Can a woman's hemline really predict the weather?

The following is an article that appeared in Melbourne's Herald Sun (after they pinched it from England's Mail on May 31st 2010)...

If you want to know what the weather is up to, it might be best to start looking down. For apparently the length of a woman’s skirt is a good way of forecasting whether to expect rain or shine. The bizarre claim is the result of sales analysis by experts at the online marketplace eBay. They insist that the length of skirts sold rises several days before the arrival of sunny skies. And when the air is about to cool, they drop. This happens, they say, at least three days before the weather changes, and sometimes even before the forecasters at the Met Office have issued their most traditional predictions. The trend, they claim, correctly predicted Britain’s temperamental weather last year, as well as the recent short-lived heat wave.

‘We’re calling it Mini Skirt Meteorology,’ said eBay’s Ruth Szyszkowski. ‘If you want to know what the weather is going to be like in three days, just take a look at hemlines.’

During the recent hot spell, it is claimed that demand for short skirts soared by 200 per cent a full seven days ahead of the actual event – well ahead of the Met Office forecast. There are said to be distinct changes in skirt length with the rise and fall of the thermometer, typically starting in advance of the actual change.

Mostly, I find this humourous. But I wonder if it's actually got some merit? eBay sales analysts reckon it accurately predicted both the general weather in Britain last year, including a heatwave. So that means sales of higher hemlines went up days before the temperature, and vice versa. Several questions are raised:

1. Is the trend applicable only to Britain, or can it be observed worldwide? Does it only occur in temperate climates, or polar and equatorial ones as well? What garment would you use in those other regions to predict the weather?
2. I wonder if someone could get a bucketload of money from the relevant grants authority in order to test the hypothesis that hemlines predict weather. They could gather data from online as well as physical retail stores, and let us all know what great expense the taxpayer went to in order for us all to find out.
3. Could eBay start to offer a weather forecasting service?
4. Is there actually some truth to this? Do we have some innate, subconscious ability to anticipate the clothing required days before the mercury changes? Ants, spiders and flies are known to move to higher or drier ground before rain, and I bet they don't think about why they're doing it, or keep a running record of the skirt length of their colleagues; why should they monopolise this meteorological ability?