It might sound like something from the far future but, according to Professor Sugata Mitra, we should be looking to give students access to the internet in exams sooner rather than later.

Mitra, professor emeritus at NIIT University Rajasthan, made the claim during a panel session at the World Education Summit during a discussion on how assessments - both high-stakes tests and informal in-class testing - needs to evolve.

His argument was that the ability to answer open-ended questions, such as “Does dark energy exist?”, and then giving students 30 minutes on the internet to research this and put together an answer, was a more effective away of testing for skills such as the ability to compute, comprehend and communicate rather than simply memory recall, which is what most assessments focus on.

It’s a big claim but Mitra was happy to discuss the idea further with Tes, to see how he envisions it working on a practical level and what the intended (and unintended) consequences might be.

Tes: How can we test students via the internet if we can’t guarantee the information is reliable or moderated?

SM: This reminds me of another time - people used to say “don’t believe everything you read in the newspapers”. With any source of information, this whole idea of misinformation exists, and the skill that you need is to distinguish between what is right and what is likely to be wrong.

It’s possible to set up really interesting exercises for children to find out what is right and what is not, and I’ve done that in many countries and given them a contentious issue - such as climate change.

You can try this with nine-year-olds, and, if they’re working in groups, they’ll quickly tell you which information is likely to be wrong and which is probably right.

Where you get my argument off guard is when it comes to asking, “At what age should they be taught?” I think that question is not yet answered.

Would it make more sense to restrict access to certain sites instead?

What I’m proposing is that we would not restrict anything because restricting anything online is really hard to do - as soon as you stop it, they will find ways around it.

Students would also be allowed to talk to other people online [because] this reflects life - we look up stuff. So, if I make an exam where you’re able to use all of these methods, then this measures your competence.

Would wealthier children with the internet at home have an advantage because they can practise this, compared with those who don’t have it?

Yes, and we are disadvantaging children [without connections], just as much as we are disadvantaging those children who can’t read as well as other children, and those children who can read but cannot comprehend as well as other children.

Intrinsically, we have advantages and disadvantages between ourselves. So a child who [doesn’t have connections] but can still figure out a way to answer the question - maybe by asking another child who does have a good connection - is, according to me, doing just as well.

Yes, there will be advantages and disadvantages children will have, but they’re not fair or unfair. There isn’t anything called an unfair advantage. If you have it, use it, if you don’t have it, find someone who does and use it.

In the end, it would even out. But, more importantly, we are interested in the children in an examination who can answer the question, as opposed to how they answer the question. The how used to be important but we’ve left that world behind.

So, internet access in exams could create more unfairness then?

It would make it unfair but no more unfair than the system is right now. I recently worked at Liverpool University, and the Chinese students were made to do assignments in English.

There were a few Chinese students who had the money to pay for their assignments to be done by private tutors - I could see that it wasn’t their work by comparing it to other answers they had completed. But the system says give them the marks they deserve based on the answer they’ve given, not based on your knowledge of them.

It is intrinsically unfair, this division of people by financial strengths, so it is no more unfair on the internet than it is in real life. The richer children have the advantage.

Are there dangers with asking children to spend more time online, given the risk of grooming and online pornography? Should we approach online learning with caution?

I have a clear idea on how to control this problem. I learned this from my Hole-in-the-Wall experiments - checking the history, I could see people were not accessing these kinds of sites.

I realised what was going on - these screens were visible, and people passing by could see what was accessed.

If children are given screens that are large and in safe public spaces, then nothing goes wrong. It’s very simple. The size of the screen matters - unfortunately the industry pushes smaller screens.

A child with a tiny device alone is in grave danger. A child with a large screen surrounded by other children in public, where that screen is visible, is safe.

If there are groups of children chatting, then the online groomer will quickly vanish once he realises there are other children there. You need four children for one screen, and all the children can see each other’s screens. When children are sharing a screen, the whole behaviour changes.

As such, I would advise against using tablets and mobile phones in schools.

Professor Sugata Mitra was talking to Tes senior analyst Grainne Hallahan