Sample and data Collection
The sample consisted of respondents who shopped for clothing apparel at discount or department stores in Taiwan. The respondents were intercepted in front of these retail stores. Discount stores are retail stores that sell private brands, and have a higher merchandise turnover rate and limited services (e.g., mass merchandisers). Department stores are retail stores that provide customers a variety of merchandise selections and high-quality services (e.g., traditional department stores). The apparel industry has been blooming for the last two decades (Wigley and Moore 2007). With the increase in personal disposable income and industry development in Taiwan, Taiwanese consumers’ demand in the apparel industry is growing (Trappey and Lai 1996). The apparel retail market in Taiwan is profitable (Seock and Lin 2011) and increasingly competitive (RnRMarketResearch 2012). One of the research purposes of this paper is to understand how retailers can build their retail stores’ price images in order to attract consumers in a competitive retail market. Thus, the apparel retail market serves as a fine study context for our research purposes.
Two phases were included in survey development: (1) interview and focus group, and (2) survey pretest. In the first phase, we obtained the constructs and measurement items from the literature, and in-depth interviews were conducted with five experts to ensure their appropriateness in a retail context. Then, a focus group comprised of eight consumers was conducted (Dillon, Madden, and Firtle 1994). As such, we can ensure that the measurement items and each construct can be understood. The revised items were confirmed and discussed by experts. In the second phase, 42 respondents were intercepted for a personal interview survey in front of the retail store after their shopping trip for clothing apparel. As such, responses to questions regarding the retail store were more accurate. The questionnaire was pilot tested using a convenience sample (42 respondents). We chose to use bootstrap samples for the confirmatory factor analysis (CFA), as this method would not be restricted to the assumptions of normality, independence, and a large sample. The composite reliability (CR) and average variance extracted (AVE) of each construct exceeded 0.7 and 0.5, respectively. The results suggested that the constructs had convergent validity (Anderson and Gerbing 1988). All factor loadings were significant (p < 0.05). Thus, no items were removed from the questionnaire in the pilot test. Then, the questionnaire was officially distributed. To make sure that respondents were able to provide valid and reliable information regarding their attitudes toward the retail store, the interviews were conducted right after respondents’ shopping trips. The respondents were asked to respond to the questions focusing on the store where they shopped. A total of 659 respondents were then interviewed.