The standard panel data literature is moving from micro panels, where the cross-section dimension is large and the intertemporal sample size is small, to large panels, where both, the cross-section and the time dimension, are large. This thesis contributes to this new and growing area of panel data treatments called "large panel data analysis''. My dissertation consists of three essays: In the first essay, a large panel data model with an omitted factor structure is considered. The role of the factors is to control for the issue of the unobserved time-varying heterogeneity effects. A parameter cascading strategy is proposed to enable efficient estimation of all model parameters when the number of factors is unknown a priori. In the second essay, further models that combine large panel data models with different versions of unobserved latent factors are discussed. Computation-related issues are solved and new specification tests are introduced to check whether or not these factors can be interpreted as classical additive fixed effects. In the third essay, a novel method for estimating panel models with multiple structural changes is proposed. The breaks are allowed to occur at unknown points in time and may affect the multivariate slope parameters individually. Asymptotic results are derived, Monte Carlo experiments are performed, and applications for highlighting these new methods are discussed....