下周星期五（6月26日）下午我们荣幸地邀请到到纽约大学商学院杨沙教授与我院师生交流座谈，交流内容分为两部分：
1.16:00pm-16:45pm, 在明德主楼1016与我院老师与学生进行座谈
2.16:45pm-17:30pm在1016为大家做学术报告。
报告题目：
Modeling the Under Reporting Bias in Panel Survey Data
摘要：
Survey has been one of the most widely used marketing research methods to study consumer attitude and behavior. However, one challenge of estimating econometric models based on panel survey data is the underreporting bias. There has been empirical evidence suggesting that respondents may underreport their behavioral incidence when the data recording mechanism is tedious and the probability of underreporting is likely to increase over the duration of the survey period. If underreporting occurs, it will be difficult to study respondents’ true behavior. In this paper, we propose a Hierarchical Bayesian model to simultaneously study the underreporting bias (i.e. behavior occurs but was not recorded) and the true behavior of interest (i.e. the incidence of beverage consumption). The essence of the proposed method lies in modeling the actual behavioral incidence as latent variables, and the Gibbs Sampler makes it convenient to impute these non-reported consumption incidences along with making inferences on other model parameters. There are two direct benefits from the proposed method. First, it offers a model-based approach to remove the underreporting bias in panel survey data, and therefore allows marketing researchers to make accurate inferences on consumer actual behavior. Second, the method also offers a natural way to study factors that influence the propensity to underreport. Since we are treating those underreported behavioral incidences as non-missing-at-random, this underreporting propensity varies across respondents and over time. This understanding can help marketing researchers design the right strategy to intervene and incentivize respondents to actually report and hence improve the quality of the panel survey data. The proposed model and estimation approach are tested on both synthetic data and an actual panel survey data on consumer reported beverage-drinking behavior. Our analysis suggests that underreporting can significantly mask respondents' true behavior.
个人简历敬请参考：http://w4.stern.nyu.edu/faculty/facultyindex.cgi?id=271
