Panel data models were used to estimate the impact of actual work experience on wages for women. Different methods were compared to correct for selection bias in the data. The assumption of exogeneity of variables in the model was crucial, but often violated in real-world applications. The study also addressed the issue of measurement error in survey data, especially when dealing with life history variables.