24 March 2006
SETI League
PriUPS Project


Powers of Ten

What's wrong with the next sentence?

"Earlier studies of WMAP data have determined that the universe is 13.7 billion years old, give or take a few hundred thousand years."  (CNN 17 March 2006)

Nothing that couldn't be fixed easily if CNN weren't innumerate.  Not that I mean to single out CNN.  Show me a news outlet that doesn't make this sort of mistake on a daily basis and I'll show you a weekly publication.  It's taken most of this century for scientists to pin down the age of the universe to the approximate 13.7 billion years quoted above.  Doing so involved, among other issues, reconciling the age of stars in globular clusters which appeared to be older than the universe in which they resided. 

When scientists produce numerical results, they are usually accompanied by "error bars," so that if a graph shows a point with a value of 175,000 and an error bar of 4000, the scientist believes that his measurement is most likely 175,000, and is very unlikely to be more than 177,000 or less than 173,000.  These error bars can be large or small depending on the details of the measurement.  If, as was the case of the measurement of the age of the universe, one measurement fell outside error bars from those of a measurement that was supposed to be consistent, it meant that something was wrong.  After a lot of work, all the measurements agreed to a sufficient extent that 13.7 billion was believed to be a good estimate.  Assuming they're right, the universe most likely is less than 13.8 billion years old, and more than 13.6 billion.  If they had greater confidence, they might say the universe is 13.70 billion years old, or miraculously, 13.700 billion.  To agree with the CNN report, though, it would need to be an "order of magnitude" or even ten times better than that!  That's because CNN says "give or take a few hundred thousand years" instead of "give or take a few hundred million," as they should have.  Have I made my point?  Million?  Thousand? 

And that was only part one.  (No, not part 1.0027.)  

Richard Factor