This section discusses measurement error in medical statistics, focusing on the variability of repeated measurements on the same subject. The authors use an example of lung function measurements in 20 schoolchildren to illustrate the concept. They introduce the idea of a "true" average value and explain that repeated measurements will vary around this true value due to measurement error. The standard deviation of these repeated measurements, known as the within-subject standard deviation ($\zeta_w$), is used to quantify the size of the measurement error. The calculation of $\zeta_w$ involves averaging the variances of multiple subjects, and the authors provide a method using one-way analysis of variance (ANOVA) for practical implementation. They also discuss the assumption that the standard deviation is unrelated to the magnitude of the measurement, which can be checked graphically and analytically. For designs with only two measurements per subject, a simplified formula for $\zeta_w$ is provided. The measurement error can be expressed as $\zeta_w$, and the repeatability, defined as $2.77 \zeta_w$, is another useful measure. The authors conclude by noting that individual peak expiratory flow readings are rarely used due to their variability, and the mean of the last three readings is typically used for analysis.This section discusses measurement error in medical statistics, focusing on the variability of repeated measurements on the same subject. The authors use an example of lung function measurements in 20 schoolchildren to illustrate the concept. They introduce the idea of a "true" average value and explain that repeated measurements will vary around this true value due to measurement error. The standard deviation of these repeated measurements, known as the within-subject standard deviation ($\zeta_w$), is used to quantify the size of the measurement error. The calculation of $\zeta_w$ involves averaging the variances of multiple subjects, and the authors provide a method using one-way analysis of variance (ANOVA) for practical implementation. They also discuss the assumption that the standard deviation is unrelated to the magnitude of the measurement, which can be checked graphically and analytically. For designs with only two measurements per subject, a simplified formula for $\zeta_w$ is provided. The measurement error can be expressed as $\zeta_w$, and the repeatability, defined as $2.77 \zeta_w$, is another useful measure. The authors conclude by noting that individual peak expiratory flow readings are rarely used due to their variability, and the mean of the last three readings is typically used for analysis.