More retail and service firms are using mystery shoppers in Taiwan to monitor their frontline operations and to assess their customer service, since the concern over whether their customers are truly satisfied is increasing. However, virtually nothing is known about whether the quality of the data collected by professional mystery shoppers is different or more reliable than that of shoppers disguised by general customers. Here we use a traditional statistics approach to assess the quality of data collected from both professional mystery shoppers and general customers. Feedback questionnaires from general customers and the potential use of mystery shoppers are made as a means of evaluating the service quality of cram schools in Taiwan. The purpose of this thesis is to show that the reliability of mystery shopping data made from professional mystery shoppers is higher than that of data made from general customers. With this question in mind, we conducted a two-step empirical research, as the amount of samples needed was insufficient. First, a primary study to examine the equality of population among branches was carried out in order to pool data from different branches. Then, the relative reliability of the surveys of mystery shoppers to general customers was examined by pooling data of the branches together. The findings provide evidence to suggest that reliable information could be obtained by pooling both shopper reports together. This empirical result also implied that well-trained mystery shoppers can make more objective observations and more effective assessments than general customers.