Validating the search, experience, and credence product classification framework in a model of patronage intentions

Although the Internet as a shopping medium provides convenience to easily access products, the U.S. e-commerce retail sales still make up a very small percentage of the U.S. total retail sales. To better understand what influences consumers' choices to shop for products or services on the Internet versus local retail stores, this study tests the influence of antecedent factors of consumer patronage intentions for Internet and local retail stores. The study draws the antecedent factors from the extant literature, which include product classes, the importance consumers place on retailer attributes, and consumer perceived risk in product classes. Because the Internet instituted a convenient shopping medium with information distribution and search capabilities, nomological validity of the search, experience, and credence (SEC) product classification framework is tested in the online shopping context. This study tests the validity of the SEC product classification framework by examining whether significant differences exist in (1) the level of importance that consumers place on retailer attributes, (2) the amount and type of risks that online shoppers perceive in product classes (search, experience, and credence), and (3) their patronage intentions for two retailer types---Internet and local retail stores---based on product classes. In the same model, the study also tests the mediating effects of perceived risk in product classes in the relationship between the importance of retailer attributes and patronage intentions for retailer types. Although the relationships between some of the antecedent factors of patronage intentions have been tested in previous studies, they have never been tested jointly in the context of Internet shopping. Data were collected in three stages. The first two stages were the pretest studies that were conducted to select products as examples to represent each product category. The hypotheses were tested using data collected from a nationwide survey of those who previously purchased products or services on the Internet. The results of the analyses support the hypotheses.