國
立 政 治 大
Ᏸ
•
N a tio na
l C h engchi U ni ve rs it y
25
H5: Internet user experience is positively related to online purchase.
H6: Heavy Internet users buy online more often than regular Internet users.
Overall, Figure 3.2.1 shows the relationship among variables and the stated hypothesis.
Figure 3.1.2. Research Framework with Hypotheses.
3.3. Development of the Measurement
To examine the model it is needed to check the relationships between each related factors and to confirm or deny hypotheses. All the hypotheses contained in this paper are supported by theories or references. To measure the relationship between variables the various data about the consumer online behavior was collected. An online survey questionnaire was developed to obtain the responses from online users about their opinion on Internet-based purchasing. The questionnaire was anonymous.
•
Table 3.3.1. Operationalization of variables and their sources
Construct Item Source
Internet experience
1. How long have you been going online?
Less than 6 months
2. How many hours each week do you spend online?
1. Have you ever shopped online?
2. On average, how often do you buy online?
Never/almost never Less than once a month About once a month,
•
E1. Using the Internet to purchase would be easy to learn for me.
E2. Using the Internet to purchase would be easy to do for me.
E3. Using the Internet to purchase would require a lot of mental effort.
E4. Using the Internet to purchase would be easy following the instructions provided in virtual shops.
Using the Internet to purchase…
U1. Using the Internet to purchase would make shopping easier
U2. Using the Internet to purchase would enable me to shop more quickly
U3. Using the Internet to purchase would be useful to get better purchases
U4. Using the Internet to purchase would service would be a good idea.
A2 . The idea of using the Internet to buy a product of service is appealing.
A3 . I like the idea of buying a product or traditional shop to purchase goods.
I2. In the next 3 months I plan to use the
•
國立 政 治 大
Ᏸ
•
N a tio na
l C h engchi U ni ve rs it y
28
3.3.2. Dependent Variables and Independent Variables
The research involves four dependent variables: usefulness, attitude toward online purchasing, online purchase intention and online customer behavior, which is the actual purchase on the Internet. There are two independent variables in the research model: ease of use and Internet experience. The research also reports the influence of consumer’s demographics and consumer’s Internet experience on online behavior (independent variables).
3.3.3. Questionnaire
The content of the questionnaire, including the ranging or the scale, is shown in Appendix1. All questions in the survey were required in order to avoid missing variables. If the respondent did not answer one of the questions, he/she was not able to finish the survey. During the survey, only those respondents who checked a positive answer regarding the question “Have you ever shopped online?” could proceed to the next questions about online purchase experience. Those respondents who have never purchased any goods or services online were moved to the next section of the questionnaire, namely personal data. Only those respondents who have bought products or services on the Internet were studied.
The questionnaire consisted of six constructs (composed in total of 16 items):
1. Internet experience – composed of two questions about Internet usage experience in years and the user’s frequency of Internet use in hours per week;
2. Online purchase frequency – composed of two questions about the previous buying experience and frequency;
•
usefulness of Internet shopping. Items were validated by Davis (1989) and Crespo et al. (2009);5. Attitude towards Online Purchasing– composed of three questions about the idea of buying online. Items were validated by Heijden et al. (2001;
2003) and Limayem et al. (2002). Seven point Likert scale ranging from
“strongly agree” to “strongly disagree” is used to measure the items.
6. Online Purchase Intention – composed of three questions about the Internet-based purchase intention in the next 3 months. Items were validated by Mathieson (1991), Taylor and Todd (1995), Limayem, et al.
(2002), Heijden et al. (2003) and Crespo et al. (2009). Seven point Likert scale ranging from “strongly agree” to “strongly disagree” is used to measure the items.
The author found it relevant to also include personal data to the questionnaire in order to make sure that our sample contained respondents from different groups.
Demographical variables, such as: age, gender, education, occupation and household income, were collected to evaluate average online user and to check if they influenced purchasing behavior on the Internet.
A single item that asked the respondent his/her age range measured age: “less than 18”, “18 – 24”, “25 – 34”, “35 – 44”, “45 – 54”, “55 – 64” and “65 or more”.
Respondents reported their gender on a single-item dichotomous variable that asked them whether they were male or female. Education was measured by a single item that asked about education level of respondent: “elementary”, “secondary” and “higher
•
國立 政 治 大
Ᏸ
•
N a tio na
l C h engchi U ni ve rs it y
30
education”. A single item measured occupation: “student”, “unemployed”, “IT related” and “non IT related”. Household income was measured by a single item that asked the respondent his monthly income in PLN. The scale ranged from “no income”, “below 1500 PLN (500 USD)”, “1500 PLN - 3000 PLN (500 USD - 1000 USD)“, „3000 PLN - 5000 PLN (1000 USD - 1750 USD)”, “5000 PLN - 10 000 PLN (1750 UDS - 3500 USD)” and “above 10 000 PLN (3500 USD)”. For the purpose of the study the author decided to utilize average exchange rate to present the data in USD.
3.3.4. Validity and Reliability
Validity means the correctness of the instrument. It maybe considered as the degree that the scale can really testify the variable quality, which is, the efficiency of the questionnaire. Content validity requires the researcher to use the correct measures for the concepts being studied. In the present study, the variables and the scales are derived from related theories and existing literature so the level of validity can be assumed. Thus, content validity of the research can be ensured.
Reliability represents the extent to which other researchers would have the same findings if they were to analyze the same data and methods (Yin, 2003). All questionnaire items were tested using Cronbach’s alpha coefficient, which is the most frequently used estimate of internal consistency. Thus, items demonstrate internal consistency reliability Cronbach’s alpha, and also internal and external validity, according the previous researches.
•
國立 政 治 大
Ᏸ
•
N a tio na
l C h engchi U ni ve rs it y
31