Understanding UX Better
Methology
Measurement
The questionnaire was designed by taking following steps. Firstly, an initial questionnaire is developed after a thorough review of the existing relevant literature. Secondly, the measurement scales were refined and wording of some items were adjusted after a round of expert review of the initial questionnaire. Thirdly, a pilot test (N = 176) was conducted in a university in Hangzhou, China to preliminary assess reliability and validity of the measurement. This three-step process led to the final questionnaire for the survey of the study. A Likert-type 5 point scale is used to access each of the measurement items with 1 representing "strongly disagree" and 5 standing for "strongly agree". The measurement scales are presented in Table 1.
Table 1. Measurements of the constructs.
Data collection
Our respondents are users of Ctrip for more than 1 year. Ctrip is a leading TMP in China. In 2021, the number of Ctrip users reached 220 million. We collect data both online and offline. The online survey was mainly conducted through "wjx.cn" (a charge online survey website of China). As for the offline survey, the questionnaires were randomly distributed in the West Lake Scenic Area in Hangzhou, China, which is the main gathering place for visitors to Hangzhou. When approaching a respondent, we first asked whether he or she had used Ctrip, and only respondents who had relevant experience of the TMP were invited to participate in the survey. The survey was conducted in May, 2022, and 310 completed questionnaires were collected. After eliminating questionnaires with missing date and obviously consistent answers, 276 valid questionnaires were obtained and used in subsequent analyses, resulting in an effective response rate of 89%. Loehlin and Mueller suggested that the sample size for structural equation model (SEM) is at least 100, and in most cases a sample with more than 200 cases is better. Therefore, we feel that a sample size of 276 for the study was generally acceptable to generate reliable results.
The sample consists of relatively more females (56.7%) than males (43.3%).Over 75% of the participants are within the age between 25 and 34. As to education background, the sample is characterized by a rather high education level with 89.9% of the participants having a bachelor degree or above. The sample demographics are consistent with the previous research that has reported a well-educated young groups who use TMP frequently for their travel plans.
Data analysis
This study mainly used SEM to test the conceptual model and research hypothesis. The internal consistency of the scale was measured by the Cronbach's α coefficient in SPSS21.0. On this basis, we further analyze the data in two steps. Firstly, the reliability and validity of the variable measurement are verified by confirmatory factor analysis (CFA), and then the structural model was tested to evaluate the proposed theoretical model and hypotheses therein. The data analysis of this study was carried out in Amos 21.0.