Panel data models were used to estimate the effects of different factors on outcomes. Two new methods were developed to estimate these effects accurately, even with individual differences. These methods were tested and found to work well in practical situations, even with limited data. The researchers also found conditions under which these methods are reliable, helping to correct any biases in the estimates.