Weekly Tutoring Calls Nearly Doubled the Share of Kids Doing Division

When schools close, learning stops for millions of children. Pandemics, floods, typhoons and conflict all shut classrooms, sometimes for months.

The United Nations’ education emergency fund estimates that 222 million children face regular disruptions to their schooling. And even before COVID-19, fewer than half of primary-age children in poorer countries could read a story or do two-digit sums.

Laptops and internet lessons don’t reach most families in poorer countries. In low-income countries, fewer than 15% of households have internet access, the authors note.

But more than 70% have a mobile phone.

So can a simple phone call keep children learning? Five randomised trials across five countries, published in Nature, suggest it can.

And it costs about $11 a child.

Can tutoring by phone help children learn?

Yes, at least for basic maths, according to these trials. Children who got weekly tutoring calls learned clearly more than similar children who didn’t.

One-to-one tutoring is one of the best-studied ways to help children catch up. England’s Education Endowment Foundation estimates it adds about five months of progress, on average.

The problem is cost. In-person tutoring is expensive and hard to scale.

These trials tested whether a cheap phone call could deliver some of the same benefit.

The study: five trials, five countries

Where it happened

The trials ran in India, Kenya, Nepal, the Philippines and Uganda between December 2020 and July 2022. Schools in all of them had been disrupted by COVID-19.

Some had among the longest school closures in the world. One trial also had to cope with a typhoon.

What children got

Families were randomly assigned to one of three groups. Some got weekly text messages with maths problems and nudges.

Others got the texts plus a weekly 20-minute phone call with a tutor for 8 weeks. A control group got neither.

The calls used a method called “teaching at the right level”. Tutors quickly checked what each child could do, then pitched the lesson to match.

What the study found

Calls worked everywhere

Children who got phone calls plus texts improved their maths by 0.32 standard deviations on average, across all five countries. That’s a large effect for an education programme.

The researchers note a typical successful education programme achieves about 0.10. So these calls were roughly three times as effective.

A big jump in harder maths

The share of children who could solve division problems rose by 13.5 percentage points. That’s a 92% increase on the control group’s 14.6%.

Children also got better at fractions, which the calls didn’t directly teach.

Biggest gains where schools were shut longest

The effects were largest in Uganda and the Philippines. Children there had been out of school for almost two years.

In Uganda, children gained more than a full level on the maths scale, which runs from no operations to division.

Families stayed with it

Week after week, 70% to 80% of families took part in the calls. That’s high, especially compared with online, TV or radio lessons, which few families used.

Governments could run it too

In Nepal and the Philippines, the researchers also tested who made the calls. Government teachers did as well as charity staff, and slightly better on average.

That matters, because many programmes work when charities run them but fail when governments scale them up.

It changed teachers too

In Nepal, government teachers who made the calls became 9.3 percentage points more likely to tailor feedback to each child’s level. They were also more likely to say they’d choose teaching again.

These answers came from the teachers themselves, so they need some caution.

How cheap is it?

The calls cost about $11 per child. The authors estimate 3.6 years of high-quality schooling gained for every $100 spent, which they round to about 4.

That ranks it in the top 10 of 150 education programmes reviewed in poorer countries. By comparison, they cite a prominent tutoring programme costing about $2,500 per child.

Why the result needs caution

Basic maths only

The trials measured foundational numeracy, such as addition and division. They didn’t test reading or other subjects.

Emergency conditions

The trials ran while schools were closed, so the control group was getting very little teaching. The benefit might be smaller where children are already in school.

Who ran the study

Two of the lead authors co-founded Youth Impact, the charity that developed the phone tutoring programme. They say it didn’t deliver the calls in these countries and gained nothing financially from the results.

The data and code are publicly available, which allows others to check the work.

What earlier research found

The first trial

A 2022 trial in Nature Human Behaviour, in Botswana, first tested phone calls and texts during school closures. Learning improved by 0.12 standard deviations, ranking it among the most cost-effective ways to improve learning.

The new trials found larger effects, which is unusual. Programmes often lose their punch when they’re scaled up.

Tutoring in general

The Education Endowment Foundation rates one-to-one tuition as adding about five months of progress, on average. It judges this a moderate impact for a moderate cost.

How it fits together

Tutoring works, and teaching at a child’s level works. These trials show both can travel through a basic phone line.

How much should you trust this?

Strong. Five randomised trials in five countries, with consistent results, published in Nature. But they measured basic maths during school closures.

What makes it convincing

• Five separate randomised trials, in very different countries.
• Results were consistent, and larger in later trials.
• The data and code are public.
• Learning was tested directly with each child, not just reported by parents.

What makes me cautious

• Only basic maths was measured.
• The trials ran during unusual school closures.
• About a quarter of families weren’t reached at the final survey.
• The lead authors co-founded the charity behind the programme.

What this means for you

• If schools close, a regular call can help. Short, weekly, focused sessions made a real difference.
• Pitch lessons at the child’s level. The calls started from what each child could already do.
• Live contact matters. Text messages alone were much less reliable.
• Simple tools can work. These trials used basic phones, not apps or internet.
• For schools and charities, it’s a cheap option to keep ready. The authors suggest it could help outside emergencies too.

In this short J-PAL video, you can see how “teaching at the right level” helps children learn:

What we still don’t know

1. Do the gains last? The trials measured learning shortly after the calls ended.
2. Does it work for reading? Only maths was tested.
3. Would it help when schools are open? The control groups had little teaching.
4. How far can governments scale it? Two countries tested government delivery.
5. What about richer countries? All five trials were in low- and middle-income settings.

My take: a quietly brilliant result

What I like about this study is its simplicity. A phone call, 20 minutes a week, aimed at what the child actually knows.

I’m less convinced the effect would be as large in normal times. These children were missing school almost entirely.

Still, it’s rare to see an education idea hold up across five countries. It deserves to be ready for the next time schools shut.