Least squares regression analysis that ignores possible group structure of the data. For example, in the context of panel data this means stacking the observations, i.e. assuming that the covariance matrix is diagonal and all regression coefficients are the same for all cross-sectional units and all time periods.