Prediction Modeling for Mardet Segmentation of Hotel and Casual Dining Restaurant in Korea : a CHAID Analysis

With CHAID (Chi-squared Automatic Interaction Detector) analysis this paper is to establish prediction modeling for segmentation of the food service industry in Korea. Unlike other methods such as discriminant analysis for the market segmentation, CHAID analysis is an algorithm to practice multi-way split using Chi-square test or F-Test helping researchers to create prediction model to find out a group with the best results or compare differences between groups. This study is conducted to 540 people who experienced 12 hotel restaurants and 5 casual dining restaurants in Seoul, 2003. As a result, this study finds that a group of people with age younger than 30, less than 60 million won annual income and social gathering purposes is the target market of the hotel and casual dining restaurants. Rather than rely on grouping consumers based on their similarity on a profile of characteristics, CHAID analysis refines segments using similarity of response, and is recommended for evaluating segments.