Science and mainstream media

What Reuters says (from here):

MADRID (Reuters) - The smallest planet discovered outside our solar system has been found by Spanish scientists.

I think we are very close, just a few years away, from detecting a planet like Earth, team leader Ignasi Ribas told a news conference on Wednesday.

The rocky planet, with a radius about 50 percent greater than the Earth's, circles a small red dwarf star 30 light years away in the constellation of Leo, said the scientists from Spain's Superior Council for Scientific Investigations (CSIC).

The planet, known as GJ 436c, was found by analyzing distortions in the orbit of another, larger planet around the star GJ 436, a technique similar to that used more than 100 years ago to discover Neptune.

With a mass about five times greater than Earth's, it is the smallest planet yet discovered outside the solar system and improving techniques are opening the way to discovering worlds ever more like our own.

…

What the paper abstract says (from here):

Most of the presently identified exoplanets have masses similar to that of Jupiter and therefore are assumed to be gaseous objects. With the ever-increasing interest in discovering lower-mass planets, several of the so-called super-Earths (1 M_earth<M<10 M_earth), which are predicted to be rocky, have already been found. Here we report the possible discovery of a planet around the M-type star GJ 436 with a minimum mass of 4.7+/-0.6 M_earth and a true mass of ~5 M_earth, which would make it the least massive planet around a main-sequence star found to date. The planet is identified from its perturbations on an inner Neptune-mass transiting planet (GJ 436b), by pumping eccentricity and producing variations in the orbital inclination. Analysis of published radial velocity measurements indeed reveals a significant signal corresponding to an orbital period that is very close to the 2:1 mean motion resonance with the inner planet. The near-grazing nature of the transit makes it extremely sensitive to small changes in the inclination.

Key differences:

• The paper doesn't actually say it's the smallest planet yet found. This is good, because it's blatantly untrue. In fact, planets of similar size have been known for years. This is the first planet of this size found orbiting a main-sequence star (of which our Sol, I gather, is an example).
• I only skimmed the actual paper after reading the abstract, but I saw no mention of the radius. As pointed out by a Slashdot commenter, orbital perturbation can tell us about mass, but not volume, unless we know the planet's density—which we don't (although the planet is probably rocky).

As mainstream coverage of scientific topics goes, this is actually pretty good—after all, they managed to present two key points (we found a planet and it's pretty small), but even so, the truth is severely distorted—in the direction, of course, of sensationalism.

Amusingly, I came across the Bad Astronomer's criticism of the Reuter's article just a few minutes before I read the article itself, so I was inoculated. Not that I particularly need it; I generally assume that a mainstream press version of a scientific discovery will be similar only insofar as the first paragraph will vaguely resemble the paper's abstract.