A bootstrap resampling analysis of galaxy clustering

Abstract
We demonstrate how statistics and standard errors can be associated with parameters of single data sets using bootstrap resampling methods. These techniques allow significance levels to be associated with any parameter derived from a data set and measure the robustness of the data set to various known and unknown errors and biases. We apply the method to the two-point angular correlation function of the Zwicky 14 mag catalogue of galaxies. If the two-point function has the power-law form $$\omega(\theta) = (\theta_0/\theta)^\gamma$$ then standard errors of $$\sigma(\theta_0) = 0.01$$ and $$\sigma(\gamma) = 0.13$$ are found with means of $$\overline \gamma = 0.80$$ and $$\overline \theta_0 = 0.06$$ radians. These are much larger than the formal errors quoted in previous analyses. Various consequences of these results and further applications of the method employed are discussed.

This publication has 0 references indexed in Scilit: