Varsha Venugopal

I think one big question that an RCT can't answer is about the mechanism.

So

What is fundamentally changing at the community level as a result of these interventions?

So SMS reminders, we think, are increasing salience.

Ambassadors, we hypothesize, are somehow changing trust and norms in the community.

But this is something that's still an open question that we would need some additional clarity on.

I think another element of this is any sense on where most of the effects come from would be really useful to know.

Could we have a few super ambassadors, for instance, so we don't have to go through the whole nomination process?

So again, that's a question we hope to get answers to through our own M&E processes.

I think other features of the study design, a big one is, does the impact change over time?

Now,

If these ambassadors, as we suspect, are actually changing norms and trust, then we don't need to be hiring or finding ambassadors every year.

Maybe it's a one-off thing.

But obviously, again, our cost-effectiveness calculations and our ideas on impact changes if we have to go back in and find ambassadors on a routine, say, annual basis.

The other bit is the SMS.

In the J-PAL study, they only enrolled children who had already shown up for at least one appointment, and they only sent out SMSs to either 33% or 66% of those people.

In our case, we enroll all children who have been born in a public facility, and we send out messages to all of them.

So

intuitively, we think we have higher impact on that one.

But again, open to studying further.