Second Stage Of Data Collection Research Articles

Bayesian reference prior analysis for polynomial calibration models

The focus of interest in a statistical calibration problem is the estimation of an unknown regressor valuex0 corresponding to an observed responsey0. A calibration experiment is typically carried out in two stages. At a first stage of data collection, responses are observed corresponding toknown regressor values; from these observations information is obtained about the regression relationship. At a second stage of data collection, one or more responses are observed which correspond to anunknown value of the regressor; estimation of this unknown regressor value is of primary importance. In this paper, we consider the calibration problem for polynomial regression models. In particular, a reference prior and corresponding posterior inferences are derived. The results are illustrated with an example, and comparisons are made between the polynomial reference prior analysis and alternative analyses.

Apparent change in caseness in longitudinal studies.

Unless positive responses are verified at a second stage of data collection, a questionnaire-based survey has limited ability to assess change in caseness in a longitudinal study of a condition of low prevalence, assuming imperfect validity of the survey instrument.