Descriptive results
Results for hypothesis 1:
Using the Student Satisfaction Questionnaire and on the 5-point scale, the sample shows that the international students are at a good satisfaction level (M = 3.47, SD = .68).
This means that the score is above the midpoint of the scale which is 3. Using the scores below 2.5 and above 3.5 as cut-off points for poor satisfaction scores and very good satisfaction scores, approximately 9.4% and 51.6% of the international students studying in Taiwan could be said to have poor and very good satisfaction scores respectively.
Therefore we reject the Null Hypothesis 1 and accept the alternative. The results of this study indicate that a majority of the international students are satisfied with their lives here where more than 50% may be described as satisfied to very satisfied. This is based on an arbitrary cutoff point established by the researcher that has not been validated so this interpretation needs to be qualified. Still yet it satisfies the research.
The researcher also looks at the percentages of the responses, the means, and the standard deviations for the 35 questions to see the factors that students are most and least
satisfied with. The responses expressed in the five-point Likert scale system have been separated by the dimensions as seen in the four tables below.
Power distance
From the variables that belong to the factor power distance, the students show high dissatisfaction to Q13. This shows that 39.5 % are dissatisfied that there are no forums for them to express their concerns about their lives here in Taiwan. None of the other questions show great dissatisfaction; they all show about 50% or more satisfaction.
Table 4.1. Power distance by Likert scale, mean and standard deviation (n = 256)
Variables VD D N S VS Mean SD
Q1 1.2 7.0 22.7 30.5 38.7 3.98 1.002
Q4 6.3 7.0 28.5 32.8 25.4 3.64 1.122
Q7 3.1 7.8 20.3 32.4 36.3 3.91 1.075
Q10 5.5 4.7 21.5 31.3 37.1 3.90 1.123
Q13 16.8 22.7 28.9 2.5 10.2 2.86 1.227
Q15 3.1 7.8 31.6 35.2 22.3 3.66 1.009
Q18 5.5 15.2 30.1 35.5 13.7 3.37 1.069
Note: The 5-Point Likert scale is reported in percentages
Uncertainty avoidance
The uncertainty avoidance factor provides favorable results. All questions indicated fair to high satisfaction level. Q14, Q20, Q23, and Q31 show some reasonable dissatisfaction but not too significant enough to cause great concern.
Table 4.2. Uncertainty avoidance by Likert scale, mean and standard deviation (n = 256)
Variables VD D N S VS Mean SD
Q12 4.3 12.5 19.9 32.4 30.9 3.73 1.152
Q14 9.8 18.4 21.9 37.5 12.5 3.25 1.181
Q17 3.9 11.3 30.5 37.5 16.8 3.52 1.025
Q20 10.5 12.1 27.3 30.5 19.5 3.36 1.226
Q21 5.9 7.8 25.0 25.4 35.9 3.78 1.185
Q23 9.8 11.7 27.7 35.5 15.2 3.35 1.165
Q24 4.3 6.6 16.0 34.0 39.1 3.97 1.099
Q26 5.5 5.1 20.3 35.5 33.6 3.87 1.105
Q30 2.7 8.6 27.0 30.5 31.3 3.79 1.064
Q31 10.2 16.0 26.2 35.5 12.1 3.23 1.165
Q34 5.9 9.8 27.7 28.5 28.1 3.63 1.161
Note: The 5-Point Likert scale is reported in percentages
61 Masculinity
This factor shows three questions that have high dissatisfaction levels. Q22 shows that 46.7% of the students are dissatisfied with counseling and psychological services offered in English to them. Q25 illustrates that 46.5% of the students are dissatisfied with the selection of food at their cafeteria. 39.8% of the students for Q28 show dissatisfaction with their living conditions at the dorm. The other questions show fair dissatisfaction but none too significant enough for concern.
Table 4.3. Masculinity by Likert scale, mean and standard deviation (n = 256)
Variables VD D N S VS Mean SD
Q3 10.5 17.6 26.6 25.4 19.9 3.27 1.259
Q6 5.5 19.9 23.4 31.3 19.9 3.40 1.171
Q9 4.7 14.8 18.0 31.6 30.9 3.69 1.189
Q16 11.7 11.3 30.9 25.8 20.3 3.32 1.248
Q19 12.9 12.5 28.9 25.4 20.3 3.28 1.279
Q22 16.8 19.9 32.0 21.9 9.4 2.87 1.206
Q25 22.7 23.8 25.4 17.2 10.9 2.70 1.292
Q28 19.1 20.7 19.9 23.4 16.8 2.98 1.373
Q35 3.5 13.7 27.7 34.4 20.7 3.55 1.073
Note: The 5-Point Likert scale is reported in percentages
Individualism
For this factor, only Q27 shows significant dissatisfaction where 34% of the students feel that they are not being informed or updated on the student activities at their institution. With the exception of Q29 and Q32, the other remaining questions show some degrees of dissatisfaction but none too high enough for concern.
Table 4.4. Individualism by Likert scale, mean and standard deviation (n = 256)
Variables VD D N S VS Mean SD
Q2 3.5 19.5 23.0 35.5 18.4 3.46 1.105
Q5 9.8 11.3 31.3 33.6 14.1 3.31 1.146
Q8 10.2 14.5 22.3 21.1 32.0 3.50 1.340
Q11 5.1 13.7 21.1 30.5 29.7 3.66 1.184
Q27 18.0 16.0 25.4 23.4 17.2 3.06 1.343
Q29 5.9 5.5 21.1 34.4 33.2 3.84 1.126
Q32 0 5.5 30.1 50.0 14.5 3.73 .772
Q33 8.2 17.6 35.9 31.6 6.6 3.11 1.038
Note: The 5-Point Likert scale is reported in percentages
From all the tables above, the researcher has decided to show the top 5 satisfaction variables and the bottom 5 dissatisfaction variables as indicated by the respondents. In Table 4.5 we can see that students feel respected, free of racial discrimination by professors, and safe and healthy on campus. The also feel free to express themselves in class.
Table 4.5. Top 5 satisfaction responses from the Student Satisfaction Questionnaire (n = 256)
Questions Descriptions
Q24 I am satisfied that I am treated as a student free of racial discrimination towards me from my professors.
Q1 I am satisfied with the opportunity to freely express my opinion in class.
Q26 I am satisfied with my health and safety on campus.
Q7 I am satisfied with the level of equitable respect I receive from my professors.
Q10 I am satisfied with the level of equitable respect for my culture by my classmates.
In Table 4.6 we see that students want better services and in English, a better selection of food, and better living conditions, they want to feel more motivated on campus, and be kept updated about school activities.
Table 4.6. Bottom 5 dissatisfaction responses from the Student Satisfaction Questionnaire (n = 256)
Questions Descriptions
Q22 I am satisfied with the counseling and psychological services in English offered to me as an international student.
Q25 I am satisfied with the diverse selection of food at the cafeteria.
Q28 I am satisfied with the living conditions (dorms) I have.
Q3 I am satisfied that the institution (foreign affairs, department. Etc) tries o motivate me as a foreign student.
Q27 I feel that I am kept informed and updated of the students activities at my university
63 Inferential results I
In order to test hypotheses 2 to 10 of Research Questions 2 and 3, a Multiple Linear Regression Analysis was used to investigate whether the independent variables have statistical significance as predictor variables and students GPA as the criterion variable. The independent variables are the 35 Questions, Age, Gender, Nationality, Degree, and Status. The 35 Questions are subdivisions of Hofsetede’s Dimensions - Power Distance, Uncertainty Avoidance, Individualism, and Masculinity. Students’
nationality was subdivided into the geographical regions Asia, Europe, North, Central, and South Americas, Africa, Australia/Oceania. Students Status represents ICDF scholarships students and Non-ICDF scholarship students. Degree is divided into Graduates and Undergraduates.
The process used in the Multiple Linear Regression Analysis in Table 4.7 is the backward elimination procedure, which examines the p-values for the 45 independent variables, and eliminates the highest insignificant variable in each equation. This process is repeated 35 times, in 36 equations, until all remaining independent variables reach at least the 10% level of significance. In the table, the independent variables accounted for 82.2% of the variance of GPA in Equation 1 and 79.9% of the variance of GPA in Equation 36. For a clearer view, only Equations 1 and 36 are shown and the rest of the equations are reported in Appendix E. The researcher decided to show all variables and their significance in Equation 1 and how that value changes because of the backward regression process by Equation 36 where all remaining variables are significant. The 35 predictor variables which are divided by their factors are used to answer Research Question 2. The demographic variables are use to answer Research Question 3.
Table 4.7. Multiple Regression of independent variables as predictors for GPA (n = 256)
Variables Equation 1 Equation 36
β value¹ t-value β value¹ t-value
Q1 .176*** 3.429 .201 6.025***
Q4 -.070 -1.441
Q7 -.022 -.447
Power Q10 -.058 -1.139
Distance Q13 -.013 -.275
Q15 .088* 1.673
Q18 .067 1.287 .085 2.127**
Table 4.7. (Continued)
Variables Equation 1 Equation 36
β value¹ t-value β value¹ t-value
65 In Equation 1, of the 45 independent variables, 23 have negative parameters of which only 2 have statistical significance explaining the variance of GPA. From the other 22 variables that have positive parameters 6 have statistical significance to GPA. For the Power Distance dimension, two questions show significance to GPA. The parameter for Q1 is positive (.176), with the t-ratio (3.429), indicating that it is significant at the 1%
level and Q15 has a positive parameter (.088) and t-ratio (1.673) showing significance at 10%. The Individualism dimension only has one question that is statistically significant which is Q11 with a negative parameter (-.102) and a t-ratio of (-2.245) that is significant at 5%. For the Uncertainty Avoidance dimension, Q12 is the only significant question with a negative parameter of (-.090), t-ratio of (-1.690) that is significant at 10%. The final dimension Masculinity likewise only has one question Q28 significant with a positive parameter of (.071), t-ratio (2.069) that indicates a significance at 5%. From the demographic variables, Europe has a positive parameter (.388), t-ratio (2.192) that is significant at 5%, and South America also has a parameter (.303), t-ratio (1.756) that is significant at 10%. Students’ degree shows significance to performance with a positive parameter of (.476) and t-ratio (4.623) indicating significance at 1%. These results change from equation to equation. In Equation 36, 10 variables show statistical significance to students GPA. Of those 10 variables, 3 have negative parameters and 7 have positive parameters.
For easier reading Table 4.5 was divided by factors that correspond with their respective hypothesis. Each division was placed at the end of the results for each hypothesis.
Results for hypothesis 2:
The final equation shows that for the Power Distance dimension, only two questions show significance to GPA. Q1 has a positive parameter (.201) and t-ratio (6.025) which is significant at 1%; the results of this variable imply that as students’
satisfaction to freely express their opinions in class increases, their GPA increases. Q18 has a positive parameter (.088) and t-ratio (2.127) indicating significance at 5%; this implies that a rise in students’ feeling that they are being recognized for their performance and contribution at their school will produce a rise in their GPA. The other
questions show no statistical significance to GPA. As a result we reject the null hypothesis of Research Hypothesis 2 for Q1 and Q18 and accept the alternative that there is statistical difference between Power Distance and students’ academic performance.
The null hypothesis is accepted for the other questions belonging to Power Distance; they show that there is no statistical significance between Power Distance and students’
performance.
Table 4.8. Multiple Regression of Power Distance as predictor for GPA (n = 256)
Variables Equation 1 Equation 36
β value¹ t-value β value¹ t-value
Q1 .176*** 3.429 .201 6.025***
Q4 -.070 -1.441
Q7 -.022 -.447
Power Q10 -.058 -1.139
Distance Q13 -.013 -.275
Q15 .088* 1.673
Q18 .067 1.287 .085 2.127**
Note: * = p < .1, ** = p < .05, *** = p < .01
¹ The value of the beta here is from the Unstandardized Coefficients.
Results for hypothesis 3:
The Uncertainty Avoidance dimension only has one question that is statistically significant which is Q12 with a negative parameter (-.066) and a t-ratio (-2.127) that is significant at 10%. This result implies that a rise in the students’ satisfaction that their instructors have sufficient background knowledge would lead to a fall in their academic performance. As a result we reject the null hypothesis for Research Hypothesis 3 and accept the alternative that there is a statistical significant difference between Uncertainty Avoidance and students’ performance. The other questions show no significance resulting in the null hypothesis for Research Hypothesis 3 being accepted that there is no statistical difference.
67 Table 4.9. Multiple Regression of Uncertainty Avoidance as predictor for GPA (n = 256)
Variables Equation 1 Equation 36
β value¹ t-value β value¹ t-value
Q12 -.090* -1.690 -.066 -1.975**
Q14 .008 .167
Q17 .037 .650
Q20 .078 1.485
Uncertainty Q21 .047 .811
Avoidance Q23 -.019 -.333
Q24 .061 1.168
Q26 .008 .175
Q30 -.031 -.565
Q31 .012 .227
Q34 -.006 -.123
Note: * = p < .1, ** = p < .05, *** = p < .01
¹ The value of the beta here is from the Unstandardized Coefficients.
Results for hypothesis 4:
For the Masculinity dimension, Q16, Q19, and Q25 show significance to GPA.
Q16 has a negative parameter (-.060), a t-ratio (-1.975) and significant at 5%; it implies that an increase in students’ satisfaction with their international affairs office efforts to make their lives more adaptive in Taiwan would result in a decrease in their academic performance. Q19 has a positive parameter (.059) and a t-ratio (1.824) which is significant at 10%. This question suggests that as students’ satisfaction with the selection of cultural activities offered at their institution increases, their performance increases.
Q25 also has a positive parameter (.059) and t-ratio (2.007) significant at 5%. The students increase in satisfaction with their diet from the diverse selection of food at the cafeteria will produce an increase their academic performance. The null hypothesis of Research Hypothesis 4 for Q16, Q19, and Q25 is rejected and the results show that there statistical significance between Masculinity and students’ performance. The other questions are insignificant and thus the null hypothesis is accepted.
Table 4.10. Multiple Regression of Masculinity as predictor for GPA (n = 256)
Variables Equation 1 Equation 36
β value¹ t-value β value¹ t-value
Q3 .042 .893
Q6 .048 1/123
Q9 -.036 -.867
Q16 -.049 -.920 -.060 -1.739*
Masculinity Q19 .040 .899 .059 1.824*
Q22 -.045 -1.114
Q25 .046 1.289 .059 2.007**
Q28 .071** 2.069
Q35 -.006 -.119
Note: * = p < .1, ** = p < .05, *** = p < .01
¹ The value of the beta here is from the Unstandardized Coefficients.
Results for hypothesis 5:
The final dimension Individualism only has one question Q11 significant with a negative parameter (-.102) and a t-ratio (-2.245) which is significant at 5%. This result suggests that an increase in students’ satisfaction that they have equal opportunities to participate in campus activities will result in a decrease in their academic performance.
As a result we reject the null hypothesis of Research Hypothesis 5 for Q11 only and accept the alternative. The other questions gave no evidence of significance so the null hypothesis is accepted that there is no significant difference between Individualism and performance.
69 Table 4.11. Multiple Regression of Individualism as predictor for GPA (n = 256)
Variables Equation 1 Equation 36
β value¹ t-value β value¹ t-value
Q2 .053 1.037
Q5 -.002 -.041
Q8 -.041 -.916
Individualism Q11 -.102** -2.245 -.087 -2.493**
Q27 -.038 -1.026
Q29 -.048 -1.063
Q32 .059 .985
Q33 -.050 -1.100
Note: * = p < .1, ** = p < .05, *** = p < .01
¹ The value of the beta here is from the Unstandardized Coefficients.
Results for hypothesis 6:
The demographic variable Age was not statistically significant to student’s academic performance. As a result the null hypothesis for Research Hypothesis 6 of Research Question 3 is accepted; this states that Age has no statistical significance to students’ academic performance.
Results for hypothesis 7:
The demographic variable Gender was not statistically significant to student’s academic performance. As a result the null hypothesis for Research Hypothesis 7 of Research Question 3 is accepted; this states that Gender has not statistical significance to student’s academic performance.
Results for hypothesis 8:
Using regional dummy variables, and students from Central America as the baseline group, the regression shows that only Europe with positive parameter (.411) and t-ratio (3.209) significant at 5%, and South America with positive parameter (.292) t-ratio (2.309) and significant at 1% are statistically significant to GPA. This means that students from Europe and South America are performing better than those from Central America.
The other regions had no statistical significance to students’ performance. Therefore, the null hypothesis of Research Hypothesis 8 is rejected for Regions Europe and South
America and shows that there is statistical significance between students’ nationality and their performance. For Regions North America, Central America, Africa, and Oceania the null hypothesis is accepted.
Results for hypothesis 9:
Students’ Status (ICDF or Non-ICDF) was not statistically significant to student’s academic performance. As a result the null hypothesis for Research Hypothesis 9 of Research Question 3 is accepted; this states that there is no statistical significance between students’ Status and their academic performance.
Results for hypothesis 10:
Students’ degree shows significance to performance with a positive parameter of (.477) and t-ratio (6.951) which is significant at 1%. This suggests that students pursuing a Graduate Degree have better performance than those pursuing Undergraduate Degrees.
As a result the null hypothesis for Research Hypothesis 10 is rejected and shows that Degree does have statistical significance to students’ performance.
Table 4.12. Multiple Regression of Demographics, Status, and Degree as predictors for GPA (n = 256)
Equation 1 Equation 36
Variables
β value¹ t-value β value¹ t-value
Age -.001 -.075
Gender -.009 -.105
Asia .009 .064
NAmerica .150 .885
Demographic SAmerica .303* 1.756 .292 2.309**
Europe .388** 2.192 .411 3.209***
Africa -.131 -.885
AOceania -.143 -.811
Status ICDF/Non-ICDF -.032 -.297
Degree Grads/Undergrads .467*** 4.623 .477 6.951***
R² .882 .799
71 Note: * = p < .1, ** = p < .05, *** = p < .01
¹ The value of the beta here is from the Unstandardized Coefficients.
Inferential results II
In order to answer Research Hypotheses 11 and 12 for Research Questions 4 and 5 respectively, Crosstab and Chi-square tests were performed for each. The tests set out to show the difference in the means of ICDF and Non-ICDF scholarship students, and the difference between Graduate and Undergraduate students.
Results for hypothesis 11:
To test if ICDF students perform better than Non-ICDF Cross-tabulation and chi-square techniques were used to predict the results. From the table below the Pearson Chi-square value is 3.561 and p = .109 has no significance which indicates that there is similarity between ICDF and Non-ICDF students’ performances. The Table shows that that the expected count and count for ICDF and Non-ICDF for all levels of GPA are almost similar on all levels of their GPA. As a result we accept the Null Hypothesis Eleven for Research Question 5 which states that ICDF and Non-ICDF has no statistically significant difference in their academic performance.
Table 4.13. Crosstabular analysis of status of students by GPA (n = 256)
Status Statistics GPA
C = 70-79 B = 80-89 A = 90-100 Total
ICDF Count 23 71 20 114
Expected Count 25.8 63.7 24.5 114.0
% within GPA 39.7% 49.7% 36.4% 48.4%
Non-ICDF Count 35 72 35 142
Expected Count 32.2 79.3 30.5 142.0
% within GPA 60.3% 50.3% 63.6 55.5%
Total Count 58 143 55 256
Expected Count 58.0 143.0 55.0 256.0
% within GPA 100.0% 100.0% 100.0% 100.0%
Note: Chi-square value is 3.561 with df = 2 and (p = .109)
Results for hypothesis 12:
Likewise a crosstabular analysis and Chi-square was used to test the performance of Graduate and Undergraduate students. From the Table below, the Pearson Chi-square value is 38.997 and p = .000 showing a significance at 1% which indicates that there are no similarity in Graduate and Undergraduate performances. Even from the crosstabular table we see that the Graduates are performing above their expected count at A and B levels while Undergraduates are performing poorly at C level with 47, almost doubled the expected count and lower than expected in levels A and B. As a result we reject the Null Hypothesis 13 for Research Question 6 and accept the alternative.
Table 4.14. Crosstabular analysis of degree of students by GPA (n = 256)
Degree Statistics GPA
C = 70-79 B = 80-89 A = 90-100 Total
Undergraduate Count 47 64 13 124
Expected Count 28.1 69.3 26.6 124.0
% within GPA 81.0% 44.8% 23.6% 48.4%
Graduate Count 11 79 42 132
Expected Count 29.9 73.7 28.4 132.0
% within GPA 19.0% 55.2% 76.4 51.6%
Total Count 58 143 55 256
Expected Count 58.0 143.0 55.0 256.0
% within GPA 100.0% 100.0% 100.0% 100.0%
Note: Chi-square value is 38.997 with df = 2 and (p = .000)
73
Commentaries
Of the 256 students that answered the questionnaire 102 gave additional comments on their satisfaction here in Taiwan. This accounts for about 40% of the respondents. The comments has been coded, categorized and placed in a hierarchical order. Their comments were coded Positive and Negative as seen in Tables 4.4 and 4.5 below. Some of the respondents made more than one comment.
Dissatisfaction comments
Category I
In the Table 4.4 below we can see that the top problem is Communication.
Twenty students complained about communication barriers with classmates and teachers.
Eighteen are also dismayed because of being unable to access information because the materials, documents, and services, etc. provided are not in English. Finally twelve students want more time to learn Mandarin in order to communicate in and out of class.
Category II
The category ‘School’ is next; ten students finds that their comments in class either to the teacher or their classmates are not appreciated. Twelve of them do not believe in the potential of their teachers stating that they have good sound education and are very knowledgeable but were not trained as teachers so their deliverance and methodologies are not effective. Another eight say that it is the education system or curriculum of their institutions that is poor. Coupled to that, are the four PhD students, who are disturbed that they are taking some courses at Master’s level, which they have done already. Then another four view the grading system of their institutions as problematic, because too much credit is given for participation which is a subjective criterion. A final noteworthy problem is the view of two students that their professors give exams that are not testing the appropriate competency as prescribed by Bloom’s Taxonomy; instead students are expected regurgitate information committed to memory.
Category III
In the third Accommodations category, twelve students wish that the scholarship money given was increased and sometimes the payment takes long and they run out of money. Six of them complain about the small and overcrowded dorm rooms and wish that they are placed in single rooms; sharing becomes too problematic with limited space for mobility. Finally two of the students wish for the scholarship organizations to provide funds so that they can do their research at home as a lack of finance force them to change their thesis topics
Category IV
Under the fourth category labeled ‘Personal’, twelve of the students believe that the scholarship should provide for their spouses to be with them in Taiwan; they claim that they would study better if their spouses were with them. Six students mostly blacks felt discriminated against because of their nationality and skin color. Lastly for this category one student say that the education system in Taiwan does not provide encouragement for individual growth; the system encourages students to be too dependent on each other and on the professors.
Category V
The final category for Table 4.4 ‘Culture’ shows that eight of the students find it hard to adapt to Taiwan because of the difference between their culture and Taiwan’s.
This may be because of the lack of or insufficient provision of means for students to learn about Taiwan’s culture when they just arrive.