The kowledge of Geomatics Engineering.docVIP

Some Empirical Evidence on the Impact of Measurement Errors in Making Ecological Inferences WeiQing Li School of resources , HeBei University Of Engineering,199 guangmingnandajie Road,Handan,056038,China Abstract We seek to identify the impact of data measurement error problems in the context of ecological inference applications. We explore the statistical and substantive implications of using inaccurate proxy variables in the estimation and inference process. The focus of our analysis is on applications of ecological inference in cases involving the Voting Rights Act. We demonstrate our findings with a unique data set on racial registration and turnout in Louisiana and South Carolina. Key words: Ecological Inference, Maximum Entropy, Voting Rights. 1. Introduction In the ecological inference chain there are four important links. One link is related to the statistical model that describes a plausible sampling process. A second link involves the estimation method used to process and recover estimates from the observable sample data. The third link is connected to the observable data that form the basis for the estimation and inference. Finally, the fourth link is the conceptual framework that ties the three links together. 2. Notation and Basic Ecological Inference Problem Matters, the racial registration data is injected with error in the data processing stage. If this measurement error is not taken into account, the assumptions underlying the statistical model may be violated. When these assumptions are violated, our ability to recover the unknown conditional probabilities in the table cells and to trace out the corresponding political implications maybe severely restricted. To develop a model that will recect the characteristics of voter response, consider the observed outcomes for a particular election across i = 1; : : :;m electoral units (e.g., precincts or districts).Each unit has j = 1; : : : ; g types of individual voters and k = 1; : : : ; c candidates for