Scribd Logo
Upload

Welcome to Scribd!

  • 281 views

    Exer Design Based Estimation

    Uploaded by

    Jhoanne Ynion
    This document discusses simple random sampling, stratified random sampling, and systematic random sampling estimators for estimating means, totals, and proportions from sample survey data. It provides formulas for calculating point estimates and measures of precision like variances and standard errors for sample means, totals, and proportions using different sampling designs. Several examples are also included to demonstrate how to apply the estimators and compute standard errors.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd

    Exer Design Based Estimation

    Uploaded by

    Jhoanne Ynion
    281 views9 pages
    This document discusses simple random sampling, stratified random sampling, and systematic random sampling estimators for estimating means, totals, and proportions from sample survey data. It provides formulas for calculating point estimates and measures of precision like variances and standard errors for sample means, totals, and proportions using different sampling designs. Several examples are also included to demonstrate how to apply the estimators and compute standard errors.

    Original Title

    Exer;Design Based Estimation

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses simple random sampling, stratified random sampling, and systematic random sampling estimators for estimating means, totals, and proportions from sample survey data. It provides formulas for calculating point estimates and measures of precision like variances and standard errors for sample means, totals, and proportions using different sampling designs. Several examples are also included to demonstrate how to apply the estimators and compute standard errors.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    281 views9 pages

    Exer Design Based Estimation

    Uploaded by

    Jhoanne Ynion
    This document discusses simple random sampling, stratified random sampling, and systematic random sampling estimators for estimating means, totals, and proportions from sample survey data. It provides formulas for calculating point estimates and measures of precision like variances and standard errors for sample means, totals, and proportions using different sampling designs. Several examples are also included to demonstrate how to apply the estimators and compute standard errors.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 9

    STAT 166: STATISTICS FOR SOCIAL SCIENCES Exercise: Design-Based Estimation

    SIMPLE RANDOM SAMPLING ESTIMATOR SRSWOR ESTIMATING THE MEAN () sample mean variance of sample mean standard error of sample mean sample total variance of sample total standard error of sample total sample proportion variance of sample proportion standard error of sample proportion

    SRSWR

    YSRS = Yi n
    i=1

    var YSRS
    se YSRS

    N-n s2 = N n

    var YSRS
    se Y SRS

    s2 = n

    N-n s2 = N n

    s2 = n

    ESTIMATING THE TOTAL (T)

    SRS =NY T
    var SRS T

    var SRS T

    s2 = N (N-n) n

    s2 =N n
    2

    se SRS T

    s2 = N (N-n) n

    2 2 s se TSRS = N n

    ESTIMATING THE PROPORTION (P)

    PSRS =a n P (1-P ) N-1 P (1-P ) Var (P ) = Var (P ) = n-1 N n-1 P (1-P ) N-1 P (1-P ) se (P ) = se (P ) = n-1 N n-1

    STAT166 1ST SEM 2012-2013 JCYnion

    STAT 166: STATISTICS FOR SOCIAL SCIENCES Exercise: Design-Based Estimation

    2 is known and Y ~ N(, 2)

    A (1-) X 100% confidence interval about Case 1 Case 2

    2 is unknown and Y ~ N(, 2)

    ' ! N-n $ 2 * ) Y SRS Z # & , # & ) N % n, ( + 2 " ' ! 2 $* ) Y SRS Z # &, # & ) " n %, ( + 2

    ' ! N-n $ s2 * ) Y SRS t # & , # ) (n-1) " N & n , % + ( 2

    SRS WOR

    ' ! s2 $* ) Y SRS t # &, ) (n-1) # n &, " %+ ( 2

    SRSWR

    A (1-) X 100% confidence interval about P ' * ! N-1 $ P 1-P , ) & )P Z # , # & " N % n-1 , ) 2 ( +

    SRS WOR

    ( )
    $ & & & %

    ! P 1-P # #P Z n-1 # 2 "

    STAT166 1ST SEM 2012-2013 JCYnion

    SRSWR

    ( )

    STAT 166: STATISTICS FOR SOCIAL SCIENCES Exercise: Design-Based Estimation

    STRATIFIED RANDOM SAMPLING ESTIMATION FOR EACH STRATUM PARAMETER STATISTIC Weight

    Ni W= i N
    Ni

    mean

    i= j=1
    Ni

    Y
    Ni
    ij

    ij

    Y= j=1 i ni
    2 ni i

    Y
    ij

    ni

    ij

    variance

    i2= j=1

    ( Y - )

    proportion

    Ni -1 P=Ai Ni i

    si2= j=1

    ( Y -Y )
    i

    ni -1 i P=ai ni

    STAT166 1ST SEM 2012-2013 JCYnion

    STAT 166: STATISTICS FOR SOCIAL SCIENCES Exercise: Design-Based Estimation

    Mean Variance of the mean total

    Ni Yst = Yi i=1 N
    L

    ESTIMATOR

    Yst = V

    ( )
    i=1
    L

    (Ni-ni ) si2 W2
    i

    Ni

    ni

    =NY Tst st
    Pst = Wi Pi
    i =1
    L

    si2 Variance of the V ( Tst ) = Ni2 (Ni -ni ) total ni i=1


    L

    proportion

    Ni -ni Pi 1-Pi Variance of the V (Pst ) = Wi2 proportion Ni ni -1 i=1


    SYSTEMATIC RANDOM SAMPLING Mean ESTIMATOR Total
    i

    ( )

    Proportion

    Ysy = i=1 n

    sy=NYsy T

    sy= a P n

    STAT166 1ST SEM 2012-2013 JCYnion

    STAT 166: STATISTICS FOR SOCIAL SCIENCES Exercise: Design-Based Estimation

    EXAMPLE: 1. A random sample of 10 households from an exclusive subdivision of 50 households gave the following monthly household incomes: 200,000 350,000 500,000 175,000 650,000 750,000 800,000 550,000 385,000 675,000 a. Estimate the true average household income of the households in that exclusive subdivision 10 200,000 + 350,000 + ... + 675,000 Yi y = = 10 10 i =1

    5,035,000 = 503,500 10 s = 220,945.57 # 50 10 & 220945.57 ( se y = % = 62,492.844 % ( $ 50 ' 10 CV y = 12.41%

    y =

    () ()

    b. Estimate the total income of households in that exclusive subdivision and its standard error.
    N y = 50 503,500 = 25,150,000
    se N y = "50 50 10 % $ ' # &

    ( )

    220,945.57 10

    = 3,124,642.216 se N y CV N y = = 12.42% Ny

    ( )

    ( )

    STAT166 1ST SEM 2012-2013 JCYnion

    STAT 166: STATISTICS FOR SOCIAL SCIENCES Exercise: Design-Based Estimation

    2. Five years ago, the Philippine population was considered relatively young. Of the 76.5 million Filipinos, about 32.9 million are aged 17 years old and below. This year, a random sample of 10,200 Filipinos was taken and 4,328 are aged 17 years old and below. a. Estimate the true proportion of Filipinos aged 17 years old and below and its standard error.
    4328 = 0.4243 10200 0.4243 1 0.4243 Var p = = 0.00002395 10200 se p = 0.00002395 = 0.0049

    p=

    ( ) ( )

    b. Construct a 90% CI for the proportion of Filipinos aged 17 years old and below

    ( ) ! $ # p Z ( ) se ( p )& " % !0.4243 1.645 0.0049 $ )% ( )( # & "


    /2

    p = 0.4243; se p = 0.0049; Z 0.10/2 = 1.645

    !0.4243 0.0080605$ " % !0.4162,0.4324$ " %

    3. An advertising firm is interested in determining how much to emphasize TV advertising in a given district. The firm decides to conduct a sample survey using STRS to estimate the number of hours per week that households watch TV. The following data
    STAT166 1ST SEM 2012-2013 JCYnion

    STAT 166: STATISTICS FOR SOCIAL SCIENCES Exercise: Design-Based Estimation

    were obtained from a sample of 40 households drawn proportionally from the 3 towns of the district.

    Town A Stratum size Sample Size Sample mean Sample variance Number of Household viewing Walang Hanggan 155 20 33.9 35.358 16

    Town B 62 8 25.125

    Town C 93 12 19.0

    232.411 87.636 2 6

    a. Estimate the true mean number of hours per week that households watch TV in this district. Give a measure of its precision. N N N y st = 1 y1 + 2 y2 + 3 y3 N N N 155 62 93 = 33.9 ) + 25.125 ) + ( ( (19.0 ) 310 310 310
    y st = 27.675

    STAT166 1ST SEM 2012-2013 JCYnion

    STAT 166: STATISTICS FOR SOCIAL SCIENCES Exercise: Design-Based Estimation

    V ( y st ) =

    Ni ni si2 Wi N n i =1 i i
    3 2 2 2

    155 155 20 35.358 62 62 8 232.411 = + 8 310 155 20 310 62 93 93 12 87.636 + = 1.97 310 93 12
    2

    b. Estimate instead the total number of hours per week that households in the district watch TV. Give a measure of its precision.

    se ( y st ) = 1.97 = 1.4036

    N y st = 310 27.675 = 8,579.25


    Var N y st = N 2Var y st = 3102 1.97 = 189,317 se N y st = 189317 = 435.1057

    ( )

    c. Estimate of the proportion of HHs viewing Walang Hanggan. Give a measure of its precision. Stratum Ni ni No. of HH Viewing MMS Pi 0.80 0.25 0.5

    Town A 155 20 16 Town B 62 8 2 Town C 93 12 6

    STAT166 1ST SEM 2012-2013 JCYnion

    STAT 166: STATISTICS FOR SOCIAL SCIENCES Exercise: Design-Based Estimation

    P = pst =

    Wi pi =
    i =1

    155 62 93 (0.80) + (0.25) + (0.50) 310 310 310

    P = pst = 0.6
    V ( pst )
    2 155 155 20 (0.8 )(0.2 ) = 19 310 155

    62 62 8 (0.25)(0.75 ) + 7 310 62
    2 93 93 12 (0.5)(0.5) + 11 310 93

    V ( pst ) = 0.0045 se ( pst ) = 0.0045 = .06708

    STAT166 1ST SEM 2012-2013 JCYnion

    You might also like