If your data contains a large number of individuals (cross-sectional units), which is quite common with panel data, then the DV approach can be computationally burdensome (even for a computer) and impractical. A better alternative to this approach is the fixed effects (FE) estimator, which I describe in the next section.