A Scaling Model for Survey Monitoring

Abstract
Social indicators work has been dominated by "objective" measures expressed in various physical units (for example, crime rates). The present work is part of an attempt to correct this imbalance by providing a method for monitoring subjective indicators over different demographic segments and time points. The present monitoring model, which is based on psychometric scaling theory, is also similar to models in test theory and choice theory, and its parameters are estimated noniteratively. A principal advantage of this model, when applied to survey rating scales, stems from built-in parameters which provide for its evaluation. This evaluation reveals encouraging results in the present applications of the model to cross-classified survey data. Other potential applications, including quasi- experimental methods, are also discussed.

This publication has 8 references indexed in Scilit: