Clustering of loose groups and galaxies in the Perseus-Pisces survey

We investigate the clustering properties of loose groups in the Perseus-Pisces redshift survey. Previous analyses based on CfA and SSRS surveys led to apparently contradictory results. We investigate the source of such discrepancies, finding satisfactory explanations for them. Furthermore, we find a definite signal of group clustering, whose amplitude A(G) exceeds the amplitude A(S) of galaxy clustering (A(G) = 14.5(-3.0)(+3.8) A(g) = 7.42(0.19)(+0.20) for the most significant case; distances are measured in h(-1) Mpc). Groups are identified with the adaptive friends-of-friends (FOF) algorithms HG (Huchra & Geller) and NW (Nolthenius & White). We debate the dependence of clustering parameters on group individuation recipes, systematically varying all search parameters. Correlation strength is especially sensitive to the sky link D-L (increasing for stricter normalization D-0) and to the (depth m(lim) of the) galaxy data. It is only moderately dependent on the galaxy luminosity function phi(L), while it is almost insensitive to the redshift link V-L(both to the normalization V-0 and to the scaling recipes HG or NW).