TWO-stage least squares methods

The two-stage least squares method is a special regression analysis method using instrumental variables, which helps to ensure that the predictions obtained from the analysis, are accurate and consistent. In this study, two-stage least squares method was examined and an application was performed. In the application, free access data set with 18 variables was used and sample size was 300. First, stepwise regression analysis was applied to the data and the variables to be entered into the model were determined. In the next step, standard regression analysis and two-stage least squares regression analyses were performed. As a result, it is emphasized that two-stage least squares regression method can be used as an alternative to standard regression analysis method in case of endogeneity problem, errors in the X variables and feed-back loop in the model. Keywords: Endogeneity problem unbiased estimation, measurement error