AEA RC T Registr y

The American Economic Association's registry for randomized controlled trials

Three types of interventions are used in the trial:
- The psychologically active intervention consists of two back-to-back ten-minute aspirational videos, a facilitated goal-setting exercise, and the distribution of a reminder.
- The psychologically inactive intervention consists of two back-to-back ten-minute placebo videos, a facilitated placebo exercise, and the distribution of a reminder.
- The cash transfer intervention consists of the standard three-stage cash transfer program implemented by GiveDirectly.

In the course of the registration process for the cash transfers, eligible individuals are provided with a mobile money account as well as a mobile phone if they do not yet have these. This is followed by three mobile money transfers, made in intervals of approximately two months: a small transfer (`Token') of approximately USD100 (nominal 2016 dollars); a large transfer (`Lump Sum A') of approximately USD450; and a second large transfer (`Lump Sum B') of USD450 minus the price of the mobile phone.

2016-11-01

2017-06-30

The various outcomes will be specified in different pre-analysis plans that will be lodged before the outcomes are analyzed. Two initial analysis plans are enclosed in the initial registration (November 2016) for outcomes collected in the vicinity of the intervetion.

See pre-analysis plans

The project follows a 2x2 factorial design, and villages randomly allocated into one of four treatment arms:
(1) Psychologically active intervention
(2) Placebo intervention
(3) Cash transfer intervention plus psychologically active intervention
(4) Cash transfer intervention plus placebo intervention
Within arm (3), the timing of the psychologically active intervention is randomized at the level of individuals for the purposes of a sub-trial.

Not available

Stratified randomization by computer.

Main study: Randomized at the village level. Sub-study: randomized at the level of individuals.

Yes

Estimated: 370 -430 villages. Number is not final; recruitment is ongoing.

Estimated average: 18 participants per village. Number is not final; recruitment is ongoing.

Target distribution is 1/4 of villages in each arm; but randomization is phased and ongoing, and the overall number of villages may not be divisible by 4.

Manipulation Checks PAP

2018-03-13_timing_subtrial.pdf