lies, damn lies, and ...
Feb. 8th, 2007 11:18 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
astroaztecAstronomers have long known about the dangers of choosing samples of objects based only on what is possible to observe. This is the Malmquist bias. It says that if you're not careful when looking at collections of distant objects you will gather a sample which is not representative of the actual conditions in the universe. More succinctly, if you are looking at the youngest (farthest away) objects in the universe, you'll tend to gather only the brightest of them, because those are the only ones you can see.
Astronomers have taken great care to avoid this bias by teaming observations from different systems. Among other strategies, the late, lamented ACS on the Hubble Space Telescope was used to identify distant candidates for later detailed study by the larger and more sensitive ground-based spectrometers like the ones on the Keck telescopes. But even when the astronomer botches the observing program, the result is just bad science. The objects are unaffected, and the failure of human understanding can be repaired.
Of late the US executive policy has driven the army's selection effects in quite the other direction. It's more than a bit disturbing to see young people being turned into statistics.
Astronomers have taken great care to avoid this bias by teaming observations from different systems. Among other strategies, the late, lamented ACS on the Hubble Space Telescope was used to identify distant candidates for later detailed study by the larger and more sensitive ground-based spectrometers like the ones on the Keck telescopes. But even when the astronomer botches the observing program, the result is just bad science. The objects are unaffected, and the failure of human understanding can be repaired.
Of late the US executive policy has driven the army's selection effects in quite the other direction. It's more than a bit disturbing to see young people being turned into statistics.
