The article by Rumen Manolov and John M. Ferron discusses the importance of assessing the consistency of effects in single-case experimental designs (SCEDs) to ensure replication and establish the generality of intervention effects. They propose using multilevel models, also known as hierarchical linear models or mixed effects models, to evaluate the consistency of effects within a study. The authors highlight the need for both internal and external validity in SCEDs, emphasizing the importance of replication within and across studies. They propose that assessing the consistency of effects involves checking whether the confidence intervals for random effects include zero, indicating that the fixed effect estimate is plausible for each individual effect. This approach is illustrated with real data and implemented using free user-friendly software. The article also reviews existing methods for quantifying consistency, such as the between-case standardized mean difference (BC-SMD) and the intraclass correlation coefficient (ICC), and discusses the limitations of these methods. The authors suggest that multilevel models provide a more robust and flexible approach to assessing consistency, allowing for the inclusion of multiple data features and the consideration of autocorrelation. They conclude by discussing the application of multilevel models in various SCED designs and the importance of model selection based on theoretical considerations and previous evidence.The article by Rumen Manolov and John M. Ferron discusses the importance of assessing the consistency of effects in single-case experimental designs (SCEDs) to ensure replication and establish the generality of intervention effects. They propose using multilevel models, also known as hierarchical linear models or mixed effects models, to evaluate the consistency of effects within a study. The authors highlight the need for both internal and external validity in SCEDs, emphasizing the importance of replication within and across studies. They propose that assessing the consistency of effects involves checking whether the confidence intervals for random effects include zero, indicating that the fixed effect estimate is plausible for each individual effect. This approach is illustrated with real data and implemented using free user-friendly software. The article also reviews existing methods for quantifying consistency, such as the between-case standardized mean difference (BC-SMD) and the intraclass correlation coefficient (ICC), and discusses the limitations of these methods. The authors suggest that multilevel models provide a more robust and flexible approach to assessing consistency, allowing for the inclusion of multiple data features and the consideration of autocorrelation. They conclude by discussing the application of multilevel models in various SCED designs and the importance of model selection based on theoretical considerations and previous evidence.