Can we improve the survey representation of non-resident parents, and collect robust data on reasons for separation?