Chapter 3: Inference about Population Proportions

Similar documents
Sequences and Series

16. Mean Square Estimation

ANOVA Notes Page 1. Analysis of Variance for a One-Way Classification of Data

n Using the formula we get a confidence interval of 80±1.64

Chapter = 3000 ( ( 1 ) Present Value of an Annuity. Section 4 Present Value of an Annuity; Amortization

Treatment Spring Late Summer Fall Mean = 1.33 Mean = 4.88 Mean = 3.

STATISTICAL PROPERTIES OF LEAST SQUARES ESTIMATORS. x, where. = y - ˆ " 1

Math 135 Circles and Completing the Square Examples

Online Appendix: Measured Aggregate Gains from International Trade

Chapter System of Equations

APPENDIX III THE ENVELOPE PROPERTY

The analysis of annuities relies on the formula for geometric sums: r k = rn+1 1 r 1. (2.1) k=0

FINANCIAL MATHEMATICS 12 MARCH 2014

An IMM Algorithm for Tracking Maneuvering Vehicles in an Adaptive Cruise Control Environment

An Effectiveness of Integrated Portfolio in Bancassurance

IDENTIFICATION OF THE DYNAMICS OF THE GOOGLE S RANKING ALGORITHM. A. Khaki Sedigh, Mehdi Roudaki

SHAPIRO-WILK TEST FOR NORMALITY WITH KNOWN MEAN

Graphs on Logarithmic and Semilogarithmic Paper

1. The Time Value of Money

Simple Linear Regression

n. We know that the sum of squares of p independent standard normal variables has a chi square distribution with p degrees of freedom.

MATH 150 HOMEWORK 4 SOLUTIONS

Lecture 3 Gaussian Probability Distribution

Classic Problems at a Glance using the TVM Solver

Unit 29: Inference for Two-Way Tables

Software Size Estimation in Incremental Software Development Based On Improved Pairwise Comparison Matrices

Example 27.1 Draw a Venn diagram to show the relationship between counting numbers, whole numbers, integers, and rational numbers.

Chapter 3. AMORTIZATION OF LOAN. SINKING FUNDS R =

Stock Index Modeling using EDA based Local Linear Wavelet Neural Network

of the relationship between time and the value of money.

Banking (Early Repayment of Housing Loans) Order,

The Time Value of Money

A. Description: A simple queueing system is shown in Fig Customers arrive randomly at an average rate of

10.5 Future Value and Present Value of a General Annuity Due

QUADRATURE METHODS. July 19, Kenneth L. Judd. Hoover Institution

Repeated multiplication is represented using exponential notation, for example:

ADAPTATION OF SHAPIRO-WILK TEST TO THE CASE OF KNOWN MEAN

Summation Notation The sum of the first n terms of a sequence is represented by the summation notation i the index of summation

Abraham Zaks. Technion I.I.T. Haifa ISRAEL. and. University of Haifa, Haifa ISRAEL. Abstract

Numerical Methods with MS Excel

Does Immigration Induce Native Flight from Public Schools? Evidence from a Large Scale Voucher Program

Use Geometry Expressions to create a more complex locus of points. Find evidence for equivalence using Geometry Expressions.

Binary Representation of Numbers Autar Kaw

The Gompertz-Makeham distribution. Fredrik Norström. Supervisor: Yuri Belyaev

15.6. The mean value and the root-mean-square value of a function. Introduction. Prerequisites. Learning Outcomes. Learning Style

Operations with Polynomials

PROF. BOYAN KOSTADINOV NEW YORK CITY COLLEGE OF TECHNOLOGY, CUNY

Optimal multi-degree reduction of Bézier curves with constraints of endpoints continuity

WHAT HAPPENS WHEN YOU MIX COMPLEX NUMBERS WITH PRIME NUMBERS?

Experiment 6: Friction

DlNBVRGH + Sickness Absence Monitoring Report. Executive of the Council. Purpose of report

A New Bayesian Network Method for Computing Bottom Event's Structural Importance Degree using Jointree

5.2. LINE INTEGRALS 265. Let us quickly review the kind of integrals we have studied so far before we introduce a new one.

CH. V ME256 STATICS Center of Gravity, Centroid, and Moment of Inertia CENTER OF GRAVITY AND CENTROID

Newton-Raphson Method of Solving a Nonlinear Equation Autar Kaw

Preprocess a planar map S. Given a query point p, report the face of S containing p. Goal: O(n)-size data structure that enables O(log n) query time.

ECONOMIC CHOICE OF OPTIMUM FEEDER CABLE CONSIDERING RISK ANALYSIS. University of Brasilia (UnB) and The Brazilian Regulatory Agency (ANEEL), Brazil

Capacitated Production Planning and Inventory Control when Demand is Unpredictable for Most Items: The No B/C Strategy

MATHEMATICS FOR ENGINEERING BASIC ALGEBRA

THE well established 80/20 rule for client-server versus

Churches on the Digital Cloud. Sponsored by Fellowship Technologies a partner in LifeWay s Digital Church initiative

Models of migration. Frans Willekens. Colorado Conference on the Estimation of Migration September 2004

Polynomial Functions. Polynomial functions in one variable can be written in expanded form as ( )

An Integrated Honeypot Framework for Proactive Detection, Characterization and Redirection of DDoS Attacks at ISP level

Statistical Pattern Recognition (CE-725) Department of Computer Engineering Sharif University of Technology

STATUS OF LAND-BASED WIND ENERGY DEVELOPMENT IN GERMANY

Public Auditing Based on Homomorphic Hash Function in

COMPARISON OF SOME METHODS TO FIT A MULTIPLICATIVE TARIFF STRUCTURE TO OBSERVED RISK DATA BY B. AJNE. Skandza, Stockholm ABSTRACT

Fractal-Structured Karatsuba`s Algorithm for Binary Field Multiplication: FK

How To Value An Annuity

The simple linear Regression Model

Bayesian Network Representation

Bayesian Updating with Continuous Priors Class 13, 18.05, Spring 2014 Jeremy Orloff and Jonathan Bloom

Speeding up k-means Clustering by Bootstrap Averaging

The Analysis of Development of Insurance Contract Premiums of General Liability Insurance in the Business Insurance Risk

Factoring Polynomials

Physics 43 Homework Set 9 Chapter 40 Key

Basically, logarithmic transformations ask, a number, to what power equals another number?

MDM 4U PRACTICE EXAMINATION

Helicopter Theme and Variations

LECTURE #05. Learning Objectives. How does atomic packing factor change with different atom types? How do you calculate the density of a material?

ALABAMA ASSOCIATION of EMERGENCY MANAGERS

AP Statistics 2006 Free-Response Questions Form B

Example A rectangular box without lid is to be made from a square cardboard of sides 18 cm by cutting equal squares from each corner and then folding

Credibility Premium Calculation in Motor Third-Party Liability Insurance

PREMIUMS CALCULATION FOR LIFE INSURANCE

10.6 Applications of Quadratic Equations

RQM: A new rate-based active queue management algorithm

A Study of Unrelated Parallel-Machine Scheduling with Deteriorating Maintenance Activities to Minimize the Total Completion Time

Integration by Substitution

Reinsurance and the distribution of term insurance claims

On formula to compute primes and the n th prime

Mathematics of Finance

Settlement Prediction by Spatial-temporal Random Process

T = 1/freq, T = 2/freq, T = i/freq, T = n (number of cash flows = freq n) are :

19. The Fermat-Euler Prime Number Theorem

Why is the NSW prison population falling?

CHAPTER 2. Time Value of Money 6-1

Curve Fitting and Solution of Equation

Common p-belief: The General Case

Transcription:

Chpter 3: Iferece bout Populto Proportos We re ofte cocered wth mg fereces bout populto proportos For exmple: - Accordg to recet Gllup poll, 60% of Amercs re dsstsfed wth the wy thgs re gog the Uted Sttes - I 990, the proporto of femle US s 53% Secto : Ifereces bout sgle populto proporto (Revso) Recll: Norml Approxmto to the Boml The boml rdom vrble X hs me µ p d stdrd devto σ p( p) If for boml dstrbuto s lrge ( p>5 d (-p)>5)we my use orml pproxmto to the boml X p - We stdrdzed X to obt Z N(0,) p( p) Remr: The smple proporto (, ( ) / ) Norml p p p where p>5 d (-p)>5 x umber of successes pˆ s pproxmtely totl umber of observto Hece, ) If pˆ > 5 d ( pˆ) > 5, 00 ( α)% CI for the populto proporto, p, s gve by pˆ ± z( α / ) s( pˆ) where s( pˆ) pˆ( pˆ) b) To test the hypothess H : p p 0 0 H : p p 0, use the test sttstc * pˆ p0 p0( p0) z where s( pˆ ) s( pˆ ) Reject H 0 f z * > z( α / ) The P-vlue s PZ ( > z * ) Pge of 9

Exmple: It hs bee reported tht pproxmtely 60% of US households hve two or more televso sets d tht t lest hlf of Amercs sometmes wtch televso loe Suppose tht 75 US households re smpled, d of those smpled, 49 hd two or more televso sets d 35 respodets sometmes wtch televso loe ) Two clms c be tested usg the smple formto Wht re the two clms? ) Costruct 95% cofdece tervl for the proporto of the Amercs tht hve two or more televso sets? Do the dt preset suffcet evdece to show tht the 60% fgure clmed the mgze rtcle s correct? 3) Do the dt preset suffcet evdece to cotrdct the clm tht t lest hlf of Amercs sometmes wtch televso loe? Pge of 9

Secto : Ifereces bout two populto proportos (Revso) x x Let pˆ d pˆ d ssume pˆ > 5, ( pˆ ˆ ˆ ) > 5, p > 5, d ( p) > 5, the ) A00( α)% CI for dfferece betwee the populto proportos, D p p, s gve by ˆ ˆ ˆ pˆ ( pˆ ) pˆ ( pˆ ) D± z( α / ) sd ( ) where sd ( ) + where D ˆ p ˆ p ˆ A c) To test the hypothess H : p p 0 0 H : p p 0, use the test sttstc z * Dˆ 0 ˆ pˆ ˆ ˆ ˆ ( p) p( p) where sd ( ) + sd ( ˆ ) Reject H 0 f z * > z( α / ) The P-vlue s PZ ( > z * ) Exmple: A expermet ws coducted to test the effect of ew drug o vrl fecto The fecto ws duced 00 mce, d the mce were rdomly splt to two groups of 50 The frst group, the cotrol group, receved o tretmet for the fecto The secod group, the expermetl group, receved the drug After 30-dy perod, the proportos of survvors the two groups were foud to be 036 d 060, respectvely ) Is there suffcet evdece to dcte tht the drug s effectve tretg the vrl fecto? ) Use 95% cofdece tervl to estmte the ctul dfferece the cure rtes for the expermetl versus the cotrol groups Pge 3 of 9

Sec 3: Ifereces bout Severl Proportos If rdom vrble X follows the gmm dstrbuto the the probblty desty fucto s gve by f( x) x e α β Γ( α) α x/ β Ch-Squre dstrbuto wth r degree of freedom s specl cse of gmm dstrbuto where α r / d β Recll tht f X hs the boml(,p) dstrbuto the x! PX ( x) p( p) where p p p!( p)! Proportos re relly just specl cses of mes To see ths, let x be or 0 f the th US ctze s mle or femle, respectvely, d let p represet the ctul proporto of mle ctzes The f N s the populto sze, N p x N So p s relly the verge of the N s d 0s Pge 4 of 9

Ths mes we c te smple of sze d estmte p wth the ubsed pot estmtor pˆ x, where X s the umber for the th rdomly chose perso Sce the populto of the US s rther lrge, we c vew the perso should be mle wth probblty p The X s s depedet Beroull(p) rdom vrbles Tht s, the th X ~ boml( p, ), whch mes tht for {0,,, } P pˆ P X P X p ( p) Ad hece the me d stdrd devto of ˆp re p d p( p) / respectvely (these results we hve used sectos d ths chpter, but we dd ot see why But ow we ow why ) The multoml dstrbuto s jot dstrbuto geerlzto of the boml dstrbuto Tht s, suppose tht depedet expermets re to be performed, ech of whch results outcome wth probblty p, outcome wth probblty p,, outcome wth probblty p Let X deote the umber of the expermets resultg outcome, the for d p, we hve! p PX ( x, X x,, X x) p p p!!!!! Π The expresso bove s the jot probblty mss fucto for the multoml dstrbuto We c verfy ths probblty expresso by otg tht the probblty of y prtculr sequece of the outcomes where evet occurs exctly tmes for,, s exctly Π p Also, the umber of these sequeces s exctly!!!! Pge 5 of 9

Exmple: Cutthrot s three-plyer gme of pool tht Joh, George, d Rgo le to ply (Pul s ded) Joh s very good, d ws 60% of the tme, George ws 30% of the tme, d Rgo ws 0% of the tme Suppose they ply te gmes Wht s the probblty tht Joh ws t lest fve gmes d George ws t lest four? Bc to sttstcs The ch-squre goodess-of-ft test s hypothess test for determg whether cert probbltes (or proportos) te o prtculr vlues To do ths, we perform multoml expermet, d record the observed umbers, of trls resultg ech outcome type,, We the compre these umbers to the expected vlues of the umbers of outcomes of ech type uder the ull hypothess Ths depeds o the test sttstc ( E ) X where E p E (expected cell frequecy) The Ch-Squre Goodess-of-Ft Test: H : p p, p p,, p p H 0 0 0 0 : t lest oe of the proportos dffers from ts hypotheszed vlue, Reject H o f X > χ ( α, ) Pge 6 of 9

Exmple: The uts populto cosst of oe of fve types A rdom smple of 300 uts s clssfed s follows: Ctegory Observed Cout, 60 50 3 30 4 40 5 0 Totl 300 It s hypotheszed tht H0 : p 00, p 05, p3 040, p4 05, p5 00 At α 005 level, do the 300 uts pper to be from populto wth these vlues for p Pge 7 of 9

Ch-Squre Test of Homogeety The followg tble s r ccotgecy tble where the r rows correspod to the r popultos d the c colums correspod to the c ctegores of clssfcto Ctegores of clssfcto Populto c Totl c c r r r rc r Totl c To test the hypothess H p p p H 0 : r : Not ll the populto proportos re equl Or o the other word H 0 H : The populto re homogeeous : The popultos re ot homogeeous We use the test sttstc ( ) r c j Ej where Ej (expected cell frequecy) χ E j Reject H o f X > χ ( α, ( r )( c )) Pge 8 of 9

Exmple: A resercher studed the chrcterstcs of subjects ttedg fve-dy hum sexulty progrm The results re show the followg tble: Mrtl Sttus Group sgle Mrred or Totl dvorced Medcl Studets 50 0 70 Nursg Studets 5 37 Other studets 6 8 4 Group leders Totl 69 74 43 Test whether or ot the four popultos represeted the study re homogeous wth respect to mrtl sttus Pge 9 of 9