Going Global Practical Applications and Recommendations for HR and OD Professionals in the Global Workplace J-B SIOP Professional Practice Series by Kyle Lundby, Jeffrey Jolton and Allen I. Kraut_8 ppt

1 Write the test in English This is the global language of business 2 Translate the test into the local language verbatim Usually with the help of a local HR professional or translation

being used, often with the wrong clients, in the wrong way, and

in inappropriate settings, it is worth assessing the current state ofthe art in developing tests

In a personal conversation with one of the authors in March

2009, a seasoned test development expert for an MNC testing firmdescribed the steps he takes to develop a test, along with some

of the pitfalls he faces The following passage is based on thisdiscussion (S Keely, personal communication, March 22, 2009).Typically the first step in developing a new test is to determine the goal Often that will be to develop a test for an MNC that wants to select 20 entry-level managers in 14 different countries These managers should be able to eventually advance in the company, possibly even out of the country, so they should be relatively equivalent to each other when selected Making the test equivalent across cultures is the challenge, so the following steps should be considered:

1 Write the test in English (the MNC is based in an speaking country, but as English is spoken by more managers across the world than any other language, this is the norm anyway).

English-2 Translate the test into the local languages for each country, verbatim.

3 Revise the test to fit the local culture with the aid of an HR sional and a psychologist from that culture Since there are many differences between languages and cultures, there are always problems here First, some terms do not exist in some languages Even figuring out the Mandarin characters for words like ‘‘exec- utive coaching’’ was difficult a few years ago when the term was not common Also in China, mentioning a personal checking account in a numerical test will confuse people, so this terminol- ogy must be avoided Second, some concepts are simply under- stood differently across cultures For example, there is no such thing as a ‘‘good loser’’ in Italian—a loser is a loser Thus, a scale measuring competitiveness may need revision to reflect that.

profes-4 Back-translate the test Using a bilingual (and ideally bicultural) translator, translate the test back into English This is best done with a local psychologist, but one different from the one used in the previous step to ensure there is no bias As there are fluent English speakers available in most countries, including those who have been educated in English-speaking countries, this is not a

problem Usually the search for the appropriate person begins at

a local university’s psychology department.

5 Obtain norms for each country This can also be done in

conjunction with the local university psychology department Standardizing the test and producing means and standard deviations is done in this step Note that just because an ethnic group in two countries has the same language (and dialect) does not mean that they will get the same norms For example, Chinese people in Mainland China tend to do better on

analytical reasoning tests than do Chinese outside Mainland China Norming is thus done for each country separately Cheung (2004) echoes the necessity for doing this step, noting that ‘‘If interpreted directly according to the original norms, test scores of Asian respondents [on personality tests] may be misjudged to be deviant’’ (p 180).

6 Validate the test for that population This step is not always done Most local governments do not require it and it is costly and time consuming One way to get around this is to do concurrent valid- ity studies These are easier and less resource intensive for the organization and can provide acceptable approximations for pre- dictive validity.

7 Do equivalence studies It is nearly impossible to have the same cut-off scores for two different countries on the same test How- ever, doing this for validity generalization to establish a worldwide norm sometimes works The goal is to get the group to a reason- able number for selection purposes.

These steps are summarized in Table 6.2

Some of the problems faced when going through this processinclude:

1 Technical problems—the script requires special programming

on the computer (for example, it is read up and down versusside to side, or right to left, or it is created in characters orscript different from those of English)

2 Equivalence in dialects—is the MNC interested in CentralAmerican Spanish or Spanish spoken in Spain? Kuwaiti or SaudiArabian Arabic? Traditional or Simplified Chinese characters?Each dialect requires a different norm to be created as it cannot

be said that the speakers of that language are equivalent acrosscountries

Table 6.2 Steps for Developing Culturally Valid, Standardized

Selection Tests.

1 Write the test in English This is the global language of business

2 Translate the test into the

local language verbatim

Usually with the help of a local HR professional or translation vendor

3 Revise the test to fit the

local culture

Usually with the help of a local HR professional or psychology professor

4 Back translate the test Using a (different) local HR

professional or psychology professor

5 Obtain local norms for the

country

Often with the Psychology Department

of a local university or an assessment vendor with experience in that country

6 Validate the test for the

3 With senior managers and executives, testing is not always used,except for personality questionnaires

4 Verbal reasoning tests do not work when translated into ent languages; thus, this is not done

differ-Since in this company’s experience it cannot be said thatany test is equivalent to the same test translated into anotherlanguage, testing is considered to be done locally only For theMNC mentioned above, the company would try to establish cutoffsthat were meaningful but this would have to be done on acountry-by-country basis Thus, the MNC might not get completeequivalence if it was trying to get the top 20 scorers among alltest takers They may end up with five people each from threecountries, two from another country, and one each from threeothers Thus, among the 14 countries from which they wereselecting, the top 20 candidates may come from only 7 Butwith the difficulties of getting equivalence across borders, the

7 countries that were not represented may contain some of thebest candidates—they just did not do as well on the tests Thoughthere is no easy answer on how to address this problem, anothertool that may help is assessment centers.

Designing Fair Tools—Assessment Centers, Context

and Culture

As testing is generally used for lower-level managers or entry-levelcandidates, other tools are generally used for more senior can-didates Or, when testing is used, it is part of a larger process—often an assessment center An assessment center comprises

a number of different activities, usually including testing, aninterview, and various simulations The simulations may include ameeting with a direct report, a work group, a customer, or a boss,along with an ‘‘inbox’’ or series of e-mails and memos to dealwith as part of a case study (Fisher, Schoenfeldt, & Shaw, 2003;Gatewood & Feild, 2005)

The issues with transporting an assessment center across tural lines, largely involve the cueing for behaviors that either donot exist in the second culture or behaviors that do exist, butare evinced differently, as alluded to in the section above Forinstance, expecting negative feedback in a direct report meeting

cul-in a Western, cul-individualistic society is not out of the ordcul-inary Aneffective manager in that scenario will give the negative feedback,even if it is the first time she has met her ‘‘subordinate.’’ If sheshies away from doing that, it likely indicates a lack of managerialskills But to expect the same behavior in an Asian context, such

as in China, is to look for behavior that does not often appear

in real life One of the authors has gotten consistent feedbackwhile assessing hundreds of managers in Greater China that thefirst few times one meets a direct report (or just about anyoneelse for that matter) the focus is on getting to know that per-son and establishing a relationship, not changing their behavior

In fact, in such a setting, it would be impolite and unnatural toaddress shortcomings in behavior Thus, an assessment center that

is developed around Western ideals may not work as intended in

an Asian context and may actually select for those who would not

be effective in the local culture

Similarly, in a leaderless group discussion (LGD) mixing nic groups does not always work Having a few Thais among a group

eth-of Americans and Germans will likely result in the Westernersrunning the meeting and the Thais, from a more reserved andless masculine culture (Hofstede, 2001), being nearly invisible.This gives little data regarding the abilities of the Thais and wouldnot work when selecting for local positions in Thailand Moreover,even if the LGD were conducted with all Thai candidates, otherproblems would appear, such as deference to the highest-rankingperson in the room If the goal is to select for local leaders, theLGD would provide some useful information but might still maskthe capabilities of the lower-ranking people in the room At thesame time, it would not provide information on how the selectedlocal leaders would interact with those from headquarters in aWestern company

With an inbox (IB) simulation, the same problems may appear.With a reserved culture, few confrontations will occur and seriousproblems may not be addressed However, the most commonproblem is generally one of language Because the IB requires thecandidate to sort through a few dozen e-mails of varying lengthsand deal with the problems therein, all within one to one-and-a-half hours, if the IB is not in his local language, the candidatewill have problems even finishing it Again, this results in lessdata available for assessment and generally a lower score on theexercise And even if he does finish, the recommended actionswhich may work in an Asian culture may not make sense (andtherefore receive lower ratings) to Western raters It should benoted that in some cases, where it is obvious that the candidate hadproblems with the language, the IB can be judged on the quality

of work done, even if it is a small amount This can somewhatmitigate the issue of the lack of data

The interview may be problematic as well Assuming the guage ability of the candidate is good enough to answer thequestions, an unfamiliarity with the process may hinder him fromgiving useful answers to behavioral questions This will likely disap-pear as the use of behavioral interviews becomes more widespread.However, the problem around appropriate behaviors for the cul-ture may still remain; because the questions are written and scored

lan-by Westerners, behaviors that are proper in the local culture may

be scored lower than more Western-appropriate ones

Other cultural issues may also have a negative impact on theassessment center process One is power distance (Hofstede, 2004)which can be roughly thought of as the amount of hierarchy in acultural group In cultures where there is a high power distance,one might find that candidates are especially hard on their directreports and more subservient to their superiors This will obviouslyresult in skewed findings in direct report meetings, IBs, and bossmeetings when compared to Western standards

Context and culture When companies enter emerging markets,

often local selection systems based on research studies are notreadily available One possible reason for a lack of research-based systems lies in the culture itself, namely an affinity fornepotism and filial or tribal loyalty (Al-Aiban & Pierce, 1993;Brand & Slater, 2003; Common, 2008) A nepotistic system wouldinvalidate selection devices such as interviews, personality tests,and even skill-based tests because the results cannot be compared

to meaningful performance when promotions are not based onmerit This is a particular problem in the Middle East and Africa

As an example of the power of this system, consider the U.S.troop surge into Iraq in late 2007 The surge may not have beensuccessful had the army not realized the importance of nepotism

in Iraqi tribal culture Specifically, before and after the surge, U.S.General David Petraeus implemented a policy of offering workcontracts to tribal elders (Sheikhs) in exchange for cooperationagainst Al Qaida in Iraq (Woodward, 2008) What was differentabout this system was the use of local tribal Sheikhs instead ofthe centralized, merit-based bidding system used in the West Thearmy had realized that Sheikhs would only work with the UnitedStates if doing so could increase the Sheikhs’ own personal power,which they measured by the number of followers they couldattract The work contracts allowed leaders to attract and hireadditional followers within their own tribe, solidifying their localstatus, and increasing the appeal of working with the United States

In contrast, nonmembers of the tribe were provided with little inthe way of job offerings, special assignments, promotions, andsubcontracts Use of selection systems in this type of environmentmay not even be tolerated, much less taken seriously Trying toforce a merit-based selection system into a culture such as this

would result in, at best, a total lack of interest in applying for thejobs among candidates.

This type of tribal loyalty extends into major corporations

as well, including a well-known Middle Eastern energy companywhere two of the authors consulted There, a multiyear attempt

to replace tribal and familial nepotism with a meritocracy hasbeen undertaken Previously workers rose within the organizationbased on a combination of personal influence and the influence

of individuals to whom they were related directly or via tribalaffiliation Replacing such a system with tests and other measures ofmerit is no small task Administrators are literally asking individuals

to change cultural norms and expectations while they are at work,then return to them when they make their way home for thenight Clearly this effort can impact interpersonal relationshipsafter work, which is a large obstacle to success in the project.Another aspect of culture which may impact selection is level

of context According to Hall (1976) cultures run on a continuumbetween high and low context (see Table 6.3) This categorization

of cultures helps to determine how people relate to one another ondimensions such as social orientation, commitment, responsibility,confrontation, communication, and dealing with new situations.Social bonds refer to how deeply involved people are with eachother High-context cultures promote social bonds that implycommitment, expectations, good will, conformity to group norms,and greater distinctions between in-groups and out-groups (Hall,

1976, in Kim, Pan, & Park, 1998) Commitment is the degree towhich people do as they say in a culture People in high-contextcultures consider their word to be their bond and therefore arevery reluctant to give it freely (Kim, Pan, & Park, 1998) However,when they do give their word, it is, for all practical purposes,

as good as a written contract in Western culture Responsibilityrefers to how hierarchical and centralized decision making is in aculture High-context cultures see responsibility as being held atthe top, where subordinate errors are blamed on those who are

in charge In low-context cultures, by contrast, responsibility anddecision making are diffused and therefore so is accountability forerrors (Kim, Pan, & Park, 1998)

Confrontation may be the most complex aspect of culturalcontext Due to the strength and intimacy of the bonds betweenpeople in high-context cultures, confrontation is avoided, as

alluded to above Kim, Pan, & Park (1998, p 511) indicate that

‘‘people are more likely to repress self feelings and interests to maintain harmony, and there is a tendency to allow for con-

siderable bending of the system.’’ Additionally, to show emotionssuch as anger or disagreement is to lose control; this causes aloss of ‘‘face’’ which is directly related to reputation and honor(Kim, Pan, & Park, 1998) On the other hand, acknowledging thatsomething has happened between yourself and another person in

a high-context culture requires that action be taken, and ‘‘action

is very, very serious’’ (Kim, Pan, & Park, 1998, p 511) Similarly,communication relies upon different things in high-context andlow-context cultures In high-context cultures, messages are highlyeconomical and rarely contain all of the information necessary

to understand meaning Instead, meaning is obtained by placingthe statements in the context from which they were derived, such

as who the communication was for and who it came from (forexample, higher- or lower-status individuals) Again, though this

is economical in that short messages can communicate a lot, itdoes require a high level of ‘‘programming’’ to get to such apoint (Kim, Pan, & Park, 1998) Finally, people from low-contextcultures are used to the complexities of relying on context-freesystems and are therefore very creative even when dealing withnovel stimuli Individuals from high-context cultures work well atbeing creative within their contextual system, yet when confrontedwith a situation outside of that system they must create a new onebefore their innovation reaches its fullest potential

China, Korea, and Japan tend to be high-context cultures;Switzerland and Scandinavian countries such as Norway and Swe-den tend to be low-context cultures; and France, Spain, Africa,and the Middle Eastern countries all fall somewhere in the middle(Kim, Pan, & Park, 1998)

Cultural context and effect on selection When it comes to

paper-and-pencil job skill assessments there does not seem to be anykind of link between context and score Whether people arefrom a high- or low-context culture, they will likely interpretthe questions the same way on a given test, all other thingsbeing equal However, in assessment center situations, such asinterviews and the assessment of softer skills or aspects such

as personality, context may play a key role in how individuals

Table 6.3 High-Context Versus Low-Context Cultures.

Cultures (HCC)

Low-Context Cultures (LCC)

for decisions is held at top;

top-down decision-making

Responsibility for decisions is diffused

Subordinate errors blamed on top for HCC; in LCC everyone has accountability Confrontation Avoided given

intimacy of bonds between people

Done routinely

to ‘‘air grievances’’

Bending of system to ensure

no one loses face

in HCC Communication Need

programming, but can then be economical

Must be detailed and precise; anyone can understand

it regardless of culture

Those outside the HCC miss a lot of what is

communicated

Creativity Need a system

to be creative within; may have to create a new system

Potentially very creative even with novel stimuli

HCC may have to create a new system to be truly innovative

Source: Compiled from Hall, 1976, in Kim, Pan, & Park, 1998.

answer questions For example, should an interviewer ask a jobcandidate to critique a piece of work as part of a skills assessment,the lack of context in such a situation may lead the job candidate

to keep quiet for fear of appearing sassy, abrasive, or causing a loss

of face for a superior who may have done the work In simulations,the person from a high-context culture may view the situationfrom his own cultural context and respond accordingly Wherethe simulations are created in a low-context culture, then, theymay elicit the wrong cues For example, in the Middle Eastern

energy company mentioned above, one of the simulations usedwas a meeting with the candidate’s peer While the low-contextcompetencies were cued for in the simulation, typical responses

by the candidates were all around developing relationships In

a high-context culture, relationships are very important (Hall,1976) Moreover, since all of the assessors were Westerners, it islikely that many of the high-context cultural messages were missed.For both reasons, the use of that particular assessment processmight be questioned However, as discussed above, this projectwas intended to wrench the company into a flat world; hence,different cultural norms were considered more appropriate thanthe local norms As noted above, the use of local norms resulted

in the promotion of managers on the basis of affiliations ratherthan merit This, it was felt, would eventually cripple the companysince the senior management was seen as increasingly less capable(and through the nepotistic system, even lacking the incentive tobecome capable)

The promotion process may be where context and otheraspects of culture such as nepotism and familial or tribal loyaltyplay their most important role in selection systems Most Westernpromotion systems assume that performance in the current jobpredicts performance in the job to which a person is beingpromoted However, little research has been conducted on thisassumption and what has been done does not seem to support it(see Bernardin, 2009) In fact, in high-context cultures, individualswho hold their own feelings in check communicate in acceptableways that do not cause others to lose face (Kim, Pan & Park, 1998).This helps to avoid causing problems related to responsibilitybecause that person is able to avoid making errors and conform

to group requirements As a result, these individuals may be morelikely to be rewarded for their loyalty than others who do not act inthis way This has little to do with merit but everything to do withwhat is prized in a good (that is, docile) worker In fact, it usuallyresults in little challenging of the system, because avoiding errorsand not speaking up against bad ideas from others does not allowfor it In addition, in cultures where nepotism and familial ortribal loyalty are high, it is even less important how well a persondoes in their current job so long as they are properly connected atthe next level and do what they need to do to fit in (Brett, 2007)

Thus, the entire concept of merit may be turned on its head.

A high-context assessment center may be better off measuringwho ‘‘gets’’ the cultural mores most effectively to ensure thosecandidates are promoted However, as Friedman (2005) suggests,those cultures may also be left behind in the flattening world:

‘‘What is the motto of the tribalist? ‘Me and my brother against mycousin; me, my brother and my cousin against the outsider.’ And

what is the motto of the globalist ? ‘Me and my brother and my

cousin, three friends from childhood, four people in Australia, two

in Beijing all make up a global supply chain’’’ (p 326) This is

an argument for merit-based assessment, regardless of which localculture is involved As such, it is a step toward equal measurement

of TN capabilities

What to Do?

To ensure that cultural factors do not negatively affect assessmentresults, there are a number of steps that can be taken First,there must be a competency modeling project completed toclarify what is to be measured, both for the position in the homecountry and the same position to be filled in the host country.Where there are differences, there should be a conversation

by stakeholders about what is important for the host countryposition and what the expectations are for the person who fillsthe position For instance, is that person expected to stay in-country or move out of the country as her career progresses? Asnoted above, this will have implications for the knowledge andskills needed Including the requisite competencies to select TNsversus expats versus domestic employees is therefore important.Currently researchers are examining the particular competenciesthat should be assessed, usually just regarding expats (see Harzing,2004), but at present there is no agreement on exactly what thosecompetencies should be

Second, when there is agreement on what the position shouldrequire, the assessment should be designed or modified to reflectthose needs If the needs are for a TN to take on the position,then a process that reflects the corporate culture rather than thelocal culture should be used If the person is expected to remainin-country for most of her career, then the process should bedesigned around and reflect the local culture

Third, once the process is modified to fit the aims of thecenter, it is necessary to pilot it Comments should be solicited

on everything, including such details as the names for the players

in the cases One way to do this is to select some internal stafffrom the relevant culture and request input from them on thesimulations and interview One manager from Shell for example,says that in that company, they use focus groups with which to tryout a new exercise or process (Sparrow et al., 2004)

Fourth, the assessment process should be validated This can

be done by examining results of the center for individuals againstthe later progress of those individuals in the company Of course,this is not an ideal solution because in the vast majority of cases,companies use the data to make selection decisions (as is thegoal here) Thus, the data used to validate the center is clearlybiased and thus not entirely useful (Gatewood & Feild, 2005).Most companies, moreover, will not put the data aside for a fewyears to determine the validity of the center; such action would betoo costly and would not help the organizations make selectiondecisions, which again is the point of the exercise Nonetheless, avalidation effort should be attempted

Fifth, a consistent reevaluation of the assessment processshould take place periodically Even national cultures now arechanging as globalization gathers steam, and the needs of com-panies often change as quickly It is imperative that the processmeasure accurately and according to the needs of the organi-zation To risk assessing the wrong person for the wrong job

in the wrong culture may be extremely expensive, not only incash, but also in negative publicity and the loss of a valuableresource— namely, some of the organization’s talent

Finally, it is possible that as national cultures change, ple are even choosing their own culture (Tipton, 2009) Thismay complicate the development of tests and assessment centersbut also may make it easier If companies can create a culturethat draws people in, perhaps like those of some of the dot-coms, then the assessment tools can simply be made to reflectthe company culture and a person’s fit with that culture, ratherthan having to determine fit with the local and corporate cul-tures, along with any other cultures in which a TN might findhimself

In this increasingly globalized world, it is becoming more and moreimportant to find, hire, and develop people who can move betweencultures freely and easily while remaining productive and advanc-ing the organization’s goals Although many researchers havespent significant time exploring the issues surrounding expats,another type of corporate nomad, the transnational (TN) hasbecome a growing part of the global workforce, and increasinglyimportant to MNCs Thus, TNs must be considered when deciding

on which selection tools to develop and how to develop them Thegrowing numbers of both TNs and expats, moreover, requires thedevelopment of tools including testing and assessment centersthat are culture appropriate and fair We have attempted to show

in this chapter that these problems are neither gigantic in sizenor Lilliputian in scope Instead, they simply require a logical,

if more complex, consideration of the variables involved and awillingness to step out of one’s own culture to consider the morelong-term goals of the organization and of the individuals withinthat organization

References

Adler, N J., & Bartholomew, S (1992) Managing globally competent

people The Academy of Management Executive, 6(3), 52–65.

Al-Aiban, K M., & Pearce, J L (1993) The influence of values on

management practices International Studies of Management and nization, 23(3), 35–52.

Orga-Anderson, N., Lievens, F., van Dam, K & Ryan, A M (2004) Future spectives on employee selection: Key directions for future research

per-and practice Applied Psychology: An International Review, 53(4),

487–501.

Bartram, D (2001) The development of international guidelines on

test use: The International Test Commission project International Journal of Testing , 1(1), 33–54.

Bernardin, H J (2009) Human resource management: An experiential approach (5th ed.) New York: McGraw-Hill/Irwin.

Brand, V., & Slater, A (2003) Using a qualitative approach to gain insights into the business ethics experiences of Australian mangers

in China Journal of Business Ethics, 45(3), 167–182.

Brett, J M (2007) Negotiating globally: How to negotiate deals, resolve disputes, and make decisions across cultural boundaries (2nd ed.) San Francisco:

Wiley.

Briscoe, D R., & Schuler, R S (1995) International human resource management (2nd ed.) New York: Routledge.

Brookfield Global Relocation Services, National Foreign Trade Council

(2008) Global relocation trends: 2008 survey report Brookfield Global

Relocation Services, LLC, Woodridge, IL.

Caligiuri, P M (2000) Selecting expatriates for personality tics: A moderating effect of personality on the relationship between

characteris-host national contact and cross-cultural adjustment Management International Review, 40, 61–81.

Caligiuri, P., & Paul, K B (2010) Selection in multinational tions (Chapter 34) Refereed book chapter in James L Farr and

organiza-Nancy T Tippins (Eds.), Handbook of employee selection Hillsdale,

Cerdin, J L., & Bird, A (2008) Careers in a global context In Michael

M Harris (Ed.), Handbook of research in international human resource management Hillsdale, NJ: Erlbaum.

Cheung, F (2004) Use of western and indigenously developed

person-ality tests in Asia Applied Psychology: An International Review, 53(2),

173–191.

Chua-Eoan, H (2007) Crimes of the century: The collapse of Barings Bank, 1995 Time.com Retrieved August 8, 2009 from www.time com/time/2007/crimes/18.html.

Common, R (2008) Administrative change in the Gulf: Modernization

in Bahrain and Oman International Review of Administrative Sciences, 74(2), 177–193.

Enron scandal at-a-glance (2002) BBC Retrieved on August 8, 2009, from http://news.bbc.co.uk/2/hi/business/1780075.stm.

Fisher, C D., Schoenfeldt, L F., & Shaw, J B (2003) Human resource management (5th ed.) Boston: Houghton Mifflin.

Forster, N (2000) The myth of the ‘‘international’’ manager The International Journal of Human Resource Management, 11(1), 126–142.

Forster, N (1994) The forgotten employees? The experiences of

expatri-ate staff returning to the UK International Journal of Human Resource Management, 5(4), 405–27.