The article explores different ways to expand traditional factor analysis models by incorporating non-normal and asymmetric distributions. It compares placing skewness assumptions on the factors or errors in these models. The researchers found that combining skew distributions on both factors and errors can provide a unified approach to skew factor models.