THIS week I have been picking up some unexpected tips on parenting on the internet.

Apparently everything I've been doing so far is wrong. Which explains a lot.

Turns out that praising your kids is absolutely the worst thing you can do.

Well, that's if you believe the latest book on the subject — NurtureShock (subtitled 'Why everything we think about raising our children is wrong').

It's written by two American journalists, Ashley Merryman and Po Bronson, and it's hard taking advice from someone called Po, who I thought was the cute little red one with a scooter in the Teletubbies.

The pair are in fact serious, science journalists and they've taken a systematic sweep through all the relevant data over the last 100 years.

And the conclusion? Most of modern society's strategies for nurturing children are backfiring.

In truth the subtitle goes too far —it's not everything we think about raising our children that's wrong. Lots of what they've found isn't so different from what generations of sensible parents have always done.

On sleep for example, research shows that the loss of one hour for an 11-year-old is the equivalent of slipping to the academic level of a nine-year-old. And in teenagers the loss of each hour of sleep sent the odds of obesity up by 80 per cent.

Research on sleep at the University of Minnesota, where 7,000 high schoolers were surveyed about the connection between their sleeping habits and grades, also provided some fascinating results. "Teens who received A grades averaged about 15 more minutes sleep than the B students, who in turn averaged 15 more minutes than the Cs and so on," the authors say.

Sleep, the study showed, had such a remarkable effect on brain performance that even an extra quarter-hour made a difference. "Tired children can't remember what they just learned because neurons lose their plasticity, becoming incapable of forming the new synaptic connections necessary to encode a memory."

The book and its finding have caused such a stir in the states that it's already influencing policy in America, where high schools are starting later in the mornings.

It makes sense but it's not really news. We all know that tired children can't concentrate in school and tired teenagers won't feel like taking exercise, and that's linked to obesity. No shock there.

But I was truly taken aback by the evidence that you shouldn't keep telling your children how great they are.

NurtureShock says if little Sophie comes home from nursery proudly clutching a dreadful splodge painting and you shower her with compliments and tell her she's a prodigy, and you stick it on the kitchen wall (with all the others) you're telling her she's a genius who doesn't have to try. And she gives up.

Whoops! That's exactly what I did! Mind you, I genuinely did think my children's art was lovely.

"It's okay to offer a compliment if something is really, really good and the compliment is meaningful, but parents who are telling their children they're amazing for sliding down a slide? That's a problem," says Merryman, laughing. "I mean, you're praising them because gravity took its course."

In one study of 10-year-olds in New York, children were set a test. Afterwards, half of them were told, 'You must be smart at this'. The others were told, 'You must have worked really hard'.

They were then offered a choice of a second test, one was easy and one hard. All the kids who had been told they were clever picked the easy test the second time.

I'm not sure whether they thought they were already smart enough, or whether the weight of expectation was too much for them. But interestingly, those who were praised for working hard were brave enough to go on and take the harder test.

They call it 'the inverse power of praise' and describe another experiment in which girls who were praised for their intelligence then failed in tests.

The leader of the study, Dr Carol Dweck, concluded that praise can backfire. "Emphasising effort gives a child a variable they can control. They come to see themselves in control of their success. Emphasising natural intelligence takes it out of a child's control, and it provides no good recipe for responding to a failure."

In wider studies, the relentless pressure on a child branded a genius is now being seen as a prime reason for them falling back at school.

So you can praise children for their efforts. But whatever you do, don't tell them they're bright or naturally talented, because they then give up and stop trying.

"Sure, your child may be special," the authors say. "But new research suggests that if you tell him that, you'll ruin him."

Now I've read all this, I must admit it makes a kind of sense.

I was discussing it with a friend who remembered how she changed schools at 16 and was told that she could only hope for a grade D at A-level French because she hadn't had the proper grounding in grammar. The result was that she worked twice as hard and ended up with a B in French and a C in English, even though everybody kept telling her that English was her best subject.

This morning I tried it on my son. He's having violin lessons. He's no genius, but he's got a great Suzuki teacher and he's doing quite well. The only problem is getting him to practise.

He's fine once he's started, but I have to badger him into picking up his instrument.

So this morning, instead of giving him grief, I said: "I've noticed how hard you've been working on your violin lately and it's really making a difference."

The next thing I knew he was downstairs practising, all on his own, without being asked. Which seemed nothing short of miraculous.