Experimental correlation analysis: foundations and practice

Abstract

Commonalities and differences in correlation analysis in terms of phase space, conditioning and uncorrelatedness are discussed. The Poisson process is not generally appropriate as reference distribution for normalisation and cumulants, so that generalised statistics in terms of arbitrarily defined reference processes must be employed. Consideration of the sampling hierarchy leads us to a classification of current event-by-event observables.

Comments: Multiparticle Dynamics (Tihany) review talk. 12 pages, includes 4 figs, 8 .eps files and own style file