卡方自动交叉检验在人群细分中的应用

Application of CHi-squared Automatic Interaction Detector in Segmenting Subgroups

卡方自动交叉检验（CHi-squared Automatic Interaction Detector， CHAID）是一种定性的统计分类技术，主要解决根据一个因变量的不同反应确定若干预测变量的特征问题。这种算法模型可广泛应用于社会调查和市场细分中，根据不同的目的细分人群。该文主要介绍了CHAID算法模型的概念和发展，理论基础和应用过程，以及其效度检验方法，并将其与Logit模型等作了比较，指出其优势和局限。最后提出了研究展望。

CHi-squared Automatic Interaction Detector (CHAID) algorithm model is applied in social surveys and market segmentation. It is mainly used to classify cohorts by determining the characteristics of several predictors based on a dependent variables. This review article introduces the core concepts， operational precede， and development of the CHAID algorithm model. Compared with Logit model， this article also discusses the advantages， disadvantages， and restrictions of CHAID model. Furthermore， this article points out some questions for further research for the model.