Extending the applicability of stochastic dominance decision rules