http://hea-www.harvard.edu/AstroStat/slog/2007/data-doctors/ Weaving together Astronomy+Statistics+Computer Science+Engineering+Intrumentation, far beyond the growing borders Fri, 01 Jun 2012 18:47:52 +0000 hourly 1 http://wordpress.org/?v=3.4 http://hea-www.harvard.edu/AstroStat/slog/2007/data-doctors/comment-page-1/#comment-25 vlk Fri, 15 Jun 2007 04:46:16 +0000 http://hea-www.harvard.edu/AstroStat/slog/2007/data-doctors/#comment-25 I think your characterization of "experiments" is too narrow. It is true that we cannot meddle with the environments of the objects we study, and almost all that we know of them comes to us via photons. However, that does not mean astronomers are passive data collectors who simply take what the Universe gives. We can, and do, design observations to provide specific answers to specific questions. When we use statistics, it is not so much to understand what our data are like, but rather to understand how much to believe the answers we get from our analyses. I think astronomers in general -- but maybe not the theoreticians! -- understand their data very well; it is the modeling that is subject to pratfalls, and is is there that statisticians must be most vigilant when speaking with astronomers. I think your characterization of “experiments” is too narrow. It is true that we cannot meddle with the environments of the objects we study, and almost all that we know of them comes to us via photons. However, that does not mean astronomers are passive data collectors who simply take what the Universe gives. We can, and do, design observations to provide specific answers to specific questions. When we use statistics, it is not so much to understand what our data are like, but rather to understand how much to believe the answers we get from our analyses. I think astronomers in general — but maybe not the theoreticians! — understand their data very well; it is the modeling that is subject to pratfalls, and is is there that statisticians must be most vigilant when speaking with astronomers.

]]>