This article by Stacie Petter, Detmar W. Straub, and Arun Rai examines the specification of formative constructs in Information Systems (IS) research. The authors argue that while researchers often focus on theoretical links between constructs, the relationship between measurement items and constructs is frequently overlooked. They distinguish between reflective and formative constructs, where reflective constructs are defined by their indicators, and formative constructs define their indicators. The study finds that a significant number of articles in MIS Quarterly and Information Systems Research mis-specify formative constructs, leading to potential Type I and Type II errors in scientific results. To address this issue, the authors provide a roadmap for researchers to properly specify formative constructs, including decision rules for identifying formative constructs and guidelines for addressing them within research models. The article also discusses the implications of mis-specification on the validity of structural models and the need for better understanding and techniques to handle formative constructs.This article by Stacie Petter, Detmar W. Straub, and Arun Rai examines the specification of formative constructs in Information Systems (IS) research. The authors argue that while researchers often focus on theoretical links between constructs, the relationship between measurement items and constructs is frequently overlooked. They distinguish between reflective and formative constructs, where reflective constructs are defined by their indicators, and formative constructs define their indicators. The study finds that a significant number of articles in MIS Quarterly and Information Systems Research mis-specify formative constructs, leading to potential Type I and Type II errors in scientific results. To address this issue, the authors provide a roadmap for researchers to properly specify formative constructs, including decision rules for identifying formative constructs and guidelines for addressing them within research models. The article also discusses the implications of mis-specification on the validity of structural models and the need for better understanding and techniques to handle formative constructs.