Author's Keywords: Shrinkage Weight	Convention Submission, Review, and Scheduling! All Academic Convention can help with all of your abstract management needs and many more. Contact us today for a quote! Submission - Custom fields, multiple submission types, tracks, audio visual, multiple upload formats, automatic conversion to pdf. Review - Peer Review, Bulk reviewer assignment, bulk emails, ranking, z-score statistics, and multiple worksheets! Reports - Many standard and custom reports generated while you wait. Print programs with participant indexes, event grids, and more! Scheduling - Flexible and convenient grid scheduling within rooms and buildings. Conflict checking and advanced filtering. Communication - Bulk email tools to help your administrators send reminders and responses. Use form letters, a message center, and much more! Management - Search tools, duplicate people management, editing tools, submission transfers, many tools to manage a variety of conference management headaches! Click here for more information.
Association: Name: American Association For Public Opinion Association URL: http://www.aapor.org
Citation: URL: http://www.allacademic.com/meta/p16993_index.html Direct Link: HTML Code:

Griffin, Richard. "Dealing with wide Weight Variation in Polls" Paper presented at the annual meeting of the American Association For Public Opinion Association, Fontainebleau Resort, Miami Beach, FL, <Not Available>. 2009-05-25 <http://www.allacademic.com/meta/p16993_index.html>

Griffin, R. "Dealing with wide Weight Variation in Polls" Paper presented at the annual meeting of the American Association For Public Opinion Association, Fontainebleau Resort, Miami Beach, FL <Not Available>. 2009-05-25 from http://www.allacademic.com/meta/p16993_index.html

Publication Type: Paper/Poster Proposal Abstract: DEALING WITH WIDE WEIGHT VARIATION IN POLLS Richard A. Griffin, U.S. Census Bureau Key Words: Bias, Variance, Model It is well known by survey practitioners that wide weight variation is not good. If a respondent with a relatively large weight is an outlier the resulting estimate may be inappropriately skewed. However, weight variation is not always bad. Unequal selection probabilities result in unbiased estimates that have weight variation. Consider optimal allocation in stratified sampling when the variance of the characteristic of interest is different from one strata to the next. The optimal allocation results in an unbiased minimum variance estimate that has weight variation. Why do we consider weight variation bad? The problem is that most surveys have multiple purposes and many estimates are computed other than those for which the sample design was targeted. Statistically the problem is to choose the estimation procedure that produces the lowest mean square error (the variance of an estimate plus the square of its bias). A biased estimate can have lower mean square error than an unbiased estimate if its variance is enough lower than the variance of the unbiased estimate to compensate for the bias. Polls often have equal probability samples with unequal balancing weights to compensate for differential nonresponse. This paper looks at the effect of unequal weights for equal probability polls on the mean square error of estimates. Balancing weights using known population proportions as well as shrinkage weights under a model are examined.

Where Weight Waxes and Wanes: Spatial Variation in Adult Body Mass Index

How SurveyUSA weights its election poll data

Measuring Spatial Variation in Determinants of Vote Choice Using Geographically Weighted Regression

Degrees of "Scapegotability": City-Wide Variations in the Levels of Violence against the Ethnic Chinese in Indonesia