PURPOSES : This study explores the preference of shared autonomous vehicle service in an underground dedicated environment. METHODS : A stated preference survey was conducted to examine the mode choice behaviors on autonomous vehicle service competing with existing modes. Multinomial logit was employed to estimate the parameters of explanatory variables from the surveyed data. The model was estimated with alternative specific parameters rather than generic parameters. The value of time was also estimated using the parameters of the mode choice model. RESULTS : The results showed that the travel cost had the highest sensitivity to public transportation and the lowest to private cars. We also found that the value of the in-vehicle travel time was highest for private cars, lowest for public transportation, and intermediate for SAVs, suggesting that SAVs could serve as a premium public transport option. Additionally, the out-vehicle time coefficient was higher for public transportation compared to that for SAVs, indicating that users are more willing to tolerate longer out-vehicle times for SAVs due to their high-speed service compared to that of public transportation. CONCLUSIONS : This study presents a direction for policy regarding the adoption of shared autonomous vehicle services by considering the attributes that are valued by users of each mode.

While machine learning has gained popularity in choice behavior modeling, most machine learning models are often complex, difficult to interpret, and even considered as black box. This study investigates machine learning methods for choice behavior modeling that provide interpretability of models’ output. We explore various approaches including (1) explicitly descriptive models such as tree-based models, (2) interpretation of predictive models through feature importance measures, and (3) recent advancements in prediction explanation methods such as LIME and SHAP (Shapley Additive exPlanations). We demonstrate the methods on consumers’ airport choice behavior in Seoul metropolitan area. Through the comparative analysis with traditional discrete choice models, we discuss advantages as well as limitations of machine learning models in consumer choice behavior modeling.

A consumer psychologist observes someone make a choice between chocolate cake and fruit salad, a classic measure of self-control (Shiv & Fedorikhin, 1999). Since the cake is relatively tastier but unhealthier (i.e., a vice; Wertenbroch, 1998), a choice of cake is interpreted as a lack of self-control, whereas a choice of fruit salad (a virtue) is viewed as successful self-control. The simplicity and intuitive appeal of this heuristic has led it to be widely adopted by consumer psychologists. However, a comprehensive understanding of the operation of self-control requires food consumption to be analyzed as a two-stage process: choice followed by intake quantity (Wansink & Chandon, 2014). Self-control processes may be relevant to both stages of a decision of what to eat (choice stage) and a decision of how much to eat the chosen option (quantity stage)—they may influence both choice and quantity. Examining the role of self-control in terms of post-choice quantity control is critical since over-consumption of calories is the single most significant contributor to obesity (Livingston & Zylke, 2012). The overconsumption problem is not an exclusive matter of vice foods. Eating virtuous foods can result in excessive calorie intake and thus consumers need to control the intake of virtues as well as vices.

This article focuses on the use of the cosmetics channel of customers. There are many papers about channel choice behavior of consumers. Blattberg et al (2008) pointed out that the consumer decide their channel by marketing action, channel attribute, social influence, channel integration, individual difference, and situation factors. Kushwaha, et al (2013) said that in the case of buying an axiomatic product, consumers who are using only Web are excellent customers, whereas consumers using many channels are excellent customers in the case of a hedonistic product. Neslin, et al (2014) implied that the channel and brand are interrelated, so the customer think about both channel choice and brand choice at same time. Valentine et al (2014) found that multi-channel user is more information sensitive than that of single-channel user, and customer who are satisfied the service of their chosen channel do not try to use new channel. From these papers, it is clear that consumer’s channel choice behavior depends on, not only channel specialty, but also product specialty. On the other hand, there are not many studies about consumer behavior of cosmetic buying. Bloch et al (1992) implied that makeup made person more active, and an evaluation from another person lead to self-satisfaction. Fabricant (1993) pointed out that the purpose of makeup is self-expression and hiding one's fault. From the research of Coulter, et al (2000), opinion leader are more interested in cosmetics and much quantity of purchasing. Because cosmetic is very image oriented products, customer’s information sensitivity and cosmetic lifestyle are important to buy cosmetics Gathering data about channel choice behavior of consumers in Japan, it becomes clear that stylish and fashionable customers are willing to buy at department stores as their priority is channel images, while non-fashionable customers tend to buy at drugstores and GMS, as their priority is low price. It imply that stylish and fashionable score, cosmetic lifestyle of customer, and channel image are important factors for channel choice behavior of customer. Some customer use both department stores and drugstores, and some customer use both GMS and drugstores. Customers who use both department stores and drugstores are almost same lifestyle of customers who use department stores only, otherwise lifestyle scores of customers who use both GMS and drugstores are, almost same as drugstore only customers. It means that customer of department stores use drugstore as the sub channel of department stores, but customer of GMS use drugstore as the main channel. We must take care about customer combination of store channels. It becomes clear that channel images and customer lifestyle are very important to explain consumer’s channel choice.

The purpose of this study was to analyze the difference in bakery choice attributes according to consumers' general characteristics and purchasing behavior. Among 350 questionnaires distributed to bakery consumers, 277 complete questionnaires (79.1%) were analyzed. Bakery choice attributes were classified into five factors: "environment and image", "bakery product features", "location", "employee service", and "price and sales promotion"; the mean scores of these factors' importance levels were 3.59, 3.58, 3.49, 3.36, and 3.00, respectively. Males considered 'employee service' factor significantly more than did females. Further, the importance level of 'employee service' factor was significantly greater as consumer's age increased. The importance levels of 'bakery product features' and 'employee service' factors were considered significantly more by consumers who spent KRW10,000-15,000 than those who spent KRW5,000 or less. 'Price and sales promotion' was considered to be more important by consumers who obtained information from the Internet than from the TV and radio. 'Location' factor was considered to be more significant as purchasing frequency increased. Such differences in importance level of bakery choice attributes according to consumers' gender, age, job, and purchasing behavior should be considered and applied to the development of marketing strategies targeted at consumers.