How do we break this cycle?

It's unstoppable, because if you remove any one of the cogs the cycle still works (though not as intensely).

The only cogs you can remove that prevent the system from working this way are the scientist (and that would defeat the point) or the grandma -- and good luck convincing every single person in the world to evaluate "science" journalism with a skeptical mind!

There's one more step missing from the end, and it's possibly the saddest of all:

The step where people star to doubt the validity of science because of all the hype about that A -> B link that didn't pan out.

To be completely pedantic, this is one arrow short of being a cycle. It's more like "Science News Entropy".

better science edumacation at the high school level, so that the people in news reports don't have to simplify so much for everyone to understand it

Better science education at all levels is important. By high school, it's too late, kids have already decided what's stupid and what's cool. Science has for too long been relegated to "geeky stuff", uncool, something children are socially punished for pursuing.

That has got to change. Get them young. Show them science is cool, full of wonders and amazement. If you get them young, they'll pursue it on their own.

As science journo I'm very aware some of us often do a pretty appalling job (it's one of the reasons I read this an other blogs, as a constant reminder of how and where we go wrong) - I'm not going to bother defending the profession here.

But a quick comment for those scientists who are frustrated at the way their research gets treated: your best defense against this is to be available. There's very little more frustrating than to see a release, be aware that there are holes in it, or there's a spin you're a bit iffy about, and to not be able to get hold of the people who did the research to check your facts.

The second part of that, though, is to only be available to journalists who do the right thing - I've spoken to more than a few scientists who've refused comment because they spoke to a journalist at Publication X who shafted them, or got it utterly wrong. Fair enough - and if it happens, don't talk to them, and make sure your colleagues know who it was so they can be wary also. And journalist lives and dies by the reputation they build with their industry contacts - it's worth being aware of that and acting accordingly.

(And it is true that a story is occasionally buggered up by subs and editors, not journalists - on the rare occasions that's happened to me I usually try to get off a preemptive apology to the person I spoke to, if a correction isn't possible).

It's also OK to ask a journalist to talk "on background" initially, when you run through the actual science (to make sure we actually get it). You can then go on the record to give direct quotes afterwards - attitudes to doing this vary between journos and publications, but it's always worth asking the questions (and I, personally, think it's good practice as a science journo to do it that way, particularly when I'm not as familiar with the subject material as I'd like to be, or when I'm not talking to someone who I have an established relationship with).

And man, my claim to be a credible journalist will have just taken a hit, looking at the number of typos in that last comment...

Kind of frightened to read that the news follows the weirdness of the internets instead of reporting the news first. But that absolutely explains why the crazy is out of control.

Sometimes the distortion happens in the University press office - e.g. Ben Goldacre reports about the Facebook and Twitter âmake us bad peopleâ story
http://www.badscience.net/2009/04/experts-say-new-scientific-evidence-h…
To fix this the university press office has to hurt when it distorts the science - at present I bet they feel "all publicity is good publicity" because it is advertising the university.

Is there any way to tell whether this cycle increases, or decreases, funding for research?

If it increases it, nobody really has any interest in changing anything, since clearly behaving this way is locally beneficial to each participant, even the grandmother figure who gets a little much-needed excitement in her life. A funding increase means it's globally beneficial, too.

If it decreases funding, there's a chance to persuade participants that their upstream flow of goodies may dry up unless they change their ways.

That would suggest it's crucial to ensure that grant-making bodies have an evidence-based process (where "our grant rocked the socks off CNN" is not admissible evidence).

The solution is to strike first, and teach your PR department to submit the story to Fark themselves after the release is made. This is apparently where most of Cable news goes to figure out what they want to spin.

Removing the intermediate steps may reduce the noise that creeps in in the additional steps. Of course, you then have to have someone in the PR department whose job is to see exactly how silly they can possibly make the school look, which tends not to go over too well at the stuffed shirt level.

I have an idea involving journalists and inserting radio controlled tasers where only customs men dare to probe...
Hmmm, we may need some customs men to help carry this one out

:)