Measurement errors in linear regression models

Measurement is a fundamental process for experimental researches. The results of studies done in many fields are obtained by using the measurements. Measurements are made to understand the methods of the conducted to research; however, conducted measurements may sometimes contain some errors. Parameter estimates may be biased and inconsistent due to measurement errors in the variables. Some circumstances such as a faulty measurement of variables, incorrect functional formatting, ignoring the necessary variables, unnecessary variables accounted for in the model may cause some deviations. In that case, this deviation or inconsistency, some alternative estimation methods to cope with measurement errors are proposed. In this study, we are mostly interested in measurements of variables, and we examined what we can do. Therefore, techniques and regression model which is minimized the measurement errors examined and then their impact on the model is evaluated. Key Words: Linear regression, Measurement errors, Instrumental variable method, Grouping methods, Methods based on ranks