We present and investigate a simple way to generate non-normal data using linear combinations of independent generator (IG) variables. The simulated data have prespecified univariate skewness and kurtosis, and a given ...