How do we communicate science and policy? Sadly, despite all the energy going into knowledge exchange these days, the answer is not very well. Despite excellent examples of joint working and the growth of science-led journalism, the proof of the pudding is still in the eating. And the overall statistics make for depressing reading.

In the UK, our levels of car use are the same as they were 10 years ago. Girls are smoking more not less and children are unsure if milk actually comes from cows.

Worse. In a country with oil and gas reserves that’s also been described as the Saudi Arabia of renewable energy, old people still die in winter because of badly insulated homes and unaffordable heating. And construction companies report that accessories such as kitchen tiles sell houses – not energy conservation features.

The beliefs and behaviours of the general population appear not to be scientifically or even evidence-based. Whose fault is that?

Take diet. At a meeting arranged by the Food Standards Agency, one expert had the same answer for any question: ‘five portions of fruit and veg’. Whether the question was about the safety of GM crops, the danger of too many dairy products, the cancer-avoiding qualities of soya or the state of school meals, her answer was always the same. And her point was forcibly made. People prefer complicated solutions to straightforward ones. That might result from a bit of displacement to avoid immediate action. But it might also be encouraged by the mixed messages sometimes coming from scientific studies and reports. Thus, between 1991-2001, while Britain got fatter and ate less fruit and vegetables, the vitamin supplement market doubled, the number of grocers halved and the number of reflexologists increased by 800%!

So how do scientists refine their messages? I’d suggest the need to send “messages” is actually part of the problem.

The internet has revolutionised the business of being informed. BBC research suggests under 50s in particular deeply resent being handed “tablets of stone” from people seen as authority figures. According to Mori 80% of the population don’t think those in authority know best. That suggests scientists need to learn from the best communicators and to tell stories instead of sending messages.

Changing misleading terms would also help. The Scotsman recently carried a correction to a story headlined: "Doctors botch three operations every day in Scots hospitals" (6 August 2009). They did this after a doctor pointed out that the ‘alarming rise in punctured organs’ included routine biopsies which use a needle.

Are stupid journalists to blame or a medical profession that insists on using jargon – or both? The need to create an evidence-based society is urgent and important. And scientists should be aware they have a leadership role in winning public acceptance for unpopular but necessary change.

Witness the smoking debate – transformed by the BMA’s unequivocal backing for a public smoking ban. Professor David Nutt may have lost the battle with the government over the classification of drugs – but his defiant stand has already won hearts and minds and he may yet win the war over reclassification of cannabis.

So scientists – be bold. Be clear. And become storytellers. Here are a few pointers.

• Change your language to be public facing.
• Believe you have only one part of the truth.
• Think human.
• Apply your findings. What does your science mean in the welter of everything else that is known. If you can’t weigh up risks and probabilities, punters will do it for you. Badly.