Crossover Linear Modeling: Combining Multilevel Heterogeneities in Crossover Relationships
提出交叉线性建模方法，用于处理组织研究中非完全嵌套的交叉结构数据，能同时识别不同层次的异质性，对研究招聘等场景的学者有用。
In organizational research, many multilevel structures are not fully nested but have a crossover structure. For example, in early recruitment study, the relationship between job applicants and recruiting companies has the crossover structure because each applicant is interested in applying multiple companies whereas each company is of interest to multiple applicants, and applicants and companies are not fully nested within each other. This article introduces the crossover linear modeling (CLM) method and argues that CLM is useful when dealing with multilevel crossover structures by combining the heterogeneities of all levels into one model. The article uses an example in early recruitment research to illustrate the use of CLM in modeling direct effects, cross-level effects, and interaction effects in a crossover data structure. The statistical analysis on an actual data set shows that CLM is able to identify both company-level and applicant-level heterogeneities. The article presents further examples to show that CLM can be applied to many organizational research settings when multilevel crossover structures are present.