An original audio drama series from the BBC World Service, Fukushima tells the story of the 2011 disaster.

At Japan's Fukushima Daiichi Nuclear Power Plant.

We lose the Daiichi plant. Then we lose Japan as we know it.

Listen to the series now by searching for Fukushima, wherever you get your BBC podcasts.

Hello and welcome to more or less on the BBC World Service.

We are guides to the numbers all around us in the news and in life. And I'm Tim Harford.

When you read that a paper has been published in a scientific journal that it's been peer reviewed,

generally you'd like to believe that it's true or at least that it's likely to be true.

But is it? In recent years there's been increasing worry about the prevalence of dodgy research practices,

of sloppy science, even have outright fraud.

I wanted to discuss this with Michael Blasland.

Michael is well known to loyal listeners of more or less because Michael is one of the creators of more or less.

He's the author of a book called The Hidden Half.

And he is also the presenter of a recent BBC program titled The Truth Police,

which is all about the strange characters who are trying to root out fraud and sloppy research

and the struggles that they're facing.

I began by asking Michael to explain the background,

what's sometimes called a replication crisis in science.

One of the big moments was about 12 or 15 years ago now.

I think when a piece was published suggesting that about half of all published research was wrong.

Now the view is that that was probably a bit of an exaggeration.

But even if it was a lot of an exaggeration, it would still suggest that there was a significant problem

...