Published in "Errors, Bias and Uncertainties in Astronomy", Cambridge University Press, 1990.
Out of print - February 1998

For a PDF version of the article, click here.


BOOTSTRAP, DATA PERMUTING AND EXTREME VALUE DISTRIBUTIONS: GETTING THE MOST OUT OF SMALL SAMPLES

Suketu P. Bhavsar

Department of Physics and Astronomy, University of Kentucky, Lexington KY 40506-0055, U.S.A.


Abstract. The use of statistical methods on samples of astronomical data are discussed. We discuss the application of bootstrap to estimate the errors in measuring the two-point correlation function for galaxies and the topology of the large-scale structure. We discuss a technique to estimate the statistical significance of filamentary structures in the distribution of galaxies. Extreme value statistical theory is applied to the problem of the nature of the brightest galaxies in rich clusters.


Table of Contents

Next