Are spatial regression methods a panacea or a Pandora's box? A reply to Beale et al. (2007)