Hypothesis Testing with Scanner Data: The Advantage of Bayesian Methods
提出一种贝叶斯方法，用于检验多元线性模型中的非嵌套约束，并用商店级扫描数据与经典方法对比，发现贝叶斯检验在客观性、易用性和解释性上更优。
The author presents a Bayesian method of testing possibly non-nested restrictions in a multivariate linear model and, using store-level scanner data, compares it with classical methods. The Bayesian tests are shown to be either equal or superior to classical tests in terms of objectivity, ease of use, and ease of interpretation. Classical tests lack a natural metric for comparing non-nested models and often employ super models in which the entertained hypotheses are special cases (i.e., nested). Nested classical tests are almost always significant when used with scanner data, making their interpretation problematic. In contrast, large samples cause Bayesian methods to become less dependent on subjective aspects of the prior distribution and therefore more objective.