Estimation of the uncertainty bands of regression line for correlated data of variable Y using the GUM rules

The estimation of the accuracy of the linear regression method used for measurements based on the GUM recommendations is considered. The impact of correlation and autocorrelation of the variable Y data together with type A and type B uncertainties, not provided in statistical literature about regression methods, is discussed. The theoretical backgrounds are given. The simulated examples of determining the uncertainty bands of the regression line fitted to the measured points with different cases of correlated values and absolute and relative uncertainties of type A and type B of the dependent variable Y are considered.