It’s pretty difficult to get a completely accurate read on how bad things are after the Fukushima reactor. A great deal of this is due to news stories which focus on sensationalism without actually checking the facts.

Take a recent Al Jazeera story, titled “Fukushima: It’s much worse than you think.” Much of the story may or may not be true, but take this nugget:

In the US, physician Janette Sherman MD and epidemiologist Joseph Mangano published an essay shedding light on a 35 per cent spike in infant mortality in northwest cities that occurred after the Fukushima meltdown, and may well be the result of fallout from the stricken nuclear plant.

The eight cities included in the report are San Jose, Berkeley, San Francisco, Sacramento, Santa Cruz, Portland, Seattle, and Boise, and the time frame of the report included the ten weeks immediately following the disaster.

“There is and should be concern about younger people being exposed, and the Japanese government will be giving out radiation monitors to children,” Dr MV Ramana, a physicist with the Programme on Science and Global Security at Princeton University who specialises in issues of nuclear safety, told Al Jazeera.

Sounds pretty striking. However, when you actually think about it, it also sounds completely bogus. Radiation levels along the west coast of the U.S. barely showed any trace of radioactivity from Japan, and what was found was insignificant in the context of background radiation. For this to result in a 35% spike in infant mortality sounds rather ludicrous, and should not pass anyone’s initial sniff test.

Sure enough, the data was found to be incorrect, if not fraudulent. The essay cherry-picked eight cities which made a writer for Scientific American “pick up a whiff of data fixing,” and the essay used only a four-week period prior to the Fukushima disaster to establish a baseline for infant mortality. Coincidentally, those four weeks showed strikingly low mortality rates, most likely a statistical blip–from the fifth week back, rates were considerably higher on average. When the Scientific American reviewer checked the data, even for just those eight cities, for the entire year, he found that infant mortality rates were actually declining throughout the year, not spiking sharply.

The ten weeks after Fukushima did see a few high numbers, but the four weeks previous were unusually low. The low numbers before Fukushima were not, of course, related to the incident, but were the primary reason a “spike” seemed to appear when only those four weeks were used as a baseline. One could just as easily compare the data for the month before those four weeks and see a similar “spike” occurring two months before Fukushima.

In short, the essay’s claims should never have been published. Sure, the media might not want to wait around for every study to be peer-reviewed, but if you don’t, then you are bound to release, as “news,” completely erroneous data as shown above. This only serves to put all the rest of the claims in the Al Jazeera article in question–what else was not fact-checked? Was the writer going for the sensationalist angle, only choosing to focus on evidence which supported a predetermined conclusion?

This is not to suggest that Fukushima isn’t worse than we think it is–it may well be a lot worse–but it suggests that not all the alarmist information, including that released by “experts,” is as accurate as we believe. That goes for data suggesting that things aren’t so bad as well.

Because the atmosphere regarding nuclear power, even within the halls of science, is so subject to polarized views and infected by bias, added to the fact that there is so much we just don’t know for certain about the subject of radioactivity and health, it is much harder to get a clear picture of how bad things really are.