Method

We pre- tested and refined a questionnaire that was distributed to students in basic marketing classes at a large U.S. university. The questionnaires were completed during a class period for extra credit. Although this is admittedly a convenience sample, these consumers are appropriate for the topic, and the data should be suitable for use in a theory testing study such as this one (Calder, Philips, and Tybout, 1981). Completed questionnaires were collected from 316 students.

In an attempt to enhance data quality, two items were included in the questionnaire to guard against random responding (Dollinger and DiLalla, 1996, pp. 169-171). A list of scale items not used for the present study contained the item: 'I have tried to answer all of these questions honestly and accurately' with a seven-point response format where 1 = Always False and 7 = Always True. A second scale contained the item: 'If you read this item, do not respond to it,' with a seven-point response format were 1= Strongly Agree and 7 = Strongly Disagree . The eleven respondents who responded with a 1 or 2 to the first item, who recorded a response to the second item, or who left most of these scales incomplete were removed from the data set as 'bad' respondents, leaving 305 usable questionnaires. Owing to missing data for key scales , two more questionnaires were eliminated so that the effective sample was 303.

The sample consisted of 135 (44.6%) men and 168 (55.4%) women. Their ages ranged from 19 to 37 with a mean of 21.1 years (sd = 2.33). There were 242 (79.9%) whites, 31 (10.2%) African-Americans, 17 (5.6%) Hispanics, and 13 (4.3%) others. There was no statistically significant ( p < .05) difference in mean age between the men and women or across the four ethnic categories, nor was there a significant association between race and gender.

Multi-item scales were used to measure the three variables of interest (Table 15-1). These were chosen because they have been used in other studies and have demonstrated acceptable reliability and validity. Global innovativeness was measured by eight items from Hurt et al. (1977). Originally a 20-item measure, their scale (Hurt, Joseph, Cook or HJC) has been shown to be valid and reliable (Goldsmith, 1990; Pallister and Foxall, 1998) even in shortened versions such as used here. Eight items appeared in the questionnaire, and the scale was reduced to five items by factor analysis and reliability analysis. Domain-specific innovativeness (online innovativeness) was measured by the six-item Domain Specific Innovativeness Scale or DSI developed by Goldsmith and Hofacker (1991). Factor analysis and reliability analysis suggested that four of these would provide the best measure for this study. Online buying was measured by three items shown by Goldsmith and Lafferty (2001) to provide a reliable and valid measure of online purchasing. Since buying online is the innovative behavior of interest here, it represents the adoptive behavior that the model intends to explain.

Table 15-1: Scale items used in the study

Scale Item	Factor Loading
Hurt, Joseph, and Cook Global Innovativeness Scale a
I am willing to try new things.	.691 e
I am aware that I am usually one of the first people in my group to accept something new.	.620
I am generally open to accepting new ideas.	.617
I feel that I am an innovative person.	.549
I tend to feel new ways of living and doing things are improvements over the past.	.492
Domain Specific Innovativeness Scale b
In general I am the last in my circle of friends to know the names of the latest places to shop on the Internet. [reverse coded?]	.825 e
In general, I am among the last in my circle of friends to purchase something over the Internet.	.814
Compared to my friends, I do little shopping over the Internet.	.800
I know more about shopping over the Internet than other people.	.769
Online Buying
How often would you say that you purchase online? c	.940 f
Is this distinguishable from the one above? Asked another way, how often do you purchase online? d	.902
Since September 1, 2002, how many times have you purchased online?	.861
a 1 = strongly disagree, 2 = disagree, 3 = somewhat disagree, 4 = neither , 5 somewhat agree, 6 = agree, 7 = strongly agree. b 1 = strongly disagree, 2 = disagree, 3 neither, 4 = agree, 5 = strongly agree. c 1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = very often. d 1 = I never do; 2 = only once or twice a year; 3 = about once a month; 4 = less than once every two weeks, but more than once a month; 5 = only about once every two weeks; 6 = about once a week; 7 = more than once a week. e Loadings from a principal axis factor analysis f Loadings from a principal components analysis