Data Analysis
|
| < Day Day Up > |
|
Phase 1: Descriptive Statistics
Through the research survey, a total of 296 valid responses (usable response rate of 31.3%) were received from the sampled subjects, and were used for the data analysis. Twenty-eight variables with their abbreviations used in the research are in Tables 3 and 4. The 14 hypotheses used to test the objectives of the research aimed at measuring the cross-cultural influences on the IS managers' performances, job satisfaction and managerial values. Since the usable answers, 296, were larger than 30, according to the central limit theorem, the response distribution would be normal. Out of 236 American respondents, 63.6% were male, while out of 60 Korean IS expatriate managers, male response rate was 60%. In the education category, 28.3% respondents had a bachelor's degree in the U.S., while 20 out of 60 subjects had a bachelor's degree from Korea. In the age category, the majority of American IS managers were between 39 and 49 years old. On the other hand, the majority of Korean IS expatriate managers were between 29 and 39 years old. In the category of the total assets in the banks, 80.9% of American banks had over $50,000,000, whereas, 88.3% of Korean Banks were in that category. As shown in Tables 3 and 4, mean values of combined descriptive statistics for 28 variables of IS managers' value, job satisfaction and performance ranged from 2.97 to 4.52. Standard deviations (Std. Dev.) ranged from 0.56 to 1.05. Out of the mean value, "The feeling of achievement of your job performance (PERF)" for American IS managers was the variable that has the highest mean value with the next to the lowest standard deviation of 0.57, while the question item "The opportunity to develop close friendship (FRID)" has the lowest mean value. On the other hand, for the Korean IS expatriate managers, "The opportunity for personal growth and development (GROW)" was the variable that has the highest mean value, while the question item "The frequency of your expressing disagreement with supervisors (DISA)" has the lowest mean value with the lowest standard deviation of 0.72. These descriptive statistics are representing some differences between the two groups' characteristics. For the American IS managers, the mode for age is between 39 and 49 years and the average working years in the present position was about 7.5 years. On the other hand, the average age of Korean IS managers was 34 and the average work experience in that position was 3.5 years. That implies American IS managers' turn-over rate is smaller than that of Korean IS expatriate managers.
| Question Item (Abbreviation) | America Data (n = 236) | Korea Data (n = 60) | ||
|---|---|---|---|---|
| Mean | Std.Dev. | Mean | Std.Dev. | |
| 4.23 | 0.60 | 3.60 | 0.90 |
| 3.82 | 0.78 | 3.74 | 0.87 |
| 4.18 | 0.71 | 3.93 | 0.92 |
| 3.77 | 0.74 | 3.59 | 0.77 |
| 4.32 | 0.64 | 3.84 | 0.94 |
| 4.01 | 0.77 | 3.47 | 0.82 |
| 4.21 | 0.63 | 3.57 | 0.84 |
| 4.35 | 0.58 | 3.64 | 0.89 |
| 4.08 | 0.64 | 3.35 | 0.83 |
| 3.95 | 0.74 | 3.36 | 0.89 |
| 3.92 | 0.65 | 3.31 | 0.90 |
| 4.27 | 0.68 | 3.28 | 0.93 |
| 3.03 | 0.88 | 3.17 | 0.88 |
| 3.98 | 0.86 | 3.21 | 0.93 |
| Question Item (Abbreviation) | America Data (n = 236) | Korea Data (n = 60) | ||
|---|---|---|---|---|
| Mean | Std.Dev. | Mean | Std.Dev. | |
| 3.59 | 0.88 | 3.17 | 0.92 |
| 3.43 | 0.80 | 3.09 | 0.73 |
| 4.18 | 0.69 | 3.47 | 0.92 |
| 4.52 | 0.57 | 3.69 | 0.82 |
| 4.34 | 0.56 | 3.36 | 0.81 |
| 4.01 | 0.79 | 3.38 | 0.86 |
| 3.14 | 0.83 | 2.97 | 0.72 |
| 3.23 | 0.87 | 3.35 | 0.78 |
| 3.97 | 0.76 | 3.31 | 1.05 |
| 3.78 | 0.86 | 3.57 | 1.04 |
| 3.68 | 0.85 | 3.17 | 0.95 |
| 3.74 | 0.79 | 3.09 | 0.80 |
| 3.55 | 0.76 | 3.19 | 9.76 |
| 3.74 | 0.81 | 3.06 | 0.91 |
Phase 2: Reliability and Validity Test
As Kerlinger (1978) and Montazemi (1988) suggested, the reliability for each individual item in the questionnaire was calculated and the reliability coefficient for each item was above 0.80. Thus, this research instrument has proven to be reliable for further analysis without deleting any items from the questionnaire. Hair, Ghiselli, and Poter (1966) described that factor analysis can help to reveal the underlying dimensions which tend to be associated with the variables and to identify the grouping of variables which are most closely associated or aligned with those dimensions. Even though the two tables were not included, the correlations between pairs of 28 items showed that all the items were related statistically significant at a p-value < .05 level at almost all pairs of the variables in the instrument. Thus, the results have proven that this research instrument has internal consistency. Hair, Ghiselli, and Poter (1966) also explained that factor analysis based on multi-item scales is more reliable than measures based on the original individual items. As a result of that test, four factors were extracted for the American IS managers as significant factors as shown in Table 5. As a cut-off point of eigen value, only those factors with one or larger than one were extracted. For the America data, the underlying constructs of the factors are meaningfully labeled as Performance, Self-Actualization, Accomplishment, and Company-Policy; 43.84% of the total variance for America data was attributed to the four factors. On the other hand, Korea data showed that with the four factors, 61.17% of the total variances was explained. Each eigen value is a summary index of how much of the variance in the initial correlation matrix is accounted for by the associated factor. Thus, in conclusion, these four underlying constructs represent those underlying factors very well. Hair, Ghiselli, and Poter (1966) and Kerlinger (1978) explained that an absolute value of 0.30 or above for the loading factor of the factor analysis is regarded as significant for this kind of the research. Even though those two tables were not included, all the loading of the items was greater than the level of the minimum standard of 0.30 at least on one factor. The calculated reliability measures both internal reliability and consistency when using the same instrument to measure similar constructs. Since all items were reliable, elimination of items was not a necessary step in the study. Construct validity means that a measure is a valid measure of a construct if it relates to other measures of this construct and to measures of different constructs in the expected manner. Kerlinger (1978) explained that factor analysis is the most important of construct validity tools. In the Item-Factor Correlations analysis, using Varimax Rotated Factor Loading, for America data, Factor 1 was loaded significantly in the variables such as ESTM, SUPE, PERF, PROD, EFFO, ACCO, PART and INFO. For the Factor 2, these variables such as RULE, HELP, FRID, PERS, STRE, WORK, DISA, and TECH were significantly loaded. For Factor 3, distinctly loaded variables were PRES, SECU, ADVA, DEFI, METH, OUTS, and EARN. For the Factor 4, the significantly loaded variables were included FULF, AUTO, GROW, INDE, and VARI. For Korea data, Factor 1 was loaded significantly by variables such as AUTO, PERS, STRE, WORK, PERF, DISA, EARN, and ADVA. Factor 2 was loaded significantly in the variables such as HELP, PART, METH, INFO, FRID, SUPE, PROD, EFFO, and VARI. Factor 3 was loaded significantly in the variables such as ESTM, GROW, PRES, INDE, SECU, FULF, and ACCO. Finally, for Factor 4, OUTS, DEFI, RULE, and TECH are the ones that were loaded significantly. Thus, the findings are validating the 28 variables as measures of four distinct constructs for Korean IS and American IS managers. To measure the predictive validity or criterion-related validity of the questionnaire, the three self-rated performance criterion variables were utilized; the test result showed that correlation of each item was highly correlated to the self-rated items.
| Factors* | America Data | Korea Data | ||
|---|---|---|---|---|
| Eigen Value | % of Variance | Eigen Value | % of Variance | |
| Factor 1 | 6.612 | 23.61 | 10.816 | 38.63 |
| Factor 2 | 2.344 | 8.37 | 2.468 | 8.81 |
| Factor 3 | 1.762 | 6.29 | 2.213 | 7.90 |
| Factor 4 | 1.558 | 5.56 | 1.629 | 5.82 |
| Total | 43.84 | 61.17 | ||
| Factors* | ||||
| Factor 1: Performance | ||||
| Factor 2: Self-Actualization | ||||
| Factor 3: Accomplishment | ||||
| Factor 4: Company-Policy | ||||
Phase 3: Regression Analysis and ANOVA Test
In this phase, a multiple regression procedure was utilized. This method generates an average measure of the degree of association (R2) among the criterion variables, IS managers' managerial values, job satisfaction and performances, and predictor variables within and across the two cultures. It is assumed that there is no interaction and correlation among the predictor variables. In this analysis, three mediated variables (organizationally, interpersonally, and internally) were used to predict the three criterion variables shown in Table 2. The coefficients resulting from standardized data of the regression analysis are called beta coefficients. The beta coefficient reflects the impact on the criterion variable of a change of a standard deviation in 10 independent variables from the three mediated variables. Then, beta coefficients were used as the relative importance of 10 individual independent items of the three mediated variables in this study. Separate regression equations were used to analyze the relationship between each dependent variable of the three criterions (value, job satisfaction and performance) with independent variables. All predictor variables were included sequentially into the equation based on the respective contribution of each predictor to explained variance in the dependent variables. For this purpose, as shown in Tables 6 and 7, the stepwise regression analysis was tested, and then predictor variables that significantly contributed were identified. For America data, PRES, VARI, EARN, HELP, and ADVA variables in IS managers' managerial value criterion showed significant F-value at the level of 0.003 or above. For job satisfaction criterion, HELP, ACCO, RULE, AUTO, SUPE, and FRID showed significant F-value at the 0.004 level or above. For performance, related variables also showed a significant level of F-value at the 0.040 level or above. On the other hand, in the value criterion, VARI and EARN showed negative signs which means that those two variables of the value criterion are the opposite of what was expected from American IS managers. In the job satisfaction criterion, HELP, ACCO, and AUTO showed a strong positive impact on job satisfaction, which was an expected result in this study. In the performance criterion variables, PART, VARI and DEFI variables also indicated that American IS managers in the banks do not desire or value too detailed a description of job definition. On the other hand, the results showed that they strongly desire to have INFO, and HELP. In other words, American IS managers would value helping and having a close relationship among IS managers and IS workers in the organization. For Korean IS expatriate managers in the surveyed banks, variables such as VARI and DEFI showed strong impacts on the criterion variable, managerial value. Thus, IS managers would value more defined rules and policies in the organization and more variation in their job environment. For the job satisfaction, PART, SUPE, and DISA were the variables, which have positive indication on the dependent variable. These predictor variables are significant at the 0.01 level of significance. On the other hand, ADVA, HELP, PROD, and PRES variables showed negative impact on the job satisfaction of Korean IS expatriate managers. Interestingly enough, literature supported the same result (Cole, 1979) too. For the performance criterion, PRID and HELP showed a mixed result, which was not expected in this study. One explanation may be that Korean IS managers who work in a western country like the U. S. have been changing their values so that they are becoming more individualized compared with those IS managers in Korea. FRID has a strong positive indication on performance, which is not expected in Korean IS managers because this factor usually has shown in the femininity society rather than that of Korea. In comparison of the two countries for the value criterion, predictor variables entered into the regression analysis have shown very similar results. Through the multivariate analysis of variance (ANOVA) shown in Table 8, dependent variables, value, job satisfaction and performances of the groups of two national IS managers were analyzed. This analysis can identify statistically significant differences by comparing variances due to the treatment effect and the measure of chance errors of the two national IS managers' cultural influences on the value, job satisfaction and performances. This analysis attempted to show what are the differences and similarities based on the convergence hypothesis and divergence hypothesis in cross-cultural studies. Three independent variables describing the criterion variable of value for Korea data have very significant F-values at the level of 0.0005 or 0.0001. For America data, all independent variables of value criterion were very significant at the 0.0003 or 0.0001 level. In addition, for job satisfaction criterion, all five variables were very significant at the 0.0005 or above level for Korea data. On the other hand, for America data, three of five variables for job satisfaction for both the Korea and America data were very significant statistically. Table 9 presents the ANOVA test of three different levels of independent variables belonging to organizationally mediated, interpersonally mediated and internally mediated variables. As expected and supported in the literature (Boyd, Ralph, & Stanley, 1977), Korea data in the organizationally mediated variable were not significant statistically. On the other hand, America data showed that all the variables were very significant at the 0.001 of significance. For the internally mediated variables, those variables for both countries' IS managers were equally significant, except one variable INDE for Korean IS managers, which indicates that Korean IS managers were not concerned with, nor considered INDE to be an important factor in their organization.
| America Data (n = 236) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Value | Job Satisfaction | Performance | |||||||||
| Variable | β Value | R2 After Entry | Prob. > F | Variable | β Value | R2 After Entry | Prob. > F | Variable | β Value | R2 After Entry | Prob. > F |
| PRES | .1251 | .0578 | .0012 | HELP | .2496 | .1020 | .0001 | PART | -.1346 | .0443 | .1033 |
| VARI | -.1453 | .1251 | .0031 | ACCO | .1809 | .1759 | .0044 | INFO | .2027 | .1027 | .0432 |
| HELP | .2681 | .2539 | .0019 | RULE | -.1247 | .2157 | .0037 | VARI | -.1516 | .1505 | .0247 |
| EARN | -.1087 | .2984 | .0006 | AUTO | .1435 | .2638 | .0021 | HELP | .2099 | .2060 | .0106 |
| ADVA | .1134 | .3286 | .0006 | SUPE | -.1442 | .3002 | .0016 | DEFI | -.1578 | .2556 | .0052 |
| FRID | -.1552 | .3349 | .0012 | ||||||||
| Korea Data (n = 60) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Value | Job Satisfaction | Performance | |||||||||
| Variable | β Value | R2 After Entry | Prob. > F | Variable | β Value | R2 After Entry | Prob. > F | Variable | β Value | R2 After Entry | Prob. > F |
| PRES | .1251 | .0578 | .0012 | ADVA | - | .0793 | .0307 | HELP | - | .1160 | .0089 |
| VARI | .2143 | .1481 | .0122 | PART | .1593 | .1433 | .0131 | PRID | .2296 | .1840 | .0039 |
| PRES | - | .1907 | .0092 | HELP | .1766 | .1783 | .0123 | .1763 | |||
| DEFI | .1773 | .2308 | .0068 | SUPE | - | .2113 | .0110 | ||||
| ADVA | .1483 | .0813 | .0300 | PROD | .1566 | .2463 | .0088 | ||||
| - | DISA | .1525 | .2908 | .0050 | |||||||
| .1644 | PRES | - | .3284 | .0034 | |||||||
| .2176 | |||||||||||
| .2233 | |||||||||||
| - | |||||||||||
| .1759 | |||||||||||
| Korea Data (n = 60) | America Data (n = 236) | ||||
|---|---|---|---|---|---|
| Variables | R2 | Prob. > F | Variables | R2 | Prob. > F |
| Value: | Value: | ||||
| 0.1498 | 0.0858 |
| 0.1878 | 0.0003 |
| 0.2356 | 0.0005 |
| 0.1398 | 0.0001 |
| 0.1503 | 0.0846 |
| 0.2812 | 0.0001 |
| 0.3780 | 0.0001 |
| 0.3128 | 0.0001 |
| 0.3249 | 0.0001 |
| 0.1190 | 0.0001 |
| Job Satisfaction: | Job Satisfaction: | ||||
| 0.2611 | 0.0030 |
| 0.0433 | 0.0168 |
| 0.3595 | 0.0001 |
| 0.0428 | 0.0178 |
| 0.2337 | 0.0074 |
| 0.1197 | 0.0001 |
| 0.3808 | 0.0001 |
| 0.2826 | 0.0001 |
| Performance: | Performance: | ||||
| 0.5616 | 0.0038 |
| 0.2866 | 0.0001 |
| 0.7614 | 0.0001 |
| 0.3278 | 0.0001 |
| 0.5711 | 0.0028 |
| 0.2948 | 0.0001 |
| Variables (Korea Data) | Prob. >F | Variables (America Data) | Prob. >F |
|---|---|---|---|
| Organizationally Mediated | Organizationally Mediated | ||
|
| ||
| .0127 |
| .0001 |
| .0001 |
| .0001 |
| .0001 |
| .0001 |
|
| ||
| .0117 |
| .0117 |
| .0015 |
| .0015 |
| .0003 |
| .0001 |
| .0012 |
| .0003 |
| .0001 |
| .0001 |
| Interpersonally or Job Mediated | Interpersonally or Job Mediated | ||
| .0008 |
| .0001 |
|
| ||
| .0390 |
| .0001 |
| .6365 |
| .0001 |
|
| ||
| .0001 |
| .0001 |
| .0212 |
| .0001 |
| Internally or Personally Mediated | Internally or Personally Mediated | ||
|
| ||
|
| ||
| .0004 |
| .0001 |
| .0001 |
| .0001 |
| .1832 |
| .0001 |
|
| ||
| .1832 |
| .0001 |
| .0001 |
| .0001 |
| .0141 | .0141 |
|
| < Day Day Up > |
|
EAN: 2147483647
Pages: 207
- Article 312 Cabinets, Cutout Boxes, and Meter Socket Enclosures
- Article 324 Flat Conductor Cable Type FCC
- Article 326 Integrated Gas Spacer Cable Type IGS
- Article 422: Appliances
- Example No. D10 Feeder Ampacity Determination for Adjustable-Speed Drive Control [See 215.2, 430.24, 620.13, 620.14, 620.61, Tables 430.22(E), and 620.14]