Managerial vs Human Capital Training
Last registered on February 18, 2020


Trial Information
General Information
Managerial vs Human Capital Training
Initial registration date
February 18, 2020
Last updated
February 18, 2020 2:21 PM EST

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
Harvard University
Other Primary Investigator(s)
Additional Trial Information
On going
Start date
End date
Secondary IDs
This project evaluates training for managers versus front line workers in the context of the low-cost private school sector in Pakistan.
External Link(s)
Registration Citation
Macdonald, Isabel. 2020. "Managerial vs Human Capital Training." AEA RCT Registry. February 18.
Experimental Details
The intervention includes teacher and school owner training aimed at low cost private schools.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
School revenue, school enrollment
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Outcomes to assess school growth:
• School fees and fee collection
• Assets including toilets, library, computers, desks, chairs, science labs, water cooler, generator, fans, playground, and boundary wall
• Investment amount and type
• Interest in future borrowing (self-reported)
• Student drop-out rates
• Quality indicators for management from business assessment
• School closures, which is likely to be low due to solvency and school age requirements to receive loan (prerequisite to enter sample)
• School expenses (self-reported)
• Proportion owner income from school vs other sources
Other outcomes:
• Teaching and learning quality indicators
• Teacher wages
• Teacher turnover
• Teacher learning on pre/post training survey
Secondary Outcomes (explanation)
• Management and learning quality indicators assessed as a checklist
Experimental Design
Experimental Design
Schools are randomly assigned to receive worker (teacher) training, manager (school owner) training, both trainings, or no trainings.
Experimental Design Details
Not available
Randomization Method
Randomization using Stata
Randomization Unit
Randomization is stratified on urban vs. rural location due to differences in the competitive landscape across these locations. Semi urban schools are grouped with rural schools due to greater similarities in school revenue, enrollment, and growth for schools with this classification in past data sets. Classification is done by the microfinance institution.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
200 schools
Sample size: planned number of observations
200 schools 200 school owners Number of teachers depends on how many school owners send to the training, but expect around 600
Sample size (or number of clusters) by treatment arms
50 schools control, 50 schools teacher training, 50 schools school owner training, 50 schools both trainings
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We conduct power calculations using a power threshold of 0.8, significance level of 5%, sample of 200, and historic variance in school enrollment growth for low cost private schools from past projects. The minimum detectable effect is 0.2 standard deviations, or 4.9 percentage points higher growth in enrollment for treatment than control groups. The MDE is likely lower when accounting for the stratification variable (location) which were not available in the past dataset used to estimate historic variance.
IRB Name
Harvard University- Area Committee on the Use of Human Subjects
IRB Approval Date
IRB Approval Number
Analysis Plan

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information