STT 221 (W) : Sampling Techniques

Instructor: Prof. Poduri S.R.S. Rao, Program in Statistics, The college

Text: Sampling Methodologies with Applications, Poduri S.R.S. Rao,

Chapman & Hall/ CRC Press, 2000

Prerequisites: One or two courses such as STT 165 or 212. Familiarity with basic statistical concepts such as expectation, variance, standard error, confidence limits and tests of significance ( covered, for instance, in the beginning of STT 203) will be helpful. Instruction will be provided during the course to make up for the student deficiencies in these topics.

This course provides instruction in the principles and applications of sample surveys, their analysis and statistical treatment. Illustrations from the public and private surveys, market research studies and public opinion polls will be presented. There will be weekly assignments, a project to be completed by each student, one or two tests and the final exam. The tests and final exam are Open Book. Computer programs and software can be used for the assignments and projects.

Grading: Homework assignments 20%, Project 10%, Tests 25%, and Final 45%.

Syllabus

1. Illustrations of the sample surveys conducted by marketing organizations, U.S. Bureau of the Census, Depts. of Agriculture, Energy and labor, the National Center for Health Statistics (NCHS) of the U.S. Center for Disease Control, and other public and private organizations.

2. Statistical concepts: Finite and infinite populations, expectation, variance, standard deviation , covariance, correlation; estimation of population parameters and quantities such as the mean, total and standard deviation; standard error, bias, Mean square Error, precision and accuracy of an estimator.

3. Simple random sampling: The procedure and its properties. Estimation of the mean and total of a finite population. Estimation of the ratio of two means or totals, difference of means or ratios, standard errors and confidence limits.

4. Simple random sampling: Estimation of proportions and percentages, differences of percentages, standard errors, and confidence limits.

5. Sample size determination to estimate means, totals, ratios and percentages for specified precisons, available budgets or prescribed confidence widths.

6. Systematic sampling and its comparison with simple random sampling.

7. Stratification of the population and its advantages. Allocation of the sample to minimize cost and maximize precision.

8. Subpopulations and domains of interest such as male and female subgroups, age and income levels, rural and urban areas; estimation of the means, proportions and their differences; standard errors and confidence limits.

9. Regression and ratio type estimators utilizing supplementary and concomitant information, and their advantages.

10. Clusters such as market-place malls, dwelling areas, counties and districts. Equal and unequal probability sampling procedures for clusters of unequal sizes.

11. Two-stage sampling of the primary and second stage units followed by estimation of the population means and totals; equal and unequal probability sampling.

12. Nonresponse in sample surveys and remedial measures.

13. Further topics: Longitudinal and rotation samples; Taylor's series, jackknife and bootstrap methods for bias reduction and estimation of standard errors and Mean Square Errors.