Chapter 3 Research Method for B2C Mobile Tourism Service Classification
3.3 Research Method for B2C Mobile Tourism Service Evaluation
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
19
3.3 Research Method for B2C Mobile Tourism Service Evaluation
During stage 3, the aim was to evaluate the revised list of B2C mobile tourism services as well as the proposed B2C mobile tourism service classification framework with multiple perspectives. The sample population was tourists and two key B2C mobile tourism service providers including 3G operators and travel agencies. Three surveys for the B2C mobile tourism service evaluation included an online survey of the tourists, another online survey of the 3G operators, and a mail survey of the travel agencies. The questionnaire was written in Chinese. Prior to the survey was distributed, the wordings of each item in the questionnaire were discussed during the semi-structured group interviews. Moreover, the questionnaire was pre-tested, using a small convenience sample of twenty respondents to detect any logical errors of questions. The survey was lasted for 42 days - from January 11 to February 19, 2007.
Taiwan is an appropriate empirical setting for B2C mobile tourism service evaluation surveys from both telecommunication and tourism perspectives. From the telecommunication perspectives, according to the telecommunication figures in 2011 released by National Communications Commission (National Communications Commission, 2012), there are 29 millions mobile phone subscribers, and among them, 20.9 million are 3G subscribers. Among all mobile subscribers, 71.3% are mobile Internet subscribers. That is, the mobile Internet access is highly used by Taiwan residents, which makes Taiwan an appropriate empirical setting for B2C mobile service evaluation survey. From the tourism perspectives, according to the 2010 survey of travel by Taiwan citizens released by the Tourism Bureau (Tourism Bureau, 2011), 93.9% Taiwan citizens take domestic travels, and 20.1% take outbound travels.
The average number of domestic trips per person is 6.08 trips while the average number of outbound trips per person is 0.41 trips. That is, most Taiwan citizens have either domestic or outbound travel experience, and it makes Taiwan an appropriate survey site for tourism survey. Based on these two perspectives, we conclude that Taiwan is an appropriate empirical setting for B2C mobile tourism service evaluation survey.
The first survey was an online survey for Taiwan tourists who had domestic and/or oversea travelling experiences, better with mobile Internet experiences. We chose the convenience sampling method, and the website address of the online B2C mobile tourism service survey was emailed to one class of the part-time college students who had at least one-year full-time working experience, the full-time graduate students of the Department of Management Information Systems, the members of the Information Management Association in Taiwan, and the authors’ personal networks. The respondents were encouraged to forward the online survey to other qualified respondents. There were total 325 effective responses from the tourists.
The second survey was another online B2C mobile tourism service survey for the 3G operators. We targeted three major 3G mobile companies, and they were Chunghwa Telecom Co., Taiwan Mobile Co., and Far Eastone Telecommunication Co. We chose the convenience sampling method, and the online survey website address was forwarded to the contact windows of the three companies, and they promised to forward this online survey to their
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
20
coworkers whose jobs were related to the planning, management, and operations of mobile services. There were total 86 effective responses from the mobile industry.
The third survey was a mail survey to the travel agencies in Taiwan. The mailing list of the travel agencies was downloaded from the official website of the Tourism Bureau, Ministry of Transportation and Communications in Taiwan. There were three different travel agency licenses: general, type-A, and type-B. We chose the stratified random sampling method, and 300 travel agencies were randomly selected from 2708 travel agencies. There were 132 return surveys from the travel agencies, and a 44% response rate was obtained from the survey of the travel agencies. Among the 132 returns, 40 questionnaires were with too many missing data; so, remaining 92 questionnaires were usable, representing an effective response rate of 31%, which was satisfactory. Moreover, as shown in Table 2, among the 92 usable questionnaires from the travel agencies, 11 (11.96%) responses had general travel licenses, 76 (82.61%) responses were with type-A travel license, and 5 (5.43%) were with type-B travel licenses. The ratios of effective returns with different travel license types were close to the ratios of the mail surveys and the ratios of the industry profiles. So, the effective returns were representative of the tourism industry.
Table 2. Return Profiles of the Travel Agencies of B2C Survey
Travel License Type Sample Base Mail Survey Effective Return
N % N % N %
General
275 10.16 30 10.00 11 11.96
Type-A
2281 84.23 252 84.00 76 82.61
Type-B
152 5.61 18 6.00 5 5.43
Total
2708 100.00 300 100.00 92 100.00
Table 3 showed the outline of survey questions of the B2C mobile tourism service evaluation by tourists. The survey questions are extracted from literature reviews, summarized in Table 1. In part one, there were six B2C mobile tourism services from our proposed B2C mobile tourism service classification framework with twenty-one questions.
We had five questions for mobile search & notification services, two for mobile community services, three for mobile guide services, six for mobile recommendation services, two for mobile auction services, and three for mobile transaction & payment services. In part two, we wanted the tourists to evaluate the overall usefulness of the proposed B2C mobile tourism services. In part three, we had five user background information questions. The background data of the respondents were limited to travel experiences, 3G mobile Internet experiences, gender, age range, and educational level for minimum background analysis. So, there were total twenty-seven questions in the B2C mobile tourism service evaluation survey, including twenty-one in part one, one in part two, and five in part three. For the user background information, we avoided sensitive personal data like ID number or income. Anonymity was ensured for surveys, as the questionnaires were not traceable to the respondents. We wanted the tourists to evaluate the usefulness of the proposed B2C mobile tourism services. The 5-point Likert scale was used, ranging from “1=not useful at all” to “5=very useful”. The tourist survey (Chinese version) was presented in Appendix 1.
‧
Table 3. Survey Questions of the B2C Mobile Tourism Service Evaluation by Tourists Part I: Evaluation of B2C mobile tourism service
No Mobile services Questions Usefulness
1 Mobile search &
notification services
(Q1) Safety-related information services (Q2) Road condition & weather report services (Q3) Event schedule change notification services (Q4) Major pre-trip information services
(Q5) Reminder services 2 Mobile community
services
(Q6) Tour information inquiry services (Q7) Tour information sharing services 3 Mobile guide
services
(Q8) Pre-recorded mobile guide services (Q9) Interactive mobile guide services (Q10) Theme-based video and audio services 4 Mobile
recommendation services
(Q11) Hotel recommendation services (Q12) Sightseeing recommendation services (Q13) Restaurant recommendation services (Q14) Tour plan recommendation services (Q15) Tour route recommendation services (Q16) Souvenir store recommendation services 5 Mobile auction
services
(Q17) Mobile auction services
(Q18) Mobile reverse auction services 6 Mobile transaction
& payment services
(Q19) Mobile booking/reservation services (Q20) Mobile payment services
(Q21) Mobile ticketing services
Part II: Overall evaluation of the B2C mobile tourism services
The overall usefulness rating of the proposed B2C mobile tourism servicesPart III: User background
1 What are your travelling experiences (multiple choices)? Domestic budge travelling, domestic group travelling, oversea budget travelling and/or group travelling experiences 2 Did you ever access Internet via WAP, GPRS, 3G or PHS? Yes or no
3 Gender: Female or male
4 Age: Below 20, 20-29, 30-39, 40-49, 50-59 or above 60
5 Educational level: High school, junior college, college or university, master or PhD Table 4 showed the survey questions of the B2C mobile tourism service evaluation presented to the business sides, including the 3G operators and the travel agencies. The survey questions for service providers were the same as the survey questions for tourists, based on literature reviews shown in Table 1. While the tourists evaluated the usefulness of the proposed B2C mobile tourism services, we wanted the business sides to evaluate the importance, feasibility, and cost-benefit effectiveness of the proposed B2C mobile tourism services. The 5-point Likert scale was used, ranging from “1=not important at all/not feasible at all/not cost-benefit effective at all” to “5=very important/very feasible/very cost-benefit effective”. We took (Q1) safety-related information services for example. When it came to the importance dimension, a respondent strongly agreed that the safety-related information services were very important services, and he/she chose 5 to rate the importance of the safety-related information services. However, when it came to the feasibility dimension, he/she strongly disagreed that the safety-related information services were very feasible, and
‧
he/she chose 1 to rate the feasibility of the safety-related information services. He/she held neutral attitude toward the cost-benefit effectiveness of the safety-related information services, and he/she chose 3 to rate the cost-benefit effectiveness of the safety-related information services. The Chinese version of the survey questions for 3G operators and travel agencies was presented in Appendix 2-3.
Table 4. Survey Questions of the B2C Mobile Tourism Service Evaluation by Service Providers
Part I: Evaluation of B2C mobile tourism service
No Mobile services Questions Importance Feasibility Cost-benefit
Effectiveness 1 Mobile search &
notification services
(Q1) Safety-related information services (Q2) Road condition & weather report services
(Q3) Event schedule change notification services
(Q4) Major pre-trip information services (Q5) Reminder services
2 Mobile community services
(Q6) Tour information inquiry services (Q7) Tour information sharing services 3 Mobile guide
services
(Q8) Pre-recorded mobile guide services (Q9) Interactive mobile guide services (Q10) Theme-based video and audio services
4 Mobile
recommendation services
(Q11) Hotel recommendation services (Q12) Sightseeing recommendation services
(Q13) Restaurant recommendation services
(Q14) Tour plan recommendation services
(Q15) Tour route recommendation services
(Q16) Souvenir store recommendation services
5 Mobile auction services
(Q17) Mobile auction services (Q18) Mobile reverse auction services 6 Mobile
(Q20) Mobile payment services (Q21) Mobile ticketing services
Part II: Overall evaluation of the B2C mobile tourism services
The overall importance, feasibility, and cost-benefit effectiveness ratings of the proposed B2C mobile tourism services
Part III: User background
1 What are your travelling experiences (multiple choices)? Domestic budge travelling, domestic group travelling, oversea budget travelling and/or group travelling experiences
2 Did you ever access Internet via WAP, GPRS, 3G or PHS? Yes or no 3 Gender: Female or male
4 Age: Below 20, 20-29, 30-39, 40-49, 50-59 or above 60
5 Educational level: High school, junior college, college or university, master or PhD
There was no lie detector in our surveys because the employees in the mobile and tourism industries were professional workers. There was no prize for filling the questionnaire, and it avoided the unusable replies. For online survey, there was no default answer, and the
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
23
respondents needed to complete the whole on-line questionnaire before they clicked the submit button. So, there was no missing data problem for online surveys, but there was possibility for double clicks for online surveys. The website was design to have a thank notice to avoid accident double clicks. Also, IP and completion time were recorded in the online survey for further double checking for double-click problems. We did not filter the same IP address because the IP pooling was used for most companies and Internet service providers.
As for mail survey, each questionnaire’s mean value was filled as a substitute of missing data if there were no more than three missing values. Questionnaire with too many missing data were deemed unusable.
Reliability is tested in terms of internal reliability, construct reliability and variance extracted. Internal reliability is measured using Cronbach’s alpha. 0.60 is satisfactory for exploratory studies, and 0.70 is the generally agreed upon lower limit for Cronbach’s alpha (Hair et al., 2010). Construct reliability and variance extracted are calculated from standardized loading and measurement error for each indicator. A commonly used threshold value for acceptable construct reliability is 0.70, and guidelines suggest that the variance extracted value should exceed 0.50 for a construct (Hair et al., 2010).
Validity is assessed in terms of content and constructs validity. Survey questions extracted from literature reviews and refined by group interviews ensure the content validity of the instruments. Construct validity is established by measuring the convergent and discriminant validity of the survey items. When survey items with the same corresponding constructs are significant correlated in the item-to-item correlation matrixes, it indicates the measures satisfy convergent validity. In terms of discriminant validity, when none of the off-diagonal correlations among constructs are higher than the values of the corresponding internal reliability of the construct, it indicates strong discriminant validity. Factor analysis is also performed to validate the construct validity. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy (MSA) is used to determine the appropriateness of applying factor analysis. Values above 0.80 for the factor matrix indicate that the use of factor analysis is appropriate (Hair et al., 2010).
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
24