You're reading: News

Press release mayhem

On Google+ (sadly in a post with limited visibility, so I can’t link directly to it), Rongmin Lu (via David Roberts) highlights a case of “american whispers”, where a piece of research is helped along by press releases and media paraphrasing to become a completely different result.

Here’s how American whispers works:

1. You publish a paper, say on a new approximation to the discrete Fourier transform. To show the relevance of your work, you then say something like your new algorithm “improve[s] over the Fast Fourier Transform”.

2. Next, your institution’s press office issues a press release. To make it sound fun, they come up with a snazzy title “Faster-than-fast Fourier transform”. Pretty neat, huh?

3. Finally, some news website picks it up and then, suddenly, it’s all about “a new way of calculating Fast Fourier Transforms”. Ta-da!

I think you’d all agree that it’s way better than Chinese whispers.

Sergey Ten commented, saying that the press release in question wasn’t too bad, and mentions the idea that “random” data from real-world measurements is usually spread around a manifold of lower dimension than the sample space, which I think is the idea behind the paper Barcodes: the persistent topology of data, which I linked to in my last Interesting Esoterica summation.

On a similar note, Nalini Joshi points out that it isn’t news when centuries-old maths is used to solve a new problem: http://www.physorg.com/news/2012-03-combining-centuries-old-mathematical-theorems-efficient.html

Update: Rongmin’s original post is hidden to the public, so I’ve pasted it in here. I hope the limited visibility was a side-effect of the way Google+ works and not a deliberate decision to restrict the post’s audience.

(will not be published)

$\LaTeX$: You can use LaTeX in your comments. e.g. $ e^{\pi i} $ for inline maths; \[ e^{\pi i} \] for display-mode (on its own line) maths.

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>