##manager.scheduler.building##: Velodromo - Bocconi University
##manager.scheduler.room##: N03
Date: 2019-01-26 11:00 AM – 12:30 PM
Last modified: 2018-12-26
Abstract
Population estimates are an important tool across a variety of sectors, informing planning and investment. As a result of the wide variety of uses that population statistics have, it is important that they are as accurate as possible to ensure their effectiveness. This paper evaluates the Cohort Component model used to produce the official small area population estimates in Scotland, examining the accuracy of this method and exploring the factors which may contribute to error and explain differences in accuracy between small areas. To evaluate this method fully, the accuracy of estimates produced by the Cohort Component method was compared to those produced by alternative methods. This project used alternative estimates, produced using methods employed by other UK statistical agencies; the Ratio Change method used by the Office of National Statistics and the Average method used by the Northern Irish Statistical Research Agency. A multilevel model was used to compare the accuracy of the estimation methods and to explore the differences in the effect of area characteristics on error. Results of this research suggested that the Cohort Component method may not be the most effective method, but adopting one of the alternative methods may provide more robust population statistics.