CH7 Distribution Free Inference: Computer-Intensive Techniques 1.Random Sampling 2.Bootstrap sampling 3.Bootstrap Testing.

Slides:



Advertisements
Similar presentations
Inferences based on TWO samples
Advertisements

Chapter 6 Sampling and Sampling Distributions
Sections 7-1 and 7-2 Review and Preview and Estimating a Population Proportion.
Sampling: Final and Initial Sample Size Determination
Sampling Distributions (§ )
Chapter 11- Confidence Intervals for Univariate Data Math 22 Introductory Statistics.
Terminology A statistic is a number calculated from a sample of data. For each different sample, the value of the statistic is a uniquely determined number.
5-1 Introduction 5-2 Inference on the Means of Two Populations, Variances Known Assumptions.
ESTIMATION AND HYPOTHESIS TESTING
Two Sample Hypothesis Testing for Proportions
Ch 6 Introduction to Formal Statistical Inference.
Lecture 3 Miscellaneous details about hypothesis testing Type II error
Estimation Procedures Point Estimation Confidence Interval Estimation.
Sample size computations Petter Mostad
Business Statistics: A Decision-Making Approach, 6e © 2005 Prentice-Hall, Inc. Chap 7-1 Introduction to Statistics: Chapter 8 Estimation.
Lecture Inference for a population mean when the stdev is unknown; one more example 12.3 Testing a population variance 12.4 Testing a population.
Inferences About Process Quality
15-1 Introduction Most of the hypothesis-testing and confidence interval procedures discussed in previous chapters are based on the assumption that.
Bootstrap spatobotp ttaoospbr Hesterberger & Moore, chapter 16 1.
CHAPTER 23 Inference for Means.
4.1 Introducing Hypothesis Tests 4.2 Measuring significance with P-values Visit the Maths Study Centre 11am-5pm This presentation.
One Sample  M ean μ, Variance σ 2, Proportion π Two Samples  M eans, Variances, Proportions μ1 vs. μ2 σ12 vs. σ22 π1 vs. π Multiple.
4-1 Statistical Inference The field of statistical inference consists of those methods used to make decisions or draw conclusions about a population.
Copyright © 2013, 2010 and 2007 Pearson Education, Inc. Chapter Inference on the Least-Squares Regression Model and Multiple Regression 14.
POSC 202A: Lecture 9 Lecture: statistical significance.
More About Significance Tests
NONPARAMETRIC STATISTICS
Sections 6-1 and 6-2 Overview Estimating a Population Proportion.
Population All members of a set which have a given characteristic. Population Data Data associated with a certain population. Population Parameter A measure.
PARAMETRIC STATISTICAL INFERENCE
+ Chapter 12: Inference for Regression Inference for Linear Regression.
Copyright © Cengage Learning. All rights reserved. 10 Inferences Involving Two Populations.
Ch9. Inferences Concerning Proportions. Outline Estimation of Proportions Hypothesis concerning one Proportion Hypothesis concerning several proportions.
Ch 6 Introduction to Formal Statistical Inference
Inference for Regression Simple Linear Regression IPS Chapter 10.1 © 2009 W.H. Freeman and Company.
Confidence intervals and hypothesis testing Petter Mostad
Large sample CI for μ Small sample CI for μ Large sample CI for p
Sections 7-1 and 7-2 Review and Preview and Estimating a Population Proportion.
Copyright © 2010, 2007, 2004 Pearson Education, Inc. All Rights Reserved. Section 9-1 Review and Preview.
Copyright © 2010, 2007, 2004 Pearson Education, Inc. Chapter 22 Comparing Two Proportions.
Limits to Statistical Theory Bootstrap analysis ESM April 2006.
Introduction to Statistical Inference Jianan Hui 10/22/2014.
Chap 7-1 A Course In Business Statistics, 4th © 2006 Prentice-Hall, Inc. A Course In Business Statistics 4 th Edition Chapter 7 Estimating Population Values.
Inference with Proportions Review Mr. Hardin AP STATS 2015.
Statistical Inference Drawing conclusions (“to infer”) about a population based upon data from a sample. Drawing conclusions (“to infer”) about a population.
Business Statistics: A Decision-Making Approach, 6e © 2005 Prentice-Hall, Inc. Chap 7-1 Business Statistics: A Decision-Making Approach 6 th Edition Chapter.
Inference for Proportions Section Starter Do dogs who are house pets have higher cholesterol than dogs who live in a research clinic? A.
MATB344 Applied Statistics I. Experimental Designs for Small Samples II. Statistical Tests of Significance III. Small Sample Test Statistics Chapter 10.
Statistical Data Analysis 2011/2012 M. de Gunst Lecture 4.
Chapter 22 Comparing Two Proportions.  Comparisons between two percentages are much more common than questions about isolated percentages.  We often.
1 Probability and Statistics Confidence Intervals.
Module 25: Confidence Intervals and Hypothesis Tests for Variances for One Sample This module discusses confidence intervals and hypothesis tests.
Two-Sample-Means-1 Two Independent Populations (Chapter 6) Develop a confidence interval for the difference in means between two independent normal populations.
Copyright (c) 2004 Brooks/Cole, a division of Thomson Learning, Inc. Chapter 7 Inferences Concerning Means.
 List the characteristics of the F distribution.  Conduct a test of hypothesis to determine whether the variances of two populations are equal.  Discuss.
4-1 Statistical Inference Statistical inference is to make decisions or draw conclusions about a population using the information contained in a sample.
CHAPTER 8 ESTIMATING WITH CONFIDENCE 8.2 Estimating a Population Proportion Outcome: I will state and check all necessary conditions for constructing a.
Independent Samples: Comparing Means Lecture 39 Section 11.4 Fri, Apr 1, 2005.
Estimating standard error using bootstrap
More on Inference.
One-Sample Inference for Proportions
STAT 312 Chapter 7 - Statistical Intervals Based on a Single Sample
Confidence Intervals and Hypothesis Tests for Variances for One Sample
Week 10 Chapter 16. Confidence Intervals for Proportions
When we free ourselves of desire,
CONCEPTS OF HYPOTHESIS TESTING
More on Inference.
CONCEPTS OF ESTIMATION
BOOTSTRAPPING: LEARNING FROM THE SAMPLE
CHAPTER 6 Statistical Inference & Hypothesis Testing
Presentation transcript:

CH7 Distribution Free Inference: Computer-Intensive Techniques 1.Random Sampling 2.Bootstrap sampling 3.Bootstrap Testing

Why do we need distribution free-inference? For parametric models we need to: Specifying the function form Or the parametric family of the distribution.

In this Chapter We do not rely on the function form or the parameter; We rely on the observations in our sample; Present computer-intensive techniques for inference;

7.1 Random Sampling from Reference distribution Generate the Empirical Reference Distribution for some statistics.(Once we have such Distribution, we can find out the p-value of our test). 1.Generating the Empirical Reference Distribution in the parametric model, 2.Generating the Empirical Reference Distribution in the nonparametric model.

Generating the Empirical Reference Distribution in the parametric model Idea: (1) N, the repetition time is given. (2) Generate n observations from the distribution under Null hypothesis(with the claimed parameter), calculate the value of the interested statistic. (3) Repeat (2) N times, and get N values of the statistic. (4) Build the distribution table of the statistic-the Empirical Reference Distribution.

Ex1:In a manufacturing process, the proportion of plates from the process having more than one blemish need to be controlled under 10%. Suppose that in a sample of 50 plates we find 8 blemished plates, shall we be satisfied? Solution: Generate 50 obs from Bin(1, 0.1),let X 1 =( # of blemish among the 50 obs)/50. If we repeat this 1000 times, we ’ ll get X 1,X 2, …… X From these 1000 statistics, we can build a histogram (the empirical distribution) and find the p-vlaue of the test.

Generating the Empirical Reference Distribution in the Nonparametric model Idea: (1) N, the repetition time is given. (2) Generate n observations from the distribution under Null hypothesis(with the claimed info), calculate the value of the interested statistic. (3) Repeat (2) N times, and get N values of the statistic. (4) Build the distribution table of the statistic-the Empirical Reference Distribution.

Ex2:Consider a Hypothesis on the length of aluminum pins(Lengthwcp in Almpin.dat).Our Sample avg=60.028, and want to test H 0 : pop avg=60.1 vs H 1 : pop avg<60.1 Sol:In our sample we have z 1,z 2,……z 70.Let y i = z i +( ).Then Avg of y’s is 60.1 and the dist of y’s is just a shift of dist of z’s. We draw 70 obs from y’s WR, we can calculate the ave of these 70 obs, and denote it by X 1.If we repeat this 1000 times,we’ll get X 1,X 2,……X In the Textbook page 233, there is a table7.2. Since the 0.01 quantile of the dist is >60.028, p-value of the test<1%, so reject H 0 …..

Bootstrap Sampling 7.2 Bootstrap Sampling the Bootstrap Method 1.The Bootstrap method was introduced in 1970 by B.Efron. 2. The method performs statistical inference by computer and without the extensive assumption and intricate theory.

Details of Bootstrap Method

Properties of the EBD It’s centered at the sample statistics t n. The mean of the EBD is an estimate of the mean of the sampling distribution of the statistic T over all possible sample. The SD of the EBD is the Bootstrap estimate of the standard error of T We can find the alpha quantiles of the EBD as the quantiles limits of the distribution of T(for the purpose of CI).

Example 7.3 (1)Generate n=100 obs from Normal(20.5,12.5^2). (2)From the above obs, assuming we do not know its original distribution,find out the estimator of mean and it’s Bootstrap confidence interval. (3)Use sample average (18.709) to estimate the population average(20.5).Let M=100,use Bootstrap method, then we get 100 Bootstrap sample means(BSMs). The SD of these 100 BSMs is an estimator of SE of (sample average). (4)Let alpha=5%, the 97.5% quantile of BSMs is , the 2.5% quantile of BSMs is Thus the Bootstrap confidence interval for the mean is ( , ).

7.2.2 Examining the bootstrap method Question: Is Bootstrap Method good? Answer:

7.2.3 Harnessing the Bootstrap Method Sometimes it is hard to calculate a statistic when it has a very complicated formula( or you may have no idea about the formula), use bootstrap method make things easier(e.g look at the formula for of page 167). And when sample size is large, the approximation is very precise in many cases.

7.3 Bootstrap Testing of Hypothesis Idea: (1)Under Null hypothesis, construct an empirical reference distribution for interested statistic. (2) Based on the empirical reference distribution, find out the corresponding p-value…..

7.3.1 Bootstrap testing and CI for the mean

7.3.2 Studentized Test for the mean

Ex7.4: In data file”Hybrid1.dat” for Res3 with n=32 obs. Avg=21.434, H 0 :  =2150, Sol: M=500 (1) use bootstrap test for mean/confidence interval find out that 95% confidence interval is (2109.5, ) which cover So do not reject H 0. (2) use bootstrap studentized test find out that t n =-0.374, but p-value=.708. not reject again.