Study design
This was a cross-sectional, questionnaire-based study conducted between April 2021 and June 2021. A questionnaire was distributed for medical graduates at 4 universities in China. With the survey being performed, we aim to find the level of GC of China medical graduates and the factors influencing the fostering of GC.
Questionnaire design
A self-administered questionnaire was developed by referencing previous GC scales of graduate students (Hu & Li, 2021;Lu & Guo, 2018;Liu, et al., 2020). Since medical graduates have features that differ from those in other disciplines, some items were revised, for example, “I know the internationally accepted theories and schools of thought in my field of study or profession” and “I know the main internationally accepted research methods in my field of study or profession” were removed as being inapplicable to medical students, “I can easily use MS Office, PDF Reader and other common international software” was deleted, because they are common knowledge for medical graduates. Statements that were similar in one category were combined to make it brief. After that, experts in medical graduate education, internationalisation of higher education and statistics were invited for examining the validity of the scale. A pilot study was conducted on a small group of medical graduates for collecting feedbacks to ensure that they understand each statement and each item was applicable, etc.. After discussing with the team members about those feedbacks from the pilot work, the final version of the questionnaire was formulated.
There are three sections in the questionnaire. The first section, with 7 fact-based questions, collected information concerning students’ demographics, such as gender, degree level, field of study, place of origin and year of study, etc.. The second section focused on students’ viewpoints about the internationalisation of the school and their International involvement. The third section was composed of a self-reported GC scale, which concerns three dimensions, namely, "knowledge and understanding”, "professional skills” and "attitudes and values”, they were further separated into eight sub-dimensions, totally 27 questions. In the following two sections, students were asked to indicate the degree to which they agree with each of the statements on a five-point Likert scale. The questionnaire was written in Chinese.
Prior to the statistical test, the reliability and validity of the questionnaire were examined. We adopted alpha coefficient for the reliability test, resulting in an overall score as 0.938, "knowledge and understanding” 0.897, "professional skills" 0.914 and "attitudes and values" 0.913, which indicated a high degree of internal consistency of the overall scale and the indicators of each level; we used confirmatory factor analysis to test the validity, the KMO value reached 0.897, providing evidence of validity of the overall structure.
Participants and Data collection
The participants came from 4 universities in China that providing medical education programs, they are Suzhou Medical College of Soochow University, Nantong University Medical School, Weifang Medical University and Xuzhou Medical University. Two of them are independent medical universities, two are medical schools of comprehensive universities. These are the main types of medical institutions in China.
A self-administered online questionnaire was distributed through Questionnaire Star and WeChat app by administrators of the four universities in charge of graduate education and internationalisation of higher education. It was clearly stated on the top of the questionnaire that participation in the study was voluntary, and the results would be used for research purpose. All participants gave tacit support for their opinions to be published anonymously.
We received 1159 responses in total, of which 446 were from candidates for professional degrees, 713 for academic degrees. They were all in medicine and health related courses. Those in the programs for academic degrees were retained for further investigation, those for professional degrees were disregarded. The number of respondents from the two types of universities was balanced (385: 328).
Statistical analysis
All data was downloaded from Questionnaire Star and incorporated into a Microsoft Excel spreadsheet. Data analysis was done by using SPSS, ver- sion 22.0. Descriptive statistics was made to analyze the level of medical graduates GC. Influencing factors were investigated using multiple linear regression.
Given the idea that the smaller the difference of GC between colleges and universities, the greater the possibility of clustering into the same category (Xue, 2014, pp. 288-314), this study also adopted hierarchical cluster analysis, and divided GC of medical graduates into four levels, they are A ≥ 4.039; 3.259 ≤ B ≤ 4; 2.778 ≤ C ≤ 3.2,22 and D ≤ 2.741.
Basing on previous studies (Hu & Li, 2021;Lu & Guo, 2018;Liu, et al., 2020), Influencing factors of medical graduates' GC were categorized into four types, they are demographic information, family background, school management and international involvement. Descriptive statistics were shown in Table 1.
[insert table 1 here]
In order to measure the influence of a single type of variables imposed upon the development of GC, four multiple regression models were constructed as follows:
Model 1: medical graduates' GC + demographic information;
Model 2: medical graduates' GC + demographic information + family background variables;
Model 3: medical graduates' GC + demographic information + family background variables + school management variables;
Model 4: medical graduates' GC + demographic information + family background variables + school management variables + individual international involvement variables.
The expression of the model is :
GC stands for medical graduates' GC, X1, X2, Xi represents various factors affecting GC, β 1, β 2,..., β N is the partial regression coefficient of each factor, ε is the residual term.