This article critically examines Baron and Kenny's (1986) procedure for determining if an independent variable affects a dependent variable through a mediator. The authors argue that the widespread use of this procedure has led to many research projects being terminated early or rejected by reviewers due to data not conforming to its criteria. They present a narrative review of the literature, highlighting points that have been disputed in the technical literature but not widely disseminated to practicing researchers. The article introduces a decision tree for testing mediation, classifying its type, and interpreting findings for theory building and future research. Key points include: 1. **Misapplication of Baron and Kenny's Procedure**: Many researchers abandon projects or face rejection due to data not meeting their criteria. 2. **The Importance of Partial Mediation**: Partial mediation, where both indirect and direct effects exist, is more common than full mediation and can provide valuable insights. 3. **Significance of the Indirect Effect**: The indirect effect $a \times b$ is the primary requirement for mediation, not the zero-order effect of $X$ on $Y$. 4. **Bootstrap Test Over Sobel Test**: The bootstrap test is more powerful and should be used to test the significance of the indirect effect. 5. **Interpreting Unexpected Signs**: The sign of the direct effect can indicate the presence of omitted mediators, even if the indirect effect is significant. The article aims to correct misconceptions and provide a more comprehensive framework for mediation analysis, emphasizing the importance of considering both the indirect and direct effects.This article critically examines Baron and Kenny's (1986) procedure for determining if an independent variable affects a dependent variable through a mediator. The authors argue that the widespread use of this procedure has led to many research projects being terminated early or rejected by reviewers due to data not conforming to its criteria. They present a narrative review of the literature, highlighting points that have been disputed in the technical literature but not widely disseminated to practicing researchers. The article introduces a decision tree for testing mediation, classifying its type, and interpreting findings for theory building and future research. Key points include: 1. **Misapplication of Baron and Kenny's Procedure**: Many researchers abandon projects or face rejection due to data not meeting their criteria. 2. **The Importance of Partial Mediation**: Partial mediation, where both indirect and direct effects exist, is more common than full mediation and can provide valuable insights. 3. **Significance of the Indirect Effect**: The indirect effect $a \times b$ is the primary requirement for mediation, not the zero-order effect of $X$ on $Y$. 4. **Bootstrap Test Over Sobel Test**: The bootstrap test is more powerful and should be used to test the significance of the indirect effect. 5. **Interpreting Unexpected Signs**: The sign of the direct effect can indicate the presence of omitted mediators, even if the indirect effect is significant. The article aims to correct misconceptions and provide a more comprehensive framework for mediation analysis, emphasizing the importance of considering both the indirect and direct effects.