A Framework to Study Level of Comfort between Employees of Local and Foreign Cultures in Multinational Firms

Geert Hofstede studied the cultural dimension in multinational firms which play significant role in the professional conduct of employees. Fifth dimension was supplemented on a cross cultural study of Chinese Confucius behavior. In the changing global cross cultural conduct at work place, the firms are facing more challenges than ever before in forming and performing in multi-location and multi-national teams. This study attempts to read the variables to build a framework which measures the Level of Comfort (LoC) between local culture and foreign culture in the multinational firms. The framework developed is also tested on a pilot study with 200 respondents from 10 countries. Results show that Level of Comfort can be measured through a structured questionnaire and also that this level of comfort vary among countries included in the study.


Introduction
Comfort of local cultures with foreign cultures has become an important area to understand for effectively managing culturally diverse workforce at MNEs.Estimating the level of comfort of the local cultures with colleagues from alien cultures is a qualitative attribute that can be conveniently assessed by the use of a cross cultural questionnaire.However there are no available qualified research on the structure and content of a questionnaire which can truly gauge the level of comfort of local cultures with foreign cultures especially in a multinational work place setting.Therefore a need is felt to propose a suitable questionnaire for addressing the need to understand the dynamics of comfort of the local cultures with foreign cultures.
Though questionnaires are the most widely used tools especially in large surveys due to the low cost and their ease to apply, however, are subjected to recall bias, thus validation studies should be conducted in order to determine their validity and reliability.
The aim of this research study is to propose the contents, structure, validation and reliability of the questionnaire for estimating the level of comfort of the local cultures with foreign cultures in cross cultural communication in multinational companies.

Literature Review
A small number of researches have been done referring to enquiry of the level of comfort of local cultures with foreign cultures.And therefore a 'gold standard' questionnaire which can effectively record such level is not available.However a few attempts to enquire into similar phenomenon are referred below

Culture and Cultural Comfort
Culture has been defined in several ways.There are literally hundreds of different definitions of culture as writers have attempted to provide the all-encompassing definitions of this most debated concept.Undoubtedly culture consists of language, ideas, beliefs, customs, taboos, codes, institutions, tools, techniques, works of art, rituals, ceremonies and symbols among others.It has played an important role in human evolution, allowing humans to adapt to the environment to their own liking rather than depend exclusively on natural selection to achieve adaptive success.Every human society has its own particular culture, or socio-cultural system.(Adapted A useful questionnaire was proposed in a research on 'Attitude change after taking a Virtual Global Understanding Course' (Chia, Rosina C. et al, 2009).In this research, a virtual collaborative classroom was formed at East Carolina University (ECU), using videoconference technology via regular internet to bring students from 18 different countries, 2 at a time, to the ECU classroom to learn about each other's culture.Students from two countries were partnered one on one.They met for 4-5 weeks, and subsequently submitted a joint paper.Then the same process was repeated for two other countries.Lectures and student discussions were managed with pre-determined topics and questions.Classes were conducted in English and reading assignments were placed on the website.Familiarity with technology, knowledge in cultural understanding and attitude change were assessed.After taking this course, all students stated their comfort level in working with, and their desire to interact with, culturally different others grew stronger and their xenophobia and isolationist attitudes decreased.The questionnaire which was used to record the comfort level of participants included questions related to topics like stereotype, xenophobia, isolationalism, disinterest in knowing about culturally different others, lack of desire to interact with other cultures, comfort with culturally different others and similar topics.This research questionnaire was somewhat close to estimation of comfort level between diverse cultures but was not exhaustive.

Cross-Cultural Interaction Comfort and Service Evaluation
Another notable research related to cross cultural comfort is 'Cross-Cultural Interaction Comfort and Service Evaluation' (Paswan, Audhesh K. et al, 2005).This study empirically investigates the notion of being comfortable with social interaction in a foreign country (interaction comfort) and its impact on how service quality is evaluated.In addition, two of the antecedents of this interaction comfort are also investigated.The results indicate that the home country social class of the respondents is positively associated with interaction comfort.In addition, similarities or differences between one's home and one's host country during travel also influences this cross cultural interaction comfort.The results also indicate that visitors high on interaction comfort are more satisfied with the service offered the host country and are willing to advocate it to others back home.Apart from some interesting research implications, the findings have several managerial ones, especially in the context of services aimed at the international market.

Research Methodology
A theoretical structure of the questionnaire was arrived at with the help of 100 volunteers using Delphi technique.The volunteers consisted of cross cultural experts and students of cross cultural courses.The volunteers were grouped in 10 groups.The Delphi sessions were performed in 2 stages.

Stage 1 -Identification of factors (Scales)
Each group was asked to identify most important factors which are likely to explain the difference among cultures which describe the level of their comfort with foreign cultures.Each group was asked to identify around 15 such factors which explain the objective of the questionnaire along with their explanation of reasons of choices.The group recommendations were read out to all the groups anonymously along with the reason of the choice of the factors and how they explain the phenomenon of the research enquiry.The groups were then asked to modify their choice of factors.In around 10 rounds each group unanimously came out with a common set of 13 factors (components) (2 conditional factors and 11 general factors).

Stage 2 -Identification of questions, as items of Sub -scales (components / factors)
In stage 2, each group was asked to identify a set of most appropriate questions, which represented the identified CFC factors.Each group was also asked to share the reasons of their choice.After each round the results of each group were anonymously read out to all volunteer groups along with the reasons of choice of questions by each group.The groups were then asked to rethink on their choices after every round.In around 12 rounds, each groups agreed on a selected set of questions which represented each identified factors.
The theoretical structure which came out of the above 2 stages of Delphi sessions looked similar to the structure as given in table 1.
In the next step the above model was retested empirically through applying the above proposed questionnaire to a random sample of 200 MNC employees from more than 10 regions.Study Population: The sample size was 200 from more than 10 regions, 84 of whom were men and 116 were women, each working for companies having operations in different countries.
All respondents who met the inclusion criteria of the study were working for multinational companies in their respective countries of residence.The inclusion criteria for the study were -all respondent must be well educated, exposed to multi cultural working environment in a multi country organization, and routinely working and interacting with colleagues with origin from several countries.
The method for the data collection was 'filling out online questionnaire', using the above CFC questionnaire, which is supposed to have fully met the requirements of the current study and consisted of the questions of "assessment of the level of comfort of the local cultures with foreign cultures at their work place" and descriptive characteristics.
The likert type five point scale was used to answer all questions.The five different scales were represented with the following answers: Strongly Agree, Agree, Neutral, Disagree and Strongly Disagree.The process of filling out the questionnaires usually took between 15 and 30 minutes each.
The results of the questionnaire survey were tabulated and means were compared for each sub scales.Based on the means country ranks were given on each subscales.These ranks were used to calculate overall rank scores for each country which gives a comparative estimate of each country's level of comfort with foreign cultures

Reliability and Validity Assessment of the Questionnaire
Validity: Initially, the questionnaire, immediately after the design, was submitted to 20 respondents to determine whether the questions were clear, understandable, and in a logical order (face validity).The construct validity of the questionnaire was tested using the appropriate statistical technique in order to determine the structure of the questionnaire, namely the number of sub-scales measured by this questionnaire, and the questions (items) that constitute each sub-scale.
The criterion validity of the questionnaire was not checked, as a gold standard tool for assessment of the level of comfort of local cultures with foreign cultures in Multinational Companies has not been proposed yet.
Reliability: Finally, the internal consistency and the repeatability of each sub-scale derived from the construct validity of the questionnaire were tested using the appropriate statistical tests.In particular, 50 respondents were used in order to assess the repeatability.These respondents completed the questionnaire two times.Between the two measurements there was a period of three to four days.

Base line characteristics:
A number of socio-demographic characteristics of respondents who participated were collected.More specifically, following were recorded:  Sex  Age of the respondents  Educational level  Residence status in their country of stay  Their relative income level

Statistical Analysis
Evaluation of the validity and reliability of questionnaire: The factor analysis was used to evaluate the construct validity of the CFC questionnaire.The suitability of the data for carrying out such analysis was tested by using the Bartlett sphericity test (see table 5) and the Kaiser-Meyer-Olkin (KMO) statistic test (see table 5B) that evaluates the degree of correlation among the questions included in the questionnaire.
Statistically significant results of the sphericity test indicates that the variance-covariance matrix of the initial questions of the questionnaire is not diagonal (i.e., variables are correlated with each other).Also, values of KMO> 0.8 indicate a fairly high correlation and therefore, factor analysis is meaningful.The method of "Maximum likelihood" was used for assessment / extraction of the main factors (Olsson, U., 1979).The criterion of Kaiser (eigenvalue > 1) was used to determine the number of factors derived from the factor analysis.An orthogonal rotation (in this case Varimax) was used to improve the explanatory ability of the factors.Each factor that emerged was interpreted based on the questions that had load value > 0.3.Evaluation of the internal consistency of the sub-scales for the CFC questionnaire was carried out by calculating the Standardized Cronbach Alpha coefficient and calculating reliability estimates.Cronbach coefficient ranges from 0 -1.Large Cronbach Alpha values indicate a high consistency of the questions of which the sub-scale is consisted.The "Cronbach Alpha if item deleted" index was used to identify the questions that reduced the internal consistency of the questionnaire and therefore had to be excluded.The repeatability of questionnaire was evaluated by using McDonald Omega which is a measure of generalizability of the test items.It varies between 0 to 1. Omega can be interpreted as the square of the correlation between the scale score and the latent variable common to all the indicators in the infinite universe of indicators of which the scale indicators are a subset (McDonald, 1999, page 89).
The greatest lower bound was calculated for reliability testing.The greatest lower bound (glb) to reliability represents the smallest reliability possible given observed covariance matrix under the restriction that the sum of error variances is maximized for errors that correlate 0 with other variables (Ten Berge, Snijders, & Zegers, 1981).
All statistical analysis was carried out using SPSS and FACTOR PROGRAM developed by Dr. Urbano Lorenzo-Seva & Dr. Pere Joan Ferrando, (2012).

Calculating the Difference on Level of Comfort with Foreign Cultures
The questionnaire as validated as per the above method was used to estimate differences among 10 countries by comparing the means of each country on different sub scales.

Descriptive Results
Table 2 presents the baseline characteristics of the participants from 10 regions (N=200).It was observed that the majority of the participants were in the age group of 30 to 50 years, mostly having post graduate education.

Mostly Post Graduate
Av & Above Av

58%F/42%M
The majority of participants had average or above average income level.Regarding area of residence, most respondents were natives of the country (90.0%)where they were interviewed.Regarding distribution by sex, the percentage of female participants (58.0%) was significantly higher than the men's percentage (42.0%).However country wise sex ratios differed significantly (see table 2).Regarding the socio economic status of sexes, male's percentage in the higher income group (28%) was significantly higher than females (16%).
Distribution of participant baseline characteristics is summarized in table 2. It can be seen that country wise distribution of baseline characteristics is not very significantly different from one country to another, indicating the validity of the sample design.

Statistic Results
Regarding face and content validity both participants and experts reported that the questions asked in the questionnaire were clear, easy to understand, in a logical order, and totally representative of the needs of the objectives of the research.

Construct Validity of the Questionnaire
As it is already mentioned, factor analysis was performed for the evaluation of the construct validity of the CFC questionnaire.Initially, the relevance of the data used for the factor analysis was confirmed.The statistical criteria Kaiser-Meyer-Olkin (KMO=0,948) and the Bartlett Test of sphericity (value 15944.3(df = 1275; P = 0.000010)), indicated that the raw data were suitable for the implementation of factor analysis.

(See table 5 B)
From the 50 original questions used in the factor analysis in general 11 sub -scales as proposed in Delphi session, 40 were confirmed, using the Kaiser criterion and Varimax orthogonal rotation.10 items were removed from the original version of the questionnaire due to lack of variance and inconsistency in data.See table 3, to see the list of questions removed from the original suggested questionnaire. 2 conditional sub -scales were retained as explained above.5A shows the detailed results of the evaluation of the internal consistency and reliability for each sub-scale (factor) of the CFC questionnaire, as those derived from the factor analysis.Table 5A shows that the internal consistency of the questions making up each sub-scale is reasonably high (Most cases> 0.7) except factors 6 and 8. However for subscales values of reliability greater than 0.6 is normally acceptable.The greatest lower bound for reliability is also calculated and has been found to be suitable.(See table 5B) Genralizability and repeatability of the questionnaire items has been tested using McDonald Omega which is very good as shown in table 5B.Also shown are the indices for factor simplicity.

Estimate of Differences among Countries with Respect to Comfort with Foreign Culturess
Following table gives the differences among countries on different sub scales incorporated in the CFC questionnaire and their ranks are also tabulated.Using these ranks total rank score is calculated for each country as given in table.The final rank scores for each country are given.Convergence value : 0.00001000

Discussions and Conclusions
The results support the proposition that above contents and structure of questionnaire proposed may be fairly appropriate for a large scale study on the comfort with foreign cultures (CFC) in multinational enterprises and culturally diverse workplaces around the world.

Limitations of the Above Results
While reliability of the above data is quite acceptable in view of several tests done, in the above study, Greatest Lower Bound (GLB) and Omega may not indicate reliability conclusively since GLB and Omega can only be trusted in even large samples, preferably 1,000 cases or more, due to a positive sampling bias (Ten Berge & Socan, 2004).The above study could cover only 606 respondents sue to limitation of resources, therefore a more detailed study involving more than 1000 cases may be more reliable for above testing.Similarly the items of the questionnaire may not represent complete picture of the enquiry and may be subject to change with changing cross cultural environment with time.The above questionnaire therefore may need to be relooked after a reasonable time interval of say every 5 years.

Table 1 .
Proposed CFC questionnaire, suggested by Delphi sessions

SS4: Effect of Personal Comfort
Q36: No Global problems due to religious beliefs I do not think there are problems in modern world which relates to religious beliefs SS8: Effect of Scapegoating Q37: Problem may not increase with foreigners Probability of being branded as accused would not increase if the innocent persons also happen to be foreigners Q38: No Victimization of foreigners based on their religious beliefs I have never witnessed incidents involving a person of foreign origin persecuted or victimized for his foreign religious belief in my society Q39: Unrelated person never getting randomly targeted There are no insecurities related to sudden turn of events which may result into an unrelated person getting targeted by the society (where I live in) at large Q40: Rational Society for foreigners The society (I live in) remains rational in a situation of major crisis involving those which may apparently look like to have been created by certain groups or persons of certain foreign origin or race Q41: Irrational behavior of victims blaming a foreigner In a situation of certain crisis have you witnessed victims behaving irrationally accusing certain group of persons based on their nationality or race?Q42: Branded accused without proof Do you think there is a possibility of innocent persons been branded as accused without enough proof in your society SS9:

Effect of liberal society
Q48: No problem to understand foreigners When dealing with persons of foreign origin I never had difficulty in understanding his or her point of view Q49: No Repulsion with people of other religion It is not difficult for me to feel close to people who have a different religion from mine Q50: No difficulty in understanding of diverse world cultures I never have any difficulty in understanding of diverse world cultures PART 2: (Conditional Sub -scales)

Table 2 .
Baseline characteristics of region based respondents

Table 3 .
List of items removed after empirical testing

Table 4
show the factor analysis which confirms the reliability of the questionnaire based on structure of 11 general factors as suggested by the Delphi Sessions.The overall rate of variation of initial data, the 11 factors (Sub-Scales) interpreted in common amounts to 65 % for overall sample of 200 respondents.

Table 5A .
Explained variance and reliability of rotated factors

Table 5B .
Other statistics for internal consistency of the CFC questionnaire

Table 5C .
Summary of statistical analysis for overall sample (N=606)

Table 9 .
The main sub scales / components which can best describe the phenomenon of comfort of local cultures with foreign culture are as follows:The final CFC questionnaire proposed and tested by above study looks as given in table 6.Also given intable 7, 8 and 9is the rank scores of the countries studied in terms of the level of comfort of the local employees of MNCs with foreign employees.Different behavior with countrymen I would have behaved differently if the person would have been a fellow countryman Q65: Foreigner should be more careful Person should have been more careful in dealing with me, because of his/her foreign origin while being in my country Q66: Problem related to his origin Peculiar behavior of the person was related to his/her foreign origin Table 8.Overall ranking based on mean score averages Nigeria China Germany India Italy LA Portugal Sweden UK USA Overall national ranks (based on rank totals and overall mean)