¹⁷Administration of four validity tests – content, retest, criterion and construct validity – confirmed that the pride question was sufficiently robust to be modelled. The urban pride question produced similar distributions when it was asked in in 2004, 2006, 2010 and again in 2012.