Model-Dependent and Design-Dependent Sampling Procedures: A Simulation Study (Classic Reprint) - Softcover

Inhaltsangabe

Excerpt from Model-Dependent and Design-Dependent Sampling Procedures<br/><br/>If not enough is known about the relationship between y and x to correct for imbalance in the sample, then a sample should be balanced on average, with respect to x such as in random sampling. A Iachan (1985) noted that the ratio-of-means estimator T can be badly biased. For example, for the model-based expectation (denoted by E2)

Die Inhaltsangabe kann sich auf eine andere Ausgabe dieses Titels beziehen.