Table of contents for National Survey of Family Growth, cycle 6 : sample design, weighting, imputation and variance estimation.

Bibliographic record and links to related information available from the Library of Congress catalog.

Note: Contents data are machine generated based on pre-publication provided by the publisher. Contents may have variations from the printed book or be incomplete or contain other coding.


Counter
Contents									
Abstract
Executive Summary
	Sample Design
	Responsive Design Features
	Sampling Weights
	Item Imputation
	Variance Estimation
	Conclusion
Background
Design Specifications
Sample Design
Summary of Sample Design
	Sample Size Requirements
	First-Stage Selection of Primary Sampling Units
	Second-Stage Selection of Segments
	Third-Stage Selection of Housing Units
	Fourth-Stage Selection of Sample Persons
	Sample Release
Sampling in Three Phases	
Sample Characteristics
	Response Rate Estimation
	Nonresponse Bias and Phase 3 Sampling
	Cost Efficiency and Response Rates in the Phase 3 Sample
	Responsive Design
	Sample Size
Summary of Weighting Procedures
	Inverse Probability Selection Weighting
	Nonresponse Adjustment
	Poststratification
	Weight Trimming
	Final Weight Summary	
	Estimating Equation for Totals
Imputation Procedures
	Background on Imputation for Missing Values 
Sequential Regression Imputation
Multiple Imputation Using Sequential Regression 
	Regression Imputation in NSFG Cycle 6
Variance Estimation
 Background
	Summary of Variance Estimation Principles
	Variance Estimation Software
	Application to the National Survey of Family Growth
References
Appendixes
I.	Definition of Terms
II.	Variables Imputed and Types of Imputation for Female, Male, and Pregnancy Files
III.	Distribution of Female, Male, and Pregnancy Records, by Stratum and Sampling Error Computing Units in the National Survey of Family Growth, Cycle 6
IV.	Variance Estimates Using SAS, SUDAAN, and STATA, for the Percentage of All Women Using the Oral Contraceptive Pill, by 5-Year Age Groups
V.	Variance Estimates Using SAS, SUDAAN, and STATA, for the Mean Number of Children Ever Born, on Race and Hispanic Origin, and Age for Women 20¿44 Years of Age
VI.	Variance Estimates Using SAS, SUDAAN, and STATA, for the Regression of Parity (Children Ever Born), by Race and Hispanic Origin, Age, and Education for Women 20¿44 Years of Age
VII.	Variance Estimates Using SAS, SUDAAN, and STATA, for the Probability of Strongly Agreeing with the Statement, ¿A Young Couple Should Not Live Together Unless They Are Married¿ Regressed on Age, Gender, Race and Hispanic Origin, and Education for Women and Men 15¿44 Years of Age
VIII. 	Median Design Effect for Female, Male, and Pregnancy File Means and Proportions
Text Figures
1.	History of the National Survey of Family Growth
2.	National Survey of Family Growth Cycle 6 selection of primary sampling units and segments
3.	Selection steps in sample implementation for Cycle 6 of the National Survey of Family Growth
4.	Within household screener and household roster
5.	Within household measures of size and illustration of within household selection
6.	Response rates for Phases 1 and 2 and overall for 18 subpopulations
7.	Response rates for Phases 1 and 2 and overall (reflecting subsampling of nonrespondents) by age group; National Survey of Family Growth Cycle 6, 2002
8.	Estimated proportion of females who have ever had sexual intercourse in Cycle 6 of the National Survey of Family Growth: Cumulative estimate; estimate by call number of interview, and estimate from Phase 3 interviewing
9.	Cumulative ratio of interviewer hours to number of main interviews completed by day of the data collection period: National Survey of Family Growth Cycle 6, 2002
10.	SAS 9.1 output from illustration 1 for estimation of the percent of women using the oral contraceptive pill in each of six 5-year age groups, National Survey of Family Growth Cycle 6, 2002
11.	SUDAAN 8.0.2 output from illustration 1 for estimation of the percent of women using the oral contraceptive pill in each of six 5-year age groups, National Survey of Family Growth Cycle 6, 2002
12.	Stata output from illustration 1 for estimation of the percent of women using the oral contraceptive pill in each of six 5-year age groups, National Survey of Family Growth Cycle 6, 2002
13.	Wesvar table specification window for illustration 1 for estimation of the percent of women using the oral contraceptive pill in each of six 5-year age groups, National Survey of Family Growth Cycle 6, 2002
Text Tables
A. Allocation of sample segments and sampling rates within domains: National Survey of Family Growth, Cycle 6
B. Expected and observed eligibility rates by sex, age, and race or ethnicity: National Survey of Family Growth, Cycle 6
C. Number of Phase 1 and 2 segments by number of active incomplete cases in the segment at the end of Phase 2, and the estimated total propensity of active cases to complete an interview on the next visit: National Survey of Family Growth, Cycle 6
D. Phase 3 segment sampling fractions, by number of active cases in the segment, and the estimated total propensity of active cases in the segment to complete the interview on the next visit: National Survey of Family Growth, Cycle 6
E. Response rates, weighted only for subsampling in Phase 3, but not for other probabilities of selection, using two estimates (0.64 and 0.58) of the proportion of nonscreened households that had an eligible respondent: National Survey of Family Growth, Cycle 6
F. Fully weighted response rates, adjusted for all unequal probabilities of selection, using two estimates (0.64 and 0.58) of the proportion of nonscreened households that had an eligible respondent: National Survey of Family Growth, Cycle 6
G. Mean number of calls per active screener case by number of active cases and total propensity strata in Phase 3: National Survey of Family Growth, Cycle 6
H. Mean number of calls per active main interview case by number of active cases and estimated total propensity of cases in the segment to respond on the next visit: Phase 3 of National Survey of Family Growth, Cycle 6
J. Mean screener response rates for segments in Phase 3 by number of active cases and total propensity strata: National Survey of Family Growth, Cycle 6
K. Mean main interview response rates for segments during Phase 3 by number of active cases at the beginning of Phase 3, and the total propensity of cases in the segment to respond on the next visit: National Survey of Family Growth, Cycle 6 
L. Mean ratio of number of screener interviews to number of total calls on screener cases by number of active cases in the segment, and by the estimated total propensity of cases in the segment to respond on the next interviewer visit: National Survey of Family Growth, Cycle 6
M. Mean ratios of number of main interviews to number of total calls on screener cases by number of active cases and total propensity strata: National Survey of Family Growth, Cycle 6
N. Number of completed interviews by age group, sex, and race or ethnicity: National Survey of Family Growth, Cycle 6
O. Mean, minimum, and maximum untrimmed base weight, and potential increase in variance due to weighting (1+L), by age group, sex, and race or ethnicity: National Survey of Family Growth, Cycle 6
P. Logistic Regression Model Predicting Response to the Screener, with 18 predictors (Phases 1¿3): National Survey of Family Growth, Cycle 6
Q. Nine logistic regression models predicting response to the main interview (Phases 1¿3): National Survey of Family Growth, Cycle 6
R. Mean, minimum, and maximum combined nonresponse adjustments, and potential increase in variance due to weighting, by age group, sex, and race or ethnicity: National Survey of Family Growth, Cycle 6
S. Poststratification adjustment factors for 36 groups, by age, sex, and race or ethnicity: National Survey of Family Growth, Cycle 6
T. Mean, minimum, and maximum final (trimmed) weight, and potential increase in variance due to weighting, by age group, sex, and race or ethnicity: National Survey of Family Growth, Cycle 6 

Library of Congress Subject Headings for this publication:

Family size -- Statistical methods.
Fertility, Human -- Statistical methods.
Family life surveys.
Health surveys.
National Survey of Family Growth (U.S.).
Research Design -- United States.
Data Collection -- methods -- United States.
Family Characteristics -- United States.
Parenting -- United States.
Reproductive Behavior -- United States.