Answers
  1. Load the BKF94.daf file and initialise its sampling design in sheet file1.

 
Initialising  the sampling weight: EDIT|Set Sample Design

 

  • After choosing the appropriate variable(s), click on  the button OK.

Use the DAD Application: Poverty|FGT Index

  • After choosing the appropriate files, click on the button OK.
 

  1. Compute an estimate of the poverty incidence and of its confidence interval when:

    • NVIE is the variable of interest

    • The poverty line is 41099 F CFA

    • The asymptotic approach is used.

Performing the estimation

  • After choosing the appropriate variables and parameters, click on the  Compute button.

RESULTS

FGT (Poverty)

Session Date Tue Jan 04 11:42:33 EST 2005
Execution Time 0.219 sec
FileName BKF94.daf
OBS 8625
Sampling Weight WEIGHT
Variable of interest NVIE
Size variable SIZE
Group variable No Selection
Group Number 1
Option Normalised = NO
Parameter α=0.0

Index

Estimated value

Standard deviation

Lower bound

Upper bound

Confidence Level in (%)

FGT

0.44456312

0.01612363

0.41296138

0.47616486

95.00000000

Poverty line

41099.00000000

0.00000000

41099.00000000

41099.00000000

95.00000000

EDE

0.00000000

0.00000000

0.00000000

0.00000000

95.00000000

 
 

  1. Repeat the same procedure by adopting the bootstrap approach.
  • Click on the button SD STD, the following windows appears

  • Select the appropriate aptions and click on the button Confirm

  • Click again on the button Compute

RESULTS
 

FGT (Poverty)

Session Date Tue Jan 04 11:47:54 EST 2005
Execution Time 14.984 sec
FileName BKF94.daf
OBS 8625
Sampling Weight WEIGHT
Variable of interest NVIE
Size variable SIZE
Group variable No Selection
Group Number 1
Option Normalised = NO
Parameter α=0.0

Index

Estimated value

Standard deviation

Lower bound

Upper bound

Confidence Level in (%)

FGT

0.44456312

0.01702745

0.41028485

0.47372371

95.00000000

Poverty line

41099.00000000

0.00000000

41099.00000000

41099.00000000

95.00000000

EDE

0.00000000

0.00000000

0.00000000

0.00000000

95.00000000

 

 

  1. Draw the curves of the estimated incidence indices of poverty.
For this question we use DAD and Stata 8

Steps in DAD

  • From the main menu of the Spreadsheet, select the item EDIT| Edit Last Bootstrap results
  • Select the 200 estimated values and copy its in the Sheet file2
  • Select the item Edit|Set OBS : obs=200
  • Use the application Distribution|Density function
  • Click on button Range : Min= 0.38 | Max=0.52
  • Click on button Graph
  • Select the item Files|Save As and save this coordinates of  density according to the bootstrap approach.

Steps with Stata

  • Load the file: File|Import|Unformatted ASCII Data (new variables X boot)
  • Generate the asymptotic distribution: gen asym=normden(X,0.44456312, 0.01612363)
  • Drawing the graph: line asym boot X, sort

 

 
 

  1. Load the file BKF98.daf in the sheet file2 and initialise its sampling design. Consider that the two distributions are independent.

  2. Compute an estimate of the poverty incidence as well as of its confidence interval when:

    • NVIE is the variable of interest

    • The poverty line is 72690 F CFA

    • The asymptotic approach is used.
 

RESULTS

FGT (Poverty)

Session Date Tue Jan 04 12:28:01 EST 2005
Execution Time 0.0 sec
FileName BKF98.daf
OBS 8478
Sampling Weight WEIGHT
Variable of interest NVIE
Size variable SIZE
Group variable No Selection
Group Number 1
Option Normalised = NO
Parameter α=0

Index

Estimated value

Standard deviation

Lower bound

Upper bound

Confidence Level in (%)

FGT

0.45267603

0.01092681

0.43125987

0.47409218

95.00000000

Poverty line

72690.00000000

0.00000000

72690.00000000

72690.00000000

95.00000000

EDE

0.00000000

0.00000000

0.00000000

0.00000000

95.00000000

 

  1. Estimate the difference and its standard error and use the Distribution|Confidence interval to test if the difference is null.

Use the DAD Application: Poverty|FGT Index

  • After choosing the appropriate files, click on the button OK.

After choosing the appropriate variables and parameters, click on the  Compute button.

 

RESULTS

FGT (Poverty)

Session Date Tue Jan 04 12:38:37 EST 2005  
Execution Time 0.062 sec  
FileName BKF98.daf BKF94.daf
OBS 8478 8625
Sampling Weight WEIGHT WEIGHT
Variable of interest NVIE NVIE
Size variable SIZE SIZE
Group variable No Selection No Selection
Index of Groups 1 1
Option Normalised = NO
Parameter(s) α=0 α=0
Estimate 0.45267603 0.44456312
  (0.01092681) (0.01612363)
Difference Index1-Index2 0.00811291
  (0.01947734)
Covariance Index1-Index2 0.00000000
Poverty Line 72690.00000000 41099.00000000
  (0.00000000) (0.00000000)

 

Use the DAD Application: Distribution|Confidence Interval

  • After choosing the appropriate files, click on the button OK.
 

Confidence Interval (Distribution)

Session Date Tue Jan 04 12:41:01 EST 2005
Execution Time 0.015 sec
Sampling Weight WEIGHT
Interval option Two-sided

Estimated value

Standard deviation

Lower bound

Upper bound

Confidence Level in (%)

0.00811291

0.01947734

-0.03006197

0.04628779

95.00000000

           Hypothesis testing: H0:parameter = 0.00000000 vs H1: parameter ≠ 0.00000000

Result of the test

Significance Level in (%)

P-Value

H0 is accepted

5.00000000

0.67702171