A method for drawing inferences from complex samples is based on Rubin's approach to missing data in survey research. Standard procedures for drawing such inferences do not apply when the variables of interest are not observed directly, but must be inferred from secondary random variables which depend on the variables of interest stochastically. This method allows reasonable inferences to be made. The key is to represent knowledge about latent variables in the form of a predictive distribution, conditional on manifest variables. It is then possible to obtain the expectations of statistics that would have been computed if the values of the latent variables corresponding to sampled units were known, along with variance estimators that account for uncertainty due to both subject sampling and the latency of the variables. (A numerical example is presented, using data from the Profile of American Youth (1980). Possible responses to four arithmetic reasoning items from the Armed Services Vocational Aptitude Battery were studied for Black male and female and White male and female populations). (GDC)