BlandAltmananalyyseja

BlandAltmananalyyseja, commonly known in the literature as the Bland-Altman analysis, is a statistical method used to assess agreement between two quantitative measurement techniques. It focuses on the differences between paired measurements rather than their correlation, to determine whether the methods can be used interchangeably. The method was introduced in 1983 by J. Martin Bland and Douglas G. Altman.

Typically, paired measurements x_i and y_i are collected for each subject. For each pair, compute the difference d_i = x_i − y_i and the average m_i = (x_i + y_i)/2. The analysis reports the mean difference (bias) and the standard deviation of the differences. A Bland-Altman plot displays d_i versus m_i, with lines showing the bias and the limits of agreement, defined as bias ± 1.96*SD, assuming the differences are approximately normally distributed. These limits estimate where 95% of future differences are expected to lie.

Interpretation revolves around the magnitude of the bias and the width of the limits of agreement. Small bias and narrow limits suggest good agreement; large bias or wide limits indicate insufficient interchangeability. If the differences vary with the magnitude of measurements (proportional bias), a transformation or regression-based approach may be warranted; assess this by analyzing the relationship between d_i and m_i.

Limitations include reliance on normality of differences and the assumption of independent measurements. Bland-Altman analysis does not assess accuracy relative to a gold standard and is not a substitute for accuracy assessment; it focuses on agreement between methods. It is widely used in medicine, laboratory science, and engineering to evaluate new instruments, devices, or observers.