Scribd
Upload
70 views12 pages

Lab 8-Impact Evaluation

The document discusses the process of evaluating a microcredit program in Bangladesh using a randomized controlled trial. It describes randomly assigning 300 eligible households to either a treatment group that participates in the program or a control group that does not. Additionally, it stratifies the sample to ensure equal numbers of female-headed households in each group. Finally, it proposes using a t-test to test whether the treatment and control groups have balanced proportions of gender to ensure internal validity.

Uploaded by

Rinda
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
70 views12 pages

Lab 8-Impact Evaluation

The document discusses the process of evaluating a microcredit program in Bangladesh using a randomized controlled trial. It describes randomly assigning 300 eligible households to either a treatment group that participates in the program or a control group that does not. Additionally, it stratifies the sample to ensure equal numbers of female-headed households in each group. Finally, it proposes using a t-test to test whether the treatment and control groups have balanced proportions of gender to ensure internal validity.

Uploaded by

Rinda
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 12

Lab 8-Impact Evaluation

Replikasi Modul PEP(Partnership for Economic Policy)


Outline

• Random Assignment to the Treatment


• Experimental Sample
• Replicability of Random Draws
• External Validity
• Treatment and Control Group
• Stratification
• Internal Validity
1. Random Assignment to the Treatment
2. Evaluation sample 3. Randomize
1. Population
X treatment

Comparison

Treatment
= Ineligible
External Internal Validity
= Eligible Validity
1. Random Assignment to the Treatment

Microcredit Program in Bangladesh


Suppose that we are asked to evaluate an upcoming microcredit
program for Bangladesh. Ideally, we would like to compare two identical
groups, with the only difference that one group would participate in the
program (treatment group) and the other would not (control group).

Use hh_91_practice.dta
Experimental Sample

• use "$data\hh_91_practice.dta", clear


• des
• gen random = runiform()
• sort random
• gen experiment = (_n <= 300)
• tab experiment
Replicability of Random Draws

• set seed 19320419


External Validity

gen experiment_20 = (_n <= 20)


# Plot
twoway (kdensity exptot) ///
(kdensity exptot if experiment ==1, lpattern(dash)) ///
(kdensity exptot if experiment_20 ==1, lpattern(shortdash)),///
legend(label(1 "survey") label(2 "large sample") label(3 "small sample"))
Randomization on Treatment and Control

• generate random_T = runiform()


• generate T = (random_T < 0.5) if experiment == 1 /* memilih
kelompok treatment dari 300 sampel */
• tabulate T, missing
Sratification

Suppose that we want to make sure that we have the same number of female
household heads in the control and the treatment groups. We do this by running
a separate lottery for the households with a female household head.

• sort sexhead random


• bysort sexhead: generate random_T_strat = runiform()
• bysort sexhead: generate T_strat = (random_T_strat < 0.5) if experiment == 1
• tabulate T_strat sexhead
Internal Validity

• Untuk membuktikan bahwa counterfactual yang dipakai adalah valid


menggunakan t-test
• misal : untuk menguji bahwa pada kelompok treatment dan control
memiliki proporsi gender yang seimbang (equal)

ttest sexhead, by(T)


Ho : kelompok treatment dan
control memiliki proporsi
gender yang seimbang (equal)

Ha : kelompok treatment dan


control memiliki proporsi
gender yang tidak seimbang
(inequal)

You might also like