2. Material and methods
Data from the Brazilian students was collected in spring 2009, while
the data from the Finnish and Chinese B.Sc. students was collected during
spring 2010. The questionnaire was a translated version, from
English into the respective national language, of the one designed by
the authors in the previous study. Each translation involved the participation
of two independent persons from the targeted university.
The first part of the questionnaire consisted of closed-ended
Likert-type questions that asked for a numerical evaluation of the importance
that students give to a list of 39 subject-specific and generic competencies
and experiences. The importance was asked on a 5-point
scale, from “of little importance (1)” to “extremely important (5)”. The
second part consisted on closed-ended questions on the students' future
plans and preferences with regard to work and further studies, as well as
on their satisfaction with the studies and self-reported grades
[from “well-below average (1)” to “well above average (5)”]. In addition,
concerns and views from the students were attained through
open ended questions. Questionnaires in paper form were administered
during class in all cases except for the final year students from UEF and
NWAFU, who were only doing their thesis, and for whom a course representative
distributed and collected the questionnaires.
Data from 584 students was analysed. Respondents were 21 years
old on average (average and standard deviation of 21.3 and 3.3 for
UFPR, 21.3 and 2.0 for UEF, and 21.4 and 1.4 for NWAFU) and 47% female
(52% at UFPR, 33% at UEF, and 46% at NWAFU). Response rates
ranged from 65% to 98% depending on the study year and university
(Table 2). Questionnaire data was supplemented by curriculum information
and other relevant literature as well as with interviews with a sample
of the students and staff members of the targeted universities.
Likert-type responses were statistically examined for differences by
study year within universities using One-Way ANOVA and Tukey's Post
Hoc Multiple Comparisons test. Following the methodology of the previous
study on Brazilian students, second and third year Chinese students
were grouped as middle year students, forming three groups – first year,
middle year(s) and final year(s) (as indicated in Table 2). Chi-Square test
was used to examine differences across gender. Spearman's correlation
coefficients were calculated to investigate the relationships between
performance, satisfaction and future study and work plans. Principle
Components Analysis (PCA) with Varimax rotation was used to explore
patterns of responses and latent dimensions in the value given by students
to generic competencies and experiences, followed by an examination
of the differences to the resulting components. Kaiser–Meyer–Olkin
and Barttlet's Test of Sphericity indicated that the data was suitable for
PCA. Mean values given by the students to the competency and experience
items were used for establishing the rankings of importance within
universities. As commonly done in cross-cultural studies, rankings rather
than mean values were used for comparing results across countries, as
mean values are affected by the different response styles that exist between
countries [as in the study of Harzing (2006), the percentage of
4/5 answers in our results was greater among the Chinese and the
Brazilians as compared to the Finns].
2. Material and methodsData from the Brazilian students was collected in spring 2009, whilethe data from the Finnish and Chinese B.Sc. students was collected duringspring 2010. The questionnaire was a translated version, fromEnglish into the respective national language, of the one designed bythe authors in the previous study. Each translation involved the participationof two independent persons from the targeted university.The first part of the questionnaire consisted of closed-endedLikert-type questions that asked for a numerical evaluation of the importancethat students give to a list of 39 subject-specific and generic competenciesand experiences. The importance was asked on a 5-pointscale, from “of little importance (1)” to “extremely important (5)”. Thesecond part consisted on closed-ended questions on the students' futureplans and preferences with regard to work and further studies, as well ason their satisfaction with the studies and self-reported grades[from “well-below average (1)” to “well above average (5)”]. In addition,concerns and views from the students were attained throughopen ended questions. Questionnaires in paper form were administeredduring class in all cases except for the final year students from UEF andNWAFU, who were only doing their thesis, and for whom a course representativedistributed and collected the questionnaires.Data from 584 students was analysed. Respondents were 21 yearsold on average (average and standard deviation of 21.3 and 3.3 forUFPR, 21.3 and 2.0 for UEF, and 21.4 and 1.4 for NWAFU) and 47% female(52% at UFPR, 33% at UEF, and 46% at NWAFU). Response ratesranged from 65% to 98% depending on the study year and university(Table 2). Questionnaire data was supplemented by curriculum informationand other relevant literature as well as with interviews with a sampleof the students and staff members of the targeted universities.Likert-type responses were statistically examined for differences bystudy year within universities using One-Way ANOVA and Tukey's PostHoc Multiple Comparisons test. Following the methodology of the previousstudy on Brazilian students, second and third year Chinese studentswere grouped as middle year students, forming three groups – first year,middle year(s) and final year(s) (as indicated in Table 2). Chi-Square testwas used to examine differences across gender. Spearman's correlationcoefficients were calculated to investigate the relationships betweenperformance, satisfaction and future study and work plans. PrincipleComponents Analysis (PCA) with Varimax rotation was used to explorepatterns of responses and latent dimensions in the value given by studentsto generic competencies and experiences, followed by an examinationof the differences to the resulting components. Kaiser–Meyer–Olkinand Barttlet's Test of Sphericity indicated that the data was suitable forPCA. Mean values given by the students to the competency and experienceitems were used for establishing the rankings of importance withinuniversities. As commonly done in cross-cultural studies, rankings ratherthan mean values were used for comparing results across countries, asmean values are affected by the different response styles that exist betweencountries [as in the study of Harzing (2006), the percentage of4/5 answers in our results was greater among the Chinese and theBrazilians as compared to the Finns].
การแปล กรุณารอสักครู่..
