Next Contents


What can a statistician, feet firmly planted on the ground, offer an astronomer whose head is up in the stars? The answer of course is: "quite a lot", which is why we are at this meeting. I shall talk about my own experience, giving an account of how statistics has enriched my astronomical research. Besides formalizing errors and making estimates of uncertainties more rigorous, it has actually opened up new ways to do analysis, providing new insights and approaches to problems which previously had proved elusive.

Small samples are a common occurrence in the field of Astronomy. Data has to be gathered from the laboratory of the universe, over which one has no control. Astronomers have to take what is given, and very often that is very little. In this context, I shall discuss three separate topics which I have been involved in, where statistics has played a major role in furthering our understanding. Please note that this in no way is an attempt at a review, or is representative, of the many uses of statistics in astronomy. It is a description of my personal involvement and excitement at the realization of how much the rich field of statistics has to offer to astronomy. The three applications that I shall talk about here are:

  1. The application of bootstrap to estimate the standard errors in measuring the galaxy-galaxy correlation function, and other measures of galaxy clustering.

  2. A technique developed to estimate whether large-scale filamentary structures in the universe are statistically significant.

  3. An application at extreme value statistical theory to the understanding of the nature of the brightest galaxies in rich clusters.

Next Contents