Fundamental Sampling Techniques

Course Description: EXST 7012 - Fundamental Sampling Techniques (3 Credit Hours). Simple and stratified random sampling; ratio and regression estimation; cluster, multistage, and multiphase sampling procedures; systematic sampling; unequal probability sampling; non-response and non-sampling errors; links between methodology and application emphasized. This course will not deal directly with the construction of survey instruments and their validation.

Course Prerequisites: Statistical Methods I (EXST 7003, 7004, 7005) or equivalent.
You should be familiar with the use of and the computation of the mean, median, proportion, ratio, sum of squares and cross-products, variance, covariance, correlation, normal or Gaussian distribution Z-scores, confidence interval for a mean or proportion, simple linear regression, and the one-way analysis of variance. You should also be familiar with the use of statistical software (SAS will be used in this course) to make the aforementioned statistical computations.

Course Objectives: General objectives include the following:

• Familiarize the student with sampling terminology and the general objectives of sampling studies
• Familiarize the student with several of the "well known" statistical sampling methods and conditions and situations where each is appropriate
• Provide practice in applying the sampling methods discussed in class and the calculations of desired estimates and their measures of precision
• Provide a foundation for further study in the area of statistical sampling
• Stimulate each student's general interest in and an appreciation for statistics and its application

Instructor: Dr. Barry Moser, 149A Ag. Administration Bldg, 578-8376, bmoser@lsu.edu
-- PLEASE USE BLACKBOARD TO COMMUNICATE WITH ME TO THE EXTENTS POSSIBLE

Office Hours: Fridays 10:00a-11:00a or by appointment.

Text: Lohr, S.L. 1999. Sampling: Design and Analysis. Duxbury Press, Pacific Grove, CA, 494pp.

Course Workbook: Moser, E.B. 2004. Fundamental Sampling Techniques: EXST 7012. Availability to be posted.

Supplementary Text: Scheaffer, R.L., W. Mendenhall, and L. Ott. 1996. Elementary Survey Sampling, fifth edition. PWS-Kent Publishing Co., Boston, 464pp.

Assignments and Grading: Grades will be based upon four (4) regular exams (100 pts each) and a final exam (150 pts) with the total score (550 pts max) divided by 5.5 and rounded to the nearest integer to produce the final score. Missed exams will be scored as 0 points unless authorization to make the exam up is first obtained from Dr. Moser prior to the exam, or due to emergencies recognized by the University (Provost or Dean of the Graduate School) as such. All excused exams MUST BE COMPLETED before the end of the final exam for the course. No additional projects or works will be permitted for additional course credit. Please bring pencils, erasers, your text book, and calculators TO EACH EXAM.

Grading Scale: Letter grade assignments will be made according to the following scale based upon the final course score:

Final Exam: Saturday, July 10, 10:00a-12:00N, 248 Ag Administration

Content Overview

1. Sample surveys overview: Purposes of statistical sampling, basic concepts, sampling and statistical terminology and notation (Lohr: 1, 2.1)
2. Mathematical preliminaries: Expected values, variances, covariances (Lohr: 2.2)
3. Simple random sampling (SRS): Selecting the sample, estimation of the mean, total, and proportion and other population parameters, variances of the estimators, sample size estimation, confidence intervals on the parameters (Lohr 2)
4. SRS: Estimation of parameters for subdomains (Lohr 3.3)
5. Sampling weights (Lohr: 7.2)
6. Systematic sampling: Selecting the sample, parameter estimation, measures of precision for estimates, repeated systematic sampling (Lohr: 2.6, 5.6)
7. Ratio estimation: Using auxiliary information, estimators, estimation of the variance of a ratio estimator (Lohr: 3)
8. Difference and regression estimation: Estimators, comparisons with ratio estimation (Lohr: 3)
9. Double sampling: Uses of multiphase sampling, Applications with ratio estimation (Lohr: 12.1)
10. Relative efficiency and the design effect: Comparing sampling and estimation strategies (Lohr: 7.5)
11. Stratified random sampling (StRS): Selecting the sample, estimation of the parameters, sample size determination and allocation, optimal strategies, post-stratification of a simple random sample (Lohr: 4)
12. Double sampling: double sampling applied to stratified random sampling (Lohr: 12.1)
13. Combining ratio or regression estimation with stratified random sampling: Estimation approaches
14. Single stage cluster sampling: Selecting the sample, parameter estimation and sample sizes, clusters of equal size, clusters of unequal size (ratio estimation and pps sampling), relationship to systematic sampling (Lohr: 5)
15. Two stage cluster sampling: Selecting the sample, parameter estimation and sample sizes (Lohr: 5)
16. Unequal probability sampling: Horvitz-Thompson estimator, Hansen-Hurwitz estimator (Lohr: 6)
17. Adaptive random sampling: Defining a network, selecting the sample, estimation

Topics as time permits

18. Non-response and non-sampling errors: Sample imputation approaches, methods to reduce non-response (Lohr: 8)
19. Sampling models: Randomized response, weighted distributions and size bias, capture-recapture, line-intercept, line-transect sampling (Lohr: 12)
20. Variance estimation: Methods for complex surveys (Lohr: 7, 9)

The course will focus on material covered in chapters 1-6, 7.2, 7.5 and chapter 12 of Lohr 1999 with material added or supplemented in class. This text covers the basics of statistical sampling, but also has material related to model-based sampling, sampling weights, and variance estimation which will be introduced in the class as time permits. To help with your understanding of the material you should work problems listed at the end of each chapter. It is not uncommon for these or similar problems to be included on exams. The material in each chapter will be supplemented in lecture with additional information, estimation methods, and sampling plans, and may be varied depending upon the makeup and interest of the class.  

Last Modified: May 27, 2004