Here's an example of the article: http://www.telegraph.co.uk/news/main.jhtml?
This "study" was released as a mixture of science and entertainment. As it turns out, there's not a spec of science in the story. It's a bogus survey produced by Veet, a hair removal company.
Ben Goldacre writes a column for The Guardian newspaper in Britain titled Bad Science. Goldacre reports on stories reported in the media as true when they are actually, well, bad science.
In the case of the Jessica Alba story, Goldacre actually received an e-mail from Veet. According to the e-mail, they already knew what they wanted the survey to say. They asked:
We are conducting a survey into the celebrity top 10 sexiest walks for my client Veet (hair removal cream) and we would like to back up our survey with an equation from an expert to work out which celebrity has the sexiest walk, with theory behind it. We would like help from a doctor of psychology or someone similar who can come up with equations to back up our findings, as we feel that having an expert comment and an equation will give the story more weight.
In a follow-up e-mail they admitted:
We haven't conducted the survey yet but we know what results we want to achieve. We want Beyonce to come out on top followed by other celebrities with curvy legs such as J-Lo and Kylie and celebrities like Kate Moss and Amy Winehouse to be at the bottom eg - skinny and pale unshapely legs are not as sexy..
The story said that the study came from Cambridge. Well, sort of. They went to Cambridge University and got a mathematician to analyze the data. There is no hard data in this survey at all. In other words, it's completely bogus.
Here's Ben Goldacre's column: http://www.guardian.co.uk/commentisfree/2007/sep/01/1