Table 2.

p-values of global Moran's I tests for spatial autocorrelation of the proportions $p_j^{\{\mathrm{day},\mathrm{week},\mathrm{month}\}}$ and the log-transformed reporting delay log(Δt_i) for LA and NY. To create the spatial neighbourhood matrix required for computing Moran's I, we use k-nearest-neighbours with k = 50 for NY and all observations in the same police district as neighbours in LA. Significance codes: 0.05*, 0.01**, 0.001***.

LA				NY
$p_j^{\mathrm{day}}$	$p_j^{\mathrm{week}}$	$p_j^{\mathrm{month}}$	log(Δti)	$p_j^{\mathrm{day}}$	$p_j^{\mathrm{week}}$	$p_j^{\mathrm{month}}$	log(Δti)
0.067	0.009**	0.002**	0.000***	0.013*	0.01**	0.006**	0.000***