Building actionable theories: The role of causal constructs
区分了三种实用理论类型（操纵X、选择X、观察X），重点讨论操纵X理论中因果构念的标准和常见缺陷，帮助管理学者构建更可操作的理论。
What makes leadership and management theories practical? Theories are practical to the extent that practitioners can enact proposed cause (X) − effect (Y) relationships. Accordingly, I distinguish three types of practical theories. Manipulate(X) theories give practitioners levers for action. These theories have causal constructs, whose operationalizations’ levels practitioners can set by themselves. For example, in a theory on charismatic leader signals, practitioners can use fewer or more signals. In select(X) theories , practitioners cannot themselves vary a construct’s levels but select the desired level. An example are trait theories of job performance. They inform practitioners at what trait level to select employees. Lastly, in observe(X) theories , practitioners can only measure levels of a causal construct. For example, managers can measure employee trust, but they cannot fix this trust at a certain level. I focus on manipulate(X) theories because they are actionable and rigorous. I discuss criteria for constructs in such theories (e.g., construct unity) and three flaws undermining the development of manipulate(X) theories: (1) the simplification fallacy involves the abstraction of complex phenomena like culture into single constructs, (2) the endogenous-cause problem, when endogenous constructs are treated as exogenous, and (3) construct conflation, the lumping of several constructs under one label.