Participatory sensing, which collects environmental information from mobile phone users, is growing in popularity. The collected information can be used for national policy or decision-making for companies. However, sensing users may violate their privacy. Recent studies have proposed negative surveys which can analyze the attributes of users statistically without precise information about each user's information. The traditional negative surveys need a lot of samples for proper estimation. These days, several types of negative surveys are used that can estimate the distribution of user attributes with a high degree of accuracy. However, privacy levels of these methods are relatively low. Moreover, existing studies assume that the privacy levels of all users are the same. In this paper, we propose a new negative survey that can estimate data distributions with more precision and can be used in a situation where the privacy levels are different based on each user's demand. By simulations of a synthetic and a real data set, we prove that our proposed method can estimate more precisely than existing methods.