Monday, November 1, 2010

Truth in data

Where did this image come from? I first noted it on Science Daily. Ignore for the moment the bombast accompanied by most any Science Daily post. Well, wait don't ignore that :
  • Science Daily : "Brilliant X-ray Explosion."
  • MAXI : "faint X-ray transient."
Hmmm. Which is it then?
Perhaps the Science Daily author was completely blown away by the accompanying image. After all, in 5 days, this object in the center of the field-of-view became significantly brighter. A close look at the two images shows that all objects in this field-of-view are brighter on Oct. 17. Why were these data sets not normalized to each other? From just these images, I think it is obvious that the relative brightness of the central object is greater on Oct. 17th. I can only guess that the two different normalizations were included to produce a more dramatic image. Unfortunately, this drama comes at the price of honesty.
There is much effort to bring science results to non-scientists. This effort is a cross between public outreach (i.e. "please keep letting the NSF, NIH spend your tax dollars") and public education (i.e. "this here galaxy is 12 billion years old. [12billion much greater than 5000]".) Some of these efforts are official, and others are part of a greater media.
The short time-scales of today's mass media are unsuited to science reporting. Science Daily is "Updated several times a day with breaking news...". I doubt it. Science is slow and painstaking work. To give the public the impression that there is such a thing as "breaking news" in science is potentially damaging to research. To generate "breaking news", a researcher may be tempted or coerced to scale the false color in an x-ray image at the behest of an editor or PI. I wish Science Daily included a minimum of technical detail in their articles and captions. A lay reader may learn a little more than what someone said. And, the editors may have to think a little more critically about the content they are including.

No comments:

Post a Comment