Key points are not available for this paper at this time.
ABSTRACT A model of determinants of repeat purchase intentions of consumers who have previously bought online is developed and empirically tested. Data for the model test comes from responses to a structured self-administered survey provided by a sample of 436 consumers in Saudi Arabia. The results confirm hypothesized positive effects of overall satisfaction with previous online purchases and attitude toward online purchasing on repeat purchase intentions. In turn, attitude is positively determined by overall satisfaction and negatively by experience with online purchase problems, while overall satisfaction is determined positively by satisfaction with the following online retailers' performance dimensions: product prices, product quality, customer service, required payment methods, perceived payment security, and delivery time. Contrary to expectations, delivery cost is not significantly related to overall satisfaction. Neither does experience with online purchase problems. Theoretical and marketing strategy implications of the findings are outlined and discussed. KEYWORDS: Online shoppinge-shoppingrepeat purchase intentionssatisfactionattitudese-commercee-satisfaction ACKNOWLEDGMENTS The findings reported in this article are part of results from a study that has been funded under the CIM Accreditation Fund of King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia. The author acknowledges KFUPM for the support, and for using its various facilities in the research and preparation of this manuscript. Notes 1. Students were used to conduct the personal interviews because it is the best way to reach female respondents who are otherwise impossible to reach for data collection in Saudi Arabia because of the strict separation of sexes. 2. Books, computer software, hotel reservations, and video/musical CDs were the dominant products that respondents reported having ever purchased. 3. Least squares regression was chosen over covariance structure analysis because the preponderance of single-item measures is likely to produce unstable parameter estimates in structural equation modeling (J. C. CitationAnderson and Gerbing 1988). 4. For a simplified exposition of the four-step procedure for testing mediation, see also http://davidakenny.net/cm/mediate.htm 5. See http://www.danielsoper.com/statcalc/calc31.aspx for a calculator of the Sobel test statistic.
Alhassan G. Abdul‐Muhmin (Tue,) studied this question.