Data Collection

---

The research questions were examined in the context of B2C e-commerce. A questionnaire was given out to experienced online shoppers—users of various websites. The respondents were instructed in the questionnaire to answer the questions by assessing the last website they bought at. Respondents were also asked to write down the name of that website. The items are shown in Table 1. The PU and PEOU items were adapted from Davis et al. (1989). The Intended Use items measured intended purchase and intended willingness to provide crucial information necessary for the vendor to provide service. Habit was measured with a set of items dealing with a habit to prefer a certain IT, in this case a website. All these items were measured on a seven-point scale ranging from strongly agree (1) through neutral (4) to strongly disagree (7). The questionnaire also collected demographic data. The questionnaire was pilot-tested by asking two experienced online shoppers to assess whether the items correspond to their related constructs.

Table 1: Item Loading and Error

Item		Std. Loading
PEOU
EO1	My interaction with the website is clear and understandable.	0.86
EO2	It is easy to interact with the website.	Dropped
EO3	The website is easy to use.	0.89
E04	It is easy to become skillful at using the website.	0.91
EO5	Learning to operate the website is easy.	0.91
EO6	The website is flexible to interact with.	Dropped
PU
PU1	The website is useful for searching and buying CDs/books.	Dropped
PU2	The website improves my performance in CD/book searching and buying.	0.85
PU3	The website enables me to search and buy CDs/books faster.	0.89
PU4	The website enhances my effectiveness in CD/book searching and buying.	0.94
PU5	The website makes it easier to search for and purchase CDs/books.	Dropped
PU6	The website increases my productivity in searching and purchasing CDs/books	0.85
Habit
HB1	This is where I am usually go to buying CDs/books.	0.83
HB2	This is my preferred online CDs/books vendor.	0.83
HB3	When I need to buy CDs/books online this is where I go first.	0.89
HB4	I often buy CDs/books online from this vendor.	0.84
Use Intentions
USE1	I would use my credit card to purchase from the online vendor.	0.85
USE2	I am very likely to provide the online vendor with the information it needs to better serve my needs.	0.80

The danger of a possible tautological relationship between Intended Use and Habit, cautioned by Thompson et al. (1994), was addressed by wording the Habit items to deal with the users' overall previous pattern of tendency and preference and wording the Intended Use items in terms of specific behavioral intentions referring to future activities. In this way, habit and behavioral intentions were worded to avoid having the two scales share the same meaning. This distinction was later confirmed in a Confirmatory Factor Analysis, discussed below.

The research questions were examined with undergraduate and graduate business students at a leading business school in the Mid-Atlantic region of the U.S. The questionnaire requested the students to relate to the last time they purchased a book or a CD online and to answer the remaining questions accordingly. The respondents were asked to complete a questionnaire during class time. There was no credit given for the task and the objective of the questionnaire was not revealed to the respondents. All the respondents reported that they were experienced online shoppers at B2C sites. Two hundred and ten students completed the questionnaire. Only the 179 questionnaires from students who previously purchased online were included in the study. The respondents had bought online on average 8.1 times in the previous year. Gender: 71 were women, 93 were men, 15 declined to state their gender. The respondents were mostly in their early 20s (n=132) or late 20s (n=26). The most popular sites mentioned by the respondents were amazon.com (37%), half.com (19%), bn.com (n=18%), cdnow.com (n=9%), bmg.com (n=6%), and ecampus.com (n=5%).

The data show that, on average, the respondents thought that the website they last bought at was very easy to use and useful across all the TAM proposed items of these constructs. The respondents also reported that they tended to use that website habitually and expected to go there again. The means and standard deviations are shown in the Appendix. Answers for all these items ranged from 1 to 7. The high correlations among PU, PEOU, and Intended Use items are common in TAM research (see Gefen et al., 2000).

There was no significant difference between the undergraduate and graduate students in their answers to the questionnaire items (Wilks' Lambda = .87678, p-value = .272), and so the data were combined for the analysis. There were also no gender effects in the data (Wilks' Lambda = .86345, p-value = .259), despite previous research on IT adoption in an organizational context that suggests such as effect (Gefen & Straub, 1997; Venkatesh & Morris, 2000), possibly because these effects come about through social interaction (Tannen, 1994), which is notably missing in online activity.

---