I have estimated an SEM model in STATA based on survey data. Once I identify the data as survey by including sample weight and bootstrap weight variables I can no longer get fit statistics. Not even the basic X2 and associated p value. The message I receive is: model was fit with svy: prefix; only stats(residuals) valid. Is this a STATA issue (in which case I may be on the wrong forum ) or does using sampling weights preclude software from generating fit statistics? Thank you!!