Similar to the quality of eWOM information, the experts also fully agreed with the scale and the appropriateness of the concept and scale. The experts made some minor adjustments to the wording of the scale.
The eWOM usefulness scale
Table 3.4: Scale of usefulness correction of eWOM
Original scale
Calibrated scale | |
Online reviews are valuable for customer purchases | Online reviews are valuable for choosing foreign tours |
Online reviews do not contain more information | Online reviews/ratings of the tour Foreign travel is informative. |
Online reviews allow you to find find information faster | Remove |
Online reviews are helpful for finding information. | Online reviews/ratings of the tour Foreign calendars are useful in finding information. |
On-line reviews make it easy buy a product or service | Remove |
Online reviews help improve the effectiveness of purchasing products or services | Online reviews/ratings of foreign tours help improve the effectiveness of your choice. choose tour |
Overall, online reviews are useful to consumers | Overall, online reviews/comments is helpful to me when choosing a tour |
Maybe you are interested!
-
The Impact of Website Quality on Tourist Satisfaction and Purchase Intention - The Case of a Travel and Tourism Business - 14 -
Foreign Studies on Improving Competitiveness for Tourism Development -
Using Cross Currency Contracts to Speculate on the Foreign Exchange Market: -
Organizational Structure of Joint Stock Commercial Bank for Foreign Trade of Vietnam - Bac Ninh Branch -
Scientific basis for perfecting state policies on foreign invested economy (FIE) in Vietnam - 2
Source: Compiled from expert interview results
Regarding the usefulness of information, the original scale was based on the scale in the study of (Bailey and Pearson, 1983; Gefen et al., 2003; Cheung, Lee & Rabjohn 2008). In the process of combining the scales of previous studies, the interviewed experts said that 2 statements: “online reviews make it easier to buy products or services” and
“Online reviews help improve the effectiveness of purchasing products or services”, so experts believe that one statement should be removed. After discussing with experts, the majority (6 out of 7 experts) of experts recommended that the statement: “Online reviews help make it easier to buy products or services” should be removed and the statement: “Online reviews allow for faster information search” should be removed instead, just the statement: “Online reviews are useful for information search” is enough.
Number of eWOM
The experts interviewed all agreed with the scale on the quantity of eWOM as well as the suitability of the scale with the research concepts. The scale of this variable did not need much adjustment, but the wording still needed to be adjusted to suit the survey respondents. The results of the scale adjustment for the variable of quantity of eWOM are as follows:
Table 3.5: Scale for adjusting eWOM quantity variables
Original scale
Calibration scale | |
Large online reviews indicate product popularity | Online reviews/comments about foreign tours often infer that the tour is popular |
Number of reviews of many products with good sales | Comments/reviews on water tours In addition to the big tour, it also has good sales revenue. |
Number of reviews many products have good reputation | Comments/reviews on water tours Large in appearance, the company has a good reputation. |
Overall, the number of product reviews Product X is large | Overall, the number of comments/reviews about many foreign tours |
Source: Compiled from expert interview results
eWOM sender expertise
The scale for eWOM sender expertise does not require the removal or addition of other variables to the study. However, the scale also has certain adjustments to make the wording and sentence structure more understandable to the survey respondents:
Table 3.6: Calibration of eWOM sender expertise scale
Original scale
Calibrated scale | |
eWOM Providers they have experience | People who provide online reviews/ratings Experienced foreign tour operator |
eWOM providers have rich knowledge | People who provide comments/reviews on foreign tours have rich knowledge about tourism. calendar |
eWOM providers have the ability to make judgments | People who provide comments/reviews on foreign tours are able to make judgments Exactly |
Professionals provide better knowledge about products and services | Professionals provide better knowledge about foreign tours |
Source: Compiled from expert interview results
Scale of intention to purchase foreign tours
The experts participating in the interview agreed with the scale of intention to purchase foreign tours and evaluated the scale as reliable for conducting research. However, the experts gave some advice on adjusting the way of using words so that the survey participants could easily understand the meaning in the interview questionnaire. Accordingly, the experts suggested replacing the word eWOM with comments/evaluations. In addition, the word product X should also be replaced with foreign tours to make the sentence clearer. Specifically, the results of adjusting the scale and the opinions of the experts are recorded in the following table:
Table 3.7: Calibration of the scale of intention to purchase foreign tours
Original ladder
Calibration scale | |
After consulting eWOM about product X on the Internet, I am interested in buying it. This product is better than others. | After consulting online reviews/comments about a foreign tour I like buy this tour over other tours |
After consulting word of mouth information about product X on the Internet, I am ready to introduce it to you. other people to buy products | After reading online reviews/comments about foreign tours, I am willing to recommend them to others. |
After consulting eWOM about product X on the Internet, I intend to buy products in the future | After consulting online reviews/comments about foreign tours, I intend to buy that tour in the future |
Source: Compiled from expert interview results
3.3 Formal research – quantitative research
3.3.1 Selecting research samples
The minimum sample size for a research paper must be from 100 to 150 samples and for EFA factor analysis, the sample size is proportional to the observed variable is 5:1. The minimum sample size is 5 samples for a parameter to be estimated, meaning N>=n*5 (N is the minimum sample size, n is the total number of observed variables) (Hair et al., 1998). On the other hand, for multivariate regression analysis, the sample size must be according to the formula: N= 50 + 8*n (Tabachnick and Fidell, 1996). Thus, if according to the requirement of Hair et al. (1998) with the number of observed variables in the paper being 25, the minimum sample size must be: 25 * 5= 125 samples and according to the requirement of Tabachnick and Fidell (1996) with 5 independent variables, the minimum sample size in the study is: N= 50+8*5 = 90 samples. The intended number of survey forms issued was greater than 125 samples, the larger the sample size, the higher the representativeness of the sample, in fact, the total number of survey forms issued was 378, of which 365 were collected. In the official study, the number of observation samples was 321, the number of samples was much larger than the minimum required sample.
The subjects selected for the survey were those who had read online reviews/ratings of foreign tours of companies. The subjects of the survey were customers who came to do business at travel companies, or customers who intended to go on a tour and had consulted online reviews/ratings of foreign tours. The survey locations were at a number of travel companies including: Intertour Vietnam Tourism Joint Stock Company (150/378 questionnaires issued); Viet Travel Company (70/378 questionnaires issued); VietTravel Company (70/378 questionnaires issued); at coffee shops in District 1 such as: Kai (30/378 questionnaires issued), Highlands (30/378 questionnaires issued), The Coffee House (28/378 questionnaires issued). The sampling method was the convenience method. Accordingly, the author distributed questionnaires to all customers when they came to do transactions or to travel companies to get advice, support, and book tours.
3.3.2 Data processing
The research data will be entered from the questionnaire into SPSS 22 software, which is also used to perform statistical techniques and data analysis. Before entering the data into SPSS 22, the author has filtered out invalid survey forms, which are surveys that customers have chosen without reading online reviews/reviews about foreign tours. At the same time, through this software, the data will be cleaned to detect errors in data entry, missing data entry fields and compared with the survey form.
3.2.2.1 Data cleaning
According to Hoang Trong and Chu Nguyen Mong Ngoc (2008), data after being entered usually cannot be processed and analyzed immediately because there may be errors such as: incorrect, missing or redundant data, interviewers misunderstanding the questions and collecting the wrong data, choosing the wrong interviewee, interviewees misunderstanding the meaning... Data needs to be cleaned before statistics and analysis. Data can be cleaned by using frequency tables, using two-variable or three-variable coordination tables, finding errors on the Data window.
view of SPSS software. In the article, the author selects valid questionnaires and uses frequency tables to check missing questionnaires.
3.3.2.2 Descriptive statistics
Descriptive statistics are often used to describe the basic characteristics of data Sternstein and Martin (1996). In this study, statistical analysis is used to describe the sample including: age, gender, education level, occupation, income and research variables.
3.3.2.3 Assessing the reliability of the scale
When conducting quantitative research on a scientific research topic, depending on the topic, the researcher can use different scales. Research concepts require elaborately constructed scales and must be tested for reliability before application (Hoang Trong and Chu Nguyen Mong Ngoc, 2008). The reliability of the scale is assessed by the Cronbach' Alpha coefficient (Hoang Trong and Nguyen Thi Mai Trang, 2009).
According to Nunnally et al. (1994), the Cronbach' Alpha coefficient must be from 0.8 or closer to 1, then the scale is considered good. According to Peterson (1994), the Cronbach' Alpha coefficient from 0.7 to 0.8 is a reliable and usable scale. For new research concepts, the Cronbach' Alpha coefficient must be from 0.6 or higher Slater (1995). At the same time, the observed variable has a correlation with the total variable greater than 0.3; variables with a correlation coefficient less than 0.3 will be eliminated from the model.
3.3.2.4 Exploratory factor analysis EFA
Factor analysis is a general name for a group of procedures used to summarize and reduce data (Hoang Trong and Chu Nguyen Mong Ngoc, 2008). According to Hair et al. (1998), factor loading is the most important indicator for EFA analysis. In which, factor loading must be greater than 0.3. A large KMO value (from 0.5 to 1) is the most suitable condition for factor analysis and vice versa.
3.3.2.5 Multiple regression analysis
The study uses multivariate regression to determine the significance level and linear correlation between the variables shown in the model. At the same time, it determines the influence of eWOM on the intention to buy a tour. Before analyzing the regression, the study considers the correlation between the independent variables (quality, usefulness, reliability, quantity, expertise of the electronic word-of-mouth sender) and the dependent variable (intention to buy a tour), F test to evaluate the suitability of the model, adjusted R2 coefficient , giving a regression model showing the relationship between the independent variable eWOM and the dependent variable intention to buy a tour.
3.3.2.6 Testing the difference in the influence of eWOM on customers' purchase intention
The study uses two techniques to test the difference between age, gender and income groups in terms of the influence of eWOM on customers' intention to buy tours. In which, Independent Sample T-test is used to test the difference between the two gender groups of men and women. Anova is used to test the difference between income groups and age groups. The difference is statistically significant with a confidence level of 95% and a significance level of Sig less than 0.05.
CHAPTER 3 SUMMARY
After presenting the model and research hypotheses, the author proceeded to design the research. The research process consists of 5 steps, in which the Likert scale is used to design the questionnaire. The original scale is given in chapter 3 based on previous foreign studies and has cited sources. In different markets, customers will have some differences in their views and perceptions of eWOM.
To make the scale suitable for the research subjects, the author conducted expert interviews in the form of 1:1 interviews and came up with a revised scale (the number of experts participating in the interview was 7 experts, the experts selected were those with experience in online marketing at travel companies). Basically, the revised scale is quite similar to the original scale, with only a few minor changes. The experts also suggested using words that are easier to understand and more friendly to the survey subjects.
In addition, chapter 3 also presents the sampling method, the number of survey samples (321 observations), the steps of analysis and statistics to produce research results, and the conditions for valid data analysis in SPSS. Research design and determination of research methods, establishment of the original scale, and methods of data processing are the prerequisites for analyzing the data in chapter 4.