Page 264 - Kỷ yếu hội thảo khoa học lần thứ 12 - Công nghệ thông tin và Ứng dụng trong các lĩnh vực (CITA 2023)
P. 264
248
3 Research Methodology
3.1 Data Collection Methods
The research employed a quantitative approach. The questionnaire is constructed with
22 observed variables for four groups of proposed research model factors. The
questionnaire's scale variables of eWOM are measured on a 5-point Likert scale, with
one representing "Absolutely disagree" and five representing "Absolutely agree." About
Social media experience and factors related to Attitude (ASM, AB) were measured by
5-point semantic differential scales anchored by not at all (1) and very much (5).
3.2 Participants and Sampling
According to Hoang Trong and Chu Nguyen Mong Ngoc [12], the sample size must
be at least 5 times the number of variables in the factor analysis. With the number of
observed variables of 22, the minimum sample size should be 110. With the expecta-
tion that a valid sample has a proportion greater than 50% of the total number of
samples collected, the study chooses a sample size of n=350. In this study, the authors
performed cluster sampling. Specifically, the authors divided students from member
universities of the University of Danang into 5 groups to 5 majors: Social Science;
Science, engineering, and technology; Natural Science; Medical and pharmaceutical
science; Humanities. Then, a simple random sample technique is used to choose
students in each group with a Google survey questionnaire via the social media page
colleges. Within one month of data collection from January 2023, there are 310
students from member universities of the University of Danang participated in the
survey.
3.3 Data Analysis Methods
After coding and screening, the data is entered into the AMOS 26 with Structural
Equation Modeling (SEM) statistical tool to analyze and evaluate the measurement
model and the structural model of the proposed research model. The goodness of fit
test model will be assessed based on the relevance of the appropriateness indexes such
as CMIN/DF, Root Mean Square Error of Approximation (RMSEA), the goodness of
index (GFI), Tucker Lewis Index (TLI), and Comparative Fit Index (CFI). Structure
validity will be based on tabulated results of structure reliability (CR) and
mean-variance extracted values (AVE). Finally, SEM analysis was performed to
assess whether the formulated research hypothesis was accepted.
CITA 2023 ISBN: 978-604-80-8083-9