National Assessment Approach to Sampling Error Estimation.

• Superior document: RTI Press Publication
• Author: Folsom, Ralph E.
• Place / Publishing House: Research Triangle Park, NC : : RTI International / RTI Press,, 2014.; Ã2014.
• Year of Publication: 2014
• Edition: 1st ed.
• Language: English
• Series: RTI Press Publication
• Physical Description: 1 online resource (96 pages)

Table of Contents

• Intro
• Figure 4-1. Montreal model (true variance vs. estimated variances)
• Figure 4-2. New model with t = 2 (true variance vs. estimated variance)
• Figure 4-3. New model with t = 1 (true variance vs. estimated variance)
• Foreword
• Preface
• About the Author
• Introduction
• National Assessment Overview
• Population Characteristics and Sample Statistics
• Statistical Inference
• Balanced Effects
• References
• Year 01 Sampling Errors
• Design Description
• Parameters of Interest
• Variance Estimators
• Computational Considerations
• Summary Analysis of Year 01 Sampling Errors
• References
• Controlled Selection: Implications for Year 02 Variances
• Introduction to Controlled Selection
• General Population Structure and the Sample Design
• References
• Simulation Study of Alternative Year 02 Variance Estimators
• Simulation Model
• Variance Approximations
• Empirical Results
• Comparison of Taylor Series and Jackknife Linearizations
• References
• Comparison of Year 01 and Year 02 Sampling Errors
• Comparative Analysis
• References
• Conclusions
• Taylor Series Linearization for Regression Coefficients
• Population Definitions
• Estimation
• Taylor Series Variance Approximation
• Taylor Series Variance Estimator
• Statistical Inference
• Table 2-1. Distribution of national design effects
• Table 2-3. Median design effects for size of community subpopulation P-value estimates
• Table 2-4. Median design effects for regional subpopulation P-value estimates (9- and 13-year-olds only)
• Table 2-5. Median design effects for sex subpopulation P-value estimates
• Table 4-1. Primary sample allocation summary for the West
• Table 4-2. Bias comparisons for four variance estimators (Montreal model)
• Table 4-3. Root mean squared error of variance estimators (Montreal model).
• Table 4-4. Average performance of variance estimators (Montreal model)
• Table 4-5. Sampling distributions for normal and t-like statistics
• Table 4-6. Bias comparisons for three variance estimators under the new model with t = 2
• Table 4-7. Design effects with N = 744 for new model with t = 2
• Table 4-8. Bias comparisons for the new model with t = 1
• Table 4-10. Average performance of variance estimates under the new model with t = 2
• Table 4-11. Average performance of variance estimates under the new model with t = 1
• Table 4-12. Sampling distributions for Gaussian-type statistics
• Table 4-13. Sampling distributions for Student's t-like statistics using V2
• Table 4-14. Bias comparisons for the new model with t = 1 and a Taylor series linearization
• Table 4-16. Sampling distributions for t-like statistics
• Table 5-1. Stem and leaf display of Year 02 national DEFFs
• Table 5-2. Distributions of Year 01 and 02 national DEFFs
• Table 5-3. Comparison of regional DEFFs
• Table 5-4. Year 02 DEFFs by size of community
• Blank Page
• Blank Page.