The source of this difficulty was due to having too small a sample to effectively model the joint distribution, and was this addressed by the best-case-available remedy of adding the (s = 466) reports to the (n = 132) sample to obtain a larger “extended sample” of 598 individuals.