The developments in the earlier chapters assumed that the values of the variables involved in an analysis are obtained with precision. More often than not, this is an unrealistic or even patently flawed assumption, but one without which our ability to make inferences would be severely curtailed. This chapter deals with the general problem of imperfect measurement, which includes measurement with error and using a variable that is a substitute for the one we would ideally like to have recorded.
KeywordsLatent Variable Conditional Distribution Sampling Variance Measurement Process Residual Variance
Unable to display preview. Download preview PDF.