population, elderly people are underrepresented in the data and the educational level was above average among the respondents. However, since we are investigating online grocery buying behaviour, not offline grocery buying behaviour, we decided not to adjust for demographic differences between online and ‘traditional’ groups. This is in line with the suggestion that adjusting online results to match phone or mail results sends the wrong message to the users of the online research results as such results do not reﬂect the actual population studied (Miller, 2001). Measures Multiple item scales were developed for each of the seven constructs shown in Figure 1. With the exception of the measurement of repeat online grocery buying intention, all responses were measured on five-point Likert scales. On each of the multi-item scales a three-step item purification procedure was accomplished. First , inter-item and item-to-total correlations were computed for each item. All items should have a significant correlation coeﬃcient at the 0.01 level. Second , Cronbach’s alpha was computed for each of the constructs. In cases of a low alpha value the lowest item-to-total correlation was removed. Third , an exploratory factor analysis was conducted for each of the constructs using an eigenvalue of 1.0 as the cut-off point. The result of this procedure revealed two items as candidates for removal from the measurement scales. When these two items were deleted, all constructs showed Cronbach alphas 4 0.60, and all inter-item and item-to-total correlations were significant at the 0.01 level. The deleted items are marked with an * below. In developing some of the measurement scales we draw on Verhoef and Langerak (2001). These authors have successfully adapted two items from Eastlick (1993) to measure physical effort, three items from Srinivasan and Ratchford (1991) to measure time pressure, and three items from Dawson et al . (1990) to measure shopping enjoyment. Physical effort : (1) I think the transportation of my purchased grocery products is hard; and (2) the transportation of my bought groceries is exhausting. Perceived time pressure : (1) I am often in a hurry when I buy groceries; (2) usually when shopping groceries there is so much to do that I wish I had more time; and (3) shopping groceries via the Internet is favourable as it makes me less dependent of opening hours*. Shopping enjoyment : (1) I like shopping in 102 T. Hansen Downloaded by [University Library Utrecht] at 14:27 04 August 2013
supermarkets; (2) I really enjoy visiting different supermarkets; and (3) I consider grocery shopping as something to be done with and to finish*. To measure perceived complexity we draw on Verhoef and Langerak (2001) among others: (1) In general, online grocery buying is very complex; (2) with electronic shopping of groceries it is diﬃcult to order products; and (3) it is diﬃcult to receive groceries purchased via the Internet and to have them home delivered. In measuring perceived Internet grocery risk we draw on numerous writers, including Belanger et al . (2002) and McKnight et al .
You've reached the end of your free preview.
Want to read all 23 pages?
- Spring '14
- The Land, University Library Utrecht