The article "When to use agreement versus reliability measures" by de Vet et al. (2006) discusses the differences and relationships between reliability and agreement parameters in the context of medical research. The authors emphasize that reproducibility concerns the consistency of repeated measurements, while agreement parameters assess how close these measurements are, and reliability parameters assess the ability to distinguish between study subjects despite measurement errors. The article uses an example of an interrater study where different physical therapists measure the range of motion of the arm in patients with shoulder complaints. It illustrates how reliability and agreement parameters differ and how they are related. Reliability parameters, such as the intraclass correlation coefficient (ICC), are influenced by the variability between subjects, while agreement parameters, such as the standard error of measurement (SEM), focus on the measurement error itself. The authors conclude that reliability parameters are appropriate when the goal is to distinguish between subjects, while agreement parameters are preferred when the aim is to measure changes in health status, which is common in clinical practice. They also highlight that agreement parameters are more stable across different population samples and are more suitable for evaluative purposes in medical research. The article provides recommendations for the correct use and interpretation of these parameters.The article "When to use agreement versus reliability measures" by de Vet et al. (2006) discusses the differences and relationships between reliability and agreement parameters in the context of medical research. The authors emphasize that reproducibility concerns the consistency of repeated measurements, while agreement parameters assess how close these measurements are, and reliability parameters assess the ability to distinguish between study subjects despite measurement errors. The article uses an example of an interrater study where different physical therapists measure the range of motion of the arm in patients with shoulder complaints. It illustrates how reliability and agreement parameters differ and how they are related. Reliability parameters, such as the intraclass correlation coefficient (ICC), are influenced by the variability between subjects, while agreement parameters, such as the standard error of measurement (SEM), focus on the measurement error itself. The authors conclude that reliability parameters are appropriate when the goal is to distinguish between subjects, while agreement parameters are preferred when the aim is to measure changes in health status, which is common in clinical practice. They also highlight that agreement parameters are more stable across different population samples and are more suitable for evaluative purposes in medical research. The article provides recommendations for the correct use and interpretation of these parameters.