4.3 METHODOLOGY
4.3.7 Data management and analysis procedure
4.3.7.1 Questionnaire data analysis
Three interrelated steps were included in my process of analysing the questionnaire data: preparing the data for analysis, the data analysis procedure and interpreting the analysis results (Creswell, 2010). The program of IBM SPSS Statistics 20 was employed for the quantitative data analysis. The whole procedure in my study is explained as follows:
In the preparation stage, the data was first scored for future analysis. This procedure operated thus: a numeric score or value was assigned to each response category for each question. For instance, as to the question of “what is your political status”, I assigned numbers such as 1= Member of Chinese Communist Party, 2=Member of the Communist Youth League of China, 3=Unaffiliated and 4=Confidential. In addition, as to the questions checking students’ responses to a statement, scores were assigned as 1=strongly disagree, 2=disagree, 3=neither agree nor disagree, 4=agree and 5= strongly agree. After the software of SPSS was installed into my computer, the main task was to convert the raw paper data into the electronic version in the SPSS program. Before it progressed to the real
analysis stage, the data was cleaned to see whether there was any error. The missing data for some questions were noticed when I was imputing the data. The questionnaires with missing data were then discarded and the original questionnaires were immediately destroyed by a paper shredder machine. Huge time and efforts were spent to finish this preparation stage, as there were large numbers of the sample and particular questions in the questionnaire in this research.
Another task in this preparation stage was to test the reliability of certain variables. According to the theoretical framework depicted earlier, some variables need to be summated for the quantitative analysis, such as variables of teaching methods, learning outcomes, campus activities, social trust, online activities, broader civic participation, political participation, civic commitment, and students’ general attitudes. Thus the term Cronbach's Alpha is introduced. Santos (1999) suggested that Cronbach's Alpha (1951) should be applied in research studies with the intention to assess and improve the reliability of variables which are derived from summated scales for further analysis need. In addition, Nunnaly (1978) indicated that a satisfactory Cronbach’s Alpha value should be above 0.7. However, it is also acceptable if it is just above 0.6. Table 4.1 below in the following page reported the internal consistency reliability (Cronbach's Alpha coefficient) of summated variables in my study, and the results showed that all Cronbach’s Alpha values are larger than 0.6, which are ranging from 0.753 to 0.926. Therefore, it can be concluded that the summated variables in this research are of a relatively high consistency, thus indicating all items can be kept in the later data analysis procedure.
The second stage was to analyse the input data using the instruments in SPSS with the intention to provide evidence for addressing my research questions. Two analysis approaches of descriptive statistics and inferential statistics were used in the analysis procedure, and the decision of employing which type of analysis was made according to the design of variables in research questions.
Scale Cronbach’s Alpha Number of items Teaching methods .894 7 Campus activities .876 7 trust .872 5 Online activities .926 11 Learning outcomes .876 9 Political participation .855 2
Broader Civic participation .753 4
Civic commitments .855 3
Attitude towards university curriculum .884 5
Attitude towards mass media .878 5
Table 4. 1: The Cronbach’s Alpha values of summated variables in this research
Descriptive statistics
Descriptive statistics indicate the general tendency of a single variable and report its mean, mode and the spread of scores (standard deviation and range). The mode is the score which appears most frequently in a list of scores. The tools of counts and percentages are generally used for categorical variables. For example, the tool of descriptive statistic was employed to obtain a general picture of participants’ background information. Besides that, the frequency analysis was used in examining participants’ perceptions. For instance, to analyse the question of “are students satisfied with the university curriculum for citizenship values and practices”, the frequency analysis of students’ learning outcomes was used to indicate the general tendency among the sample.
Inferential analysis
Inferential analysis is introduced when comparing groups or relating two or more variables. In order to test the proposed potential relations, different types of inferential statistics were used. The techniques of one sample T-test, ordinal multiple regression and one-way ANOVA test were mainly employed in this process.
One sample T- test is a statistical procedure used to examine the mean difference between the sample and the known value of a population mean. Therefore, it was used to examine differences in civic participation among different groups such as females and males; party members and non-party member students. In addition, this technique was also used to show the differences between the means of participants’ attitudes towards the university curriculum and mass media. Another tool is regression analysis, which is employed to identify the casual relationship at work among variables revealed from the literature. For instance, regression analysis was performed within a range of variables such as forms of online participation and offline participation, and mass media use. One-way ANOVA test is conducted in order to test whether there is statistically significant difference between the groups. For instance, it was used to test statistically the difference in civic intention between males and females.
In the third stage, it is the turn to report my results using the tables and figures obtained from the testing stage. The analysis results were presented according to each research question one by one in order. When presenting the results, numbers in each table and figure were carefully explained, and the results were interpreted based on SPSS knowledge. A more detailed explanation of these statistical procedures is presented in Chapter Five.