Ranked Set Sampling: Improving Estimates from a Stratified Simple Random Sample Christopher Sroka, Elizabeth Stasny, and Douglas Wolfe Department of Statistics.

Slides:



Advertisements
Similar presentations
Introduction Simple Random Sampling Stratified Random Sampling
Advertisements

7 (a) Under what circumstances is stratified random sampling procedure is considered appropriate?How would you select such samples?Explain by means of.
Sampling: Final and Initial Sample Size Determination
Statistics for Managers Using Microsoft® Excel 5th Edition
1 STRATIFIED SAMPLING Stratification: The elements in the population are divided into layers/groups/ strata based on their values on one/several.
Chapter 5 Stratified Random Sampling n Advantages of stratified random sampling n How to select stratified random sample n Estimating population mean and.
QBM117 Business Statistics Statistical Inference Sampling 1.
Complex Surveys Sunday, April 16, 2017.
Categories of Sampling Techniques n Statistical ( Probability ) Sampling: –Simple Random Sampling –Stratified Random Sampling –Cluster Random Sampling.
Introduction to Statistics
Chapter 10 Sampling and Sampling Distributions
Chapter 17 Additional Topics in Sampling
Business Statistics: A Decision-Making Approach, 7e © 2008 Prentice-Hall, Inc. Chap 1-1 Business Statistics: A Decision-Making Approach 7 th Edition Chapter.
Why sample? Diversity in populations Practicality and cost.
Aaker, Kumar, Day Ninth Edition Instructor’s Presentation Slides
Chapter 1 The Where, Why, and How of Data Collection
Chapter 1: Data Collection
STAT262: Lecture 5 (Ratio estimation)
The Excel NORMDIST Function Computes the cumulative probability to the value X Business Statistics: A First Course, 5e © 2009 Prentice-Hall, Inc
Ratio estimation with stratified samples Consider the agriculture stratified sample. In addition to the data of 1992, we also have data of Suppose.
A new sampling method: stratified sampling
Stratified Simple Random Sampling (Chapter 5, Textbook, Barnett, V
7-1 Copyright ©2011 Pearson Education, Inc. publishing as Prentice Hall Chapter 7 Sampling and Sampling Distributions Statistics for Managers using Microsoft.
STAT 4060 Design and Analysis of Surveys Exam: 60% Mid Test: 20% Mini Project: 10% Continuous assessment: 10%
Stratified Sampling Lecturer: Chad Jensen. Sampling Methods SRS (simple random sample) SRS (simple random sample) Systematic Systematic Convenience Convenience.
5.10: Stratification after Selection of Sample – Post Stratification n Situations can arise in which we cannot place sampling units into their correct.
Sampling Designs Avery and Burkhart, Chapter 3 Source: J. Hollenbeck.
Increasing Survey Statistics Precision Using Split Questionnaire Design: An Application of Small Area Estimation 1.
Sample Design.
Chapter 3 Goals After completing this chapter, you should be able to: Describe key data collection methods Know key definitions:  Population vs. Sample.
Definitions Observation unit Target population Sample Sampled population Sampling unit Sampling frame.
BPS - 5th Ed. Chapter 81 Producing Data: Sampling.
Chap 20-1 Statistics for Business and Economics, 6e © 2007 Pearson Education, Inc. Chapter 20 Sampling: Additional Topics in Sampling Statistics for Business.
1 Sampling Distributions Lecture 9. 2 Background  We want to learn about the feature of a population (parameter)  In many situations, it is impossible.
Copyright ©2011 Pearson Education 7-1 Chapter 7 Sampling and Sampling Distributions Statistics for Managers using Microsoft Excel 6 th Global Edition.
GREAT Day!!!. Producing Data Population – Entire group of individuals or objects that we want information about. Defined in terms of what we want to know.
Chapter 18 Additional Topics in Sampling ©. Steps in Sampling Study Step 1: Information Required? Step 2: Relevant Population? Step 3: Sample Selection?
Chapter 11 – 1 Chapter 7: Sampling and Sampling Distributions Aims of Sampling Basic Principles of Probability Types of Random Samples Sampling Distributions.
4.4 Statistics Notes What are Other Ways to Conduct Experimental and Observational Studies?
Performance of Resampling Variance Estimation Techniques with Imputed Survey data.
Statistics for Managers Using Microsoft Excel, 4e © 2004 Prentice-Hall, Inc. Chap 1-1 Statistics for Managers Using Microsoft ® Excel 4 th Edition Chapter.
Sampling Design and Analysis MTH 494 LECTURE-12 Ossam Chohan Assistant Professor CIIT Abbottabad.
Lohr 2.2 a) Unit 1 is included in samples 1 and 3.  1 is therefore 1/8 + 1/8 = 1/4 Unit 2 is included in samples 2 and 4.  2 is therefore 1/4 + 3/8 =
BUS216 Spring  Simple Random Sample  Systematic Random Sampling  Stratified Random Sampling  Cluster Sampling.
Section 5.1 Designing Samples AP Statistics
Copyright © 2011 by The McGraw-Hill Companies, Inc. All rights reserved. McGraw-Hill/Irwin Chapter 7 Sampling and Sampling Distributions.
Basic Business Statistics, 10e © 2006 Prentice-Hall, Inc.. Chap 7-1 Chapter 7 Sampling Distributions Basic Business Statistics.
Basic Business Statistics, 11e © 2009 Prentice-Hall, Inc. Chap 7-1 Chapter 7 Sampling and Sampling Distributions Basic Business Statistics 11 th Edition.
Chap 1-1 Chapter 3 Goals After completing this chapter, you should be able to: Describe key data collection methods Know key definitions:  Population.
Basic Business Statistics
Lecture 4 Forestry 3218 Avery and Burkhart, Chapter 3 Shiver and Borders, Chapter 5 Forest Mensuration II Lecture 4 Stratified Random Sampling.
Sampling Designs Outline
1 Chapter 11 Understanding Randomness. 2 Why Random? What is it about chance outcomes being random that makes random selection seem fair? Two things:
1. 2 DRAWING SIMPLE RANDOM SAMPLING 1.Use random # table 2.Assign each element a # 3.Use random # table to select elements in a sample.
ESTIMATING RATIOS OF MEANS IN SURVEY SAMPLING Olivia Smith March 3, 2016.
Survey sampling Outline (1 hr) Survey sampling (sources of variation) Sampling design features Replication Randomization Control of variation Some designs.
Introduction/ Section 5.1 Designing Samples.  We know how to describe data in various ways ◦ Visually, Numerically, etc  Now, we’ll focus on producing.
Statistics for Managers Using Microsoft Excel, 4e © 2004 Prentice-Hall, Inc. Chap 1-1 Statistics for Managers Using Microsoft ® Excel 4 th Edition Chapter.
CHAPTER 4 Designing Studies
John Loucks St. Edward’s University . SLIDES . BY.
Two-Phase Sampling (Double Sampling)
Chapter 7 Sampling Distributions
Estimation of Sampling Errors, CV, Confidence Intervals
Random sampling Carlo Azzarri IFPRI Datathon APSU, Dhaka
Chapter 8: Weighting adjustment
Data Sampling Jerry Post Copyright © 1997
Sampling and estimation
1st Joint Workshop Pesticides Statistics
EQ: What is a “random sample”?
Presentation transcript:

Ranked Set Sampling: Improving Estimates from a Stratified Simple Random Sample Christopher Sroka, Elizabeth Stasny, and Douglas Wolfe Department of Statistics The Ohio State University

Alternative Title – Ranked Set Sampling: Where are the Samplers? Purpose: Show that RSS can be incorporated into traditional sampling designs Compare RSS to traditional sampling designs Develop stratified ranked set sampling (SRSS) Computer simulation to evaluate relative standard error

Notation Select m random samples of size m with replacement from the population Order the m items within each set using auxiliary variable or visual judgment We do this before measuring our variable of interest

Notation Select one ranked unit from each set and quantify with respect to variable of interest X [m]1 Set X [3]1 X [2]1 X [1]1 X [m]2 Set X [3]2 X [2]2 X [1]2 X [m]3 Set X [3]3 X [2]3 X [1]3... X[m]mX[m]m Set m X [3]m X [2]m X [1]m

Notation X [1]1 X [2]1 X [3]1... X [m]1 X [1]2 X [2]2 X [3]2... X [m] X [1]k X [2]k X [3]k... X [m]k Repeat k times to get a total of mk measurements on our variable of interest

Notation Our estimator of the population mean for the variable of interest is the average of our mk quantified observations:

RSS vs. Stratified Sampling For fixed sample size n = mk, Sample DesignPopulation info needed No. of auxiliary measurements RSSNonemn = m 2 k PoststratificationTotalsn = mk SSRSEnough for stratification Entire population

RSS vs. Stratified Sampling Expect SSRS to be better than RSS, since uses more population info Can we improve on SSRS using RSS? Stratified ranked set sampling (SRSS): Use RSS to select units from each stratum We estimate the population mean by RSS estimator from before Stratum weights

Simulation USDA data on corn production in Ohio Treat the data set as a population Use computer simulation to estimate the precision of each technique –Sample from data using each method –Estimate mean accordingly –Repeat 50,000 times Use the variance of the 50,000 mean estimates to approximate the standard error of the estimator

Simulation Performed simulation multiple times, varying –Sample size –Number of strata –Number of sets –Combination of ranking variable and variable of interest (correlations vary) Reported standard error as percent of standard error under simple random sampling

Simulation Number of sets in RSS equals number of strata in SSRS and SRSS Only one cycle within strata for SRSS For example, for 3 strata and sample size of 30 RSS: 3 sets of 3, repeat for 10 cycles SSRS: 3 strata, 10 observations per stratum SRSS: 3 strata, 10 sets of 10, 1 obs. per set

Results SRSS is more precise than SSRS for almost all combinations of variables, set sizes, and sample sizes Increased precision of SRSS the highest when –Strong correlation between ranking variable and variable of interest (i.e., accurate rankings) –Large sample size SRSS less precise or not much more precise than SSRS when –Low correlation –Large number of strata combined with low sample size

Results – High Correlation (0.996) Sample size n = 15n = 30n = 60n = 120 #strata = % 61.0% 79.2% 52.2% 79.2% 44.1% 79.1% 36.1% #strata = % 59.5% 70.0% 50.4% 69.9% 43.2% 69.8% 36.2% #strata = % 51.5% 45.2% 51.0% 40.5% 51.4% 34.9% WHITE = SSRS RED = SRSS

Results – Moderate Correlation (0.620) Sample size n = 15n = 30n = 60n = 120 #strata = % 61.7% 66.1% 60.0% 66.7% 59.1% 66.5% 58.6% #strata = % 60.0% 62.2% 59.3% 62.0% 59.2% 61.8% 58.5% #strata = % 58.7% 59.0% 58.8% 58.4% 59.3% 58.0% WHITE = SSRS RED = SRSS

Conclusions Can improve precision of survey estimation by using RSS in place of SRS SRSS will improve estimation for all variables in a survey, no matter how low the correlation SRSS may not require collecting additional information

Future Research Use different variables for stratification and ranking Performance under optimal strata allocation Do results hold for any sampling design that uses SRS in its final stage? Cost considerations