

037

VC

ˆcharlocc

0.189

0.084

2.239

0.025

0.140

CC

ˆch

arlo

cc0.

099

0.05

31.

880

0.06

00.

149

RE

ˆcharlocc

0.528

0.195

2.707

0.007

0.189

AE

ˆch

arlo

cc2

0.40

60.

217

21.

868

0.06

22

0.18

0V

Eˆ

char

locc

0.19

10.

177

1.07

80.

281

0.10

1C

Eˆ

char

locc

20.

090

0.24

32

.370

0.71

22

0.03

2

Note:

Sig

nifi

can

tre

lati

onsh

ips

init

alic

s

Table AI.

PR34,1

108


RC 0.470

AC 0.523

VC 0.408

CC 0.198

RE 0.417

AE 0.704

VE 0.152

CE 0.321Table AII.

Explained variance

Model NPAR CMIN DF P CMIN/DF

Default model 158 1,353.819 471 0.000 2.874Saturated model 629 0.000 0Independence model 34 4,569.601 595 0.0000 7.680

Table AIII.Model fit

summary – CMIN

Estimate ofregression weights S.E. C.R. P

Estimate of standardisedregression weights

GEN ˆ hof_sum 0.001 0.003 0.472 0.637 0.035COMM ˆ hof_sum 20.002 0.002 20.780 0.435 20.052GEN ˆ satprev 0.084 0.110 0.767 0.443 0.076COMM ˆ satprev 20.100 0.100 20.999 0.318 20.090GEN ˆ freedom 0.222 0.071 30.107 0.002 0.233COMM ˆ freedom 0.183 0.065 2.801 0.005 0.191GEN ˆ area 0.010 0.077 0.126 0.900 0.010COMM ˆ area 20.032 0.070 20.453 0.651 20.031GEN ˆ SKILL 0.131 0.097 1.351 0.177 0.167COMM ˆ SKILL 0.330 0.091 3.608 *** 0.416GEN ˆ sufficnt 0.027 0.082 0.328 0.743 0.040COMM ˆ sufficnt 0.054 0.074 0.731 0.465 0.080GEN ˆ charlocc 20.043 0.079 20.546 0.585 20.042COMM ˆ charlocc 0.161 0.073 2.214 0.027 0.156

Note: Significant relationships in italicsTable AIV.

Regression weights

GEN 0.099COMM 0.314

Table AV.Explained variance

Model NPAR CMIN DF P CMIN/DF

Default model 65 318.604 105 0.000 3.034Saturated model 170 0.000 0Independence model 17 2168.520 153 0.000 14.173

Table AVI.Model fit

summary – CMIN


adaptation

109


Facets and Dimensions of Cross-Cultural Adaptation - Ashridge

Documents

Transcript of Facets and Dimensions of Cross-Cultural Adaptation - Ashridge