WSS NEWS

News about activities of the Washington Statistical Society
A Chapter of the American Statistical Association

---

June 2000

-->

---

Contents:

• Open Board Meeting
• WSS Seminars Videoconferenced to the Census Bureau
• Special Course Offering at GMU for Fall, 2000
• CPS Technical Paper 63
• Employment Opportunities
• Short Courses
• Seminars
• Area Meetings and Courses (Adobe Acrobat PDF file)
• Note From The WSS NEWS Editor
• WSS People

Open Board Meeting

In order to attend the meeting, you will need to contact Tom Mule by email(vincent.t.mule.jr@ccmail.census.gov) or by phone (301 457 8322) before June 23rd so your name can be added to the BLS visitor list.

You will need to check in at the guard's desk. The building is across First Street from the Union Station metro stop.

Return to top

WSS Seminars Videoconferenced to the Census Bureau

June 5 – Response Variance in the Current Population Survey Income Supplement

Reichert/Kindelberger
Room G-316 Building 3
12:30 - 2:00

June 13 – A (Latin) Square Deal for Voters: Fixing a Flaw in the Australian Preferential Voting System

K.R.W. Brewer
Room 3225 Building 4
12:30 - 2:00

June 26 – The Influence of Environmental Characteristics on Survey Cooperation

Harris-Kojetin/Fricker
Room G-316 Building 3
12:30 - 2:00

June 28 – New Sources of Demographic Data on the Web

Mather
Room 3225 Building 4
12:30 - 2:00

July 11 – Construction of efficient one-level rotation sampling designs

Room 3225 Building 4
12:30 - 2:00

Return to top

Special Course Offering at GMU for Fall, 2000

STAT 679: Topics in Survey Design and Analysis
Fall Semester, 2000

Dr. William Winkler U.S. Bureau of the Census
Dr. Phillip Kott, National Agricultural Statistics Service

Course Outline

Part I: Record Linkage (5 weeks)
Dr. Winkler

Lecture 1
1.  Examples of Record Linkage
2.  Name and Address Standardization
3.  Review
3.1.  Multinomial Distribution
3.2.  Maximum Likelihood
3.3.  Likelihood Ratio Test
4.  Fellegi-Sunter Model
5.  Elementary Contingency Tables
5.1.  Loglinear Models
5.2.  Latent Classes
6.  Parameter Estimation
6.1.  Elementary
6.2.  EM Algorithm

Reading assignment
Winkler (1995) Chapter from Business Survey Methods
Winkler (1988) EM Algorithm
Winkler (1994) Advanced Methods for Record Linkage

Homework assignment
Install and run address standardization software
From example input, produce output file

Lecture 2
1.  Review of EM algorithm
2.  Frequency-based matching
3.  String comparator metrics
4.  Assignment algorithm
5.  Some heuristics
6.  General parameter estimation and matching
7.  Summarizing comments on matching methods
8.  Review of random sampling

Homework assignment
Install and run matching software
Run example to produce outputs 

Lecture 3
1.  Blocking
1.1.  Elementary
1.2.  Advanced
1.3.  USDA example
1.4.  Administrative list example
2.  Review of Contingency Tables and Loglinear
3.  Error-rate estimation
3.1.  Sampling based
3.2.  Belin-Rubin
3.3.  Winkler heuristic
3.4.  Capture-Recapture
4.  Summarizing Questions 

Homework assignment
1.  Install and run Belin-Rubin error-rate software
    Run example.  Answer several questions.
2.  Answer one of set of elementary questions.
      
Lecture 4
1.  Questions to answer prior to matching projects 
2.  Elementary Examples
3.  Advanced Examples
4.  Review of software

Homework assignment
1.  Answer two of a series of elementary questions
2.  Do one of three problems related to running software

Lecture 5
1.  Administrative lists
2.  Analyzing merged files
2.1.  Contingency tables
2.2.  Quantitative data
3.  Microdata Confidentiality
3.1.  Masking methods
3.2.  Example of Fuller
3.3.  Example of Kim-Winkler

Part II: Randomization-assisted Model-based Sampling (9 weeks)
Dr. Kott

Lecture 6  - 
Review of Randomization-Based Survey Sampling
(Chapters 2-4 in SSW*)
Additional notes on with- versus without-replacement sampling, 
the Hartley-Rao variance estimator. 

Lecture 7
Introduction to Using Models in Survey Sampling 
Estimating means without auxiliaries.  
Random-effects models for two-stage sampling. 

Lecture 8  
Asymptotics in Randomization-Based Theory
(Chapter 5.1- 5.4 in SSW)  
Application of theory to ratios, products, and log odds.  

Lecture 9
Extension to Vectors/ Distribution Functions
(Chapter 5.5 -5.11 in SSW).
Discussion of score-based confidence intervals.

Lecture 10
The Randomization Consistent Regression Estimator (RCRE) 
(Chapter 6 in SSW)
Discussion of the Godambe-Joshi lower bound. 
Superiority of the weighted residual variance estimator. 

Lecture 11
The RCRE in Practice and Calibration
(Chapter 7 in SSW).
Discussion of calibration estimators.

Lecture 12
Two-Phase Sampling and Estimation/Unit Non-response  
The rhetorical framework for two-phase sampling 
(Chapter 9 in SSW)
The double expansion versus the reweighted expansion estimator.
Unit nonresponse as an additional phase of sampling.

Lecture 13
Replication/Item Non-response
Randomization and model-based properties of the jackknife and BRR.
Imputation for item nonresponse and replication to capture
imputation's contribution to variance. 
Discussion of multiple imputation.
  
Lecture 14
Small Domains/Analysis of Survey Data
Extending the RCRE to small-domain estimation
Theoretical interpretations of the analysis of survey data. 
Estimating degrees of freedom. 

* SSW: Model Assisted Survey Sampling by Särndal, Swensson, and Wretman. Optional text will provide source of homework assignments

Return to top

CPS Technical Paper 63

This technical paper describes the many changes made to the survey since the publication of the previous technical documentation (Technical Paper 40) over two decades ago. While the basic approach to collecting labor force and other data through the CPS has remained intact, many changes have been made to the survey, including the improvement of numerous questions and the computerization of the survey instrument, both introduced January 1994.

The document describes the CPS design and methodology as of December 1995. Appendices cover updates that have been made to the survey since then. The document is available via the Internet at the CPS website at http://www.bls.census.gov/cps. Future changes to the survey will be documented at this site.

Return to top

Note From The WSS NEWS Editor

Return to top

WSS People

Return to top

First posted May 31, 2000
Last modified June 2, 2000

Webmaster: Dan Jacobs, Maryland Sea Grant