Sources of potential bias when combining routine data linkage and a national survey of secondary school-aged children: a record linkage study

Table 3 Modelled student and school level characteristics associated with linkage success

	n (%)	Linked (%)	OR (95% CIs)	Wald test	AOR (95% CIs)
Student-level
Gender
Males	4654 (50.4)	3315 (51.5)	1.00		1.00
Female	4465 (48.4)	3098 (48.1)	0.91 [0.82, 1.00]		0.94 [0.85, 1.04]
Did not want to answer	113 (1.2)	28 (0.4)	0.11 [0.07, 0.18]	X²(2) = 31.55	0.19 [0.12, 0.32]	X²(2) = 40.33
School grade
Year 7	1980 (21.4)	1538 (23.9)	1.00		1.00
Year 8	2125 (23.0)	1608 (25.0)	0.89 [0.76, 1.04]		0.89 [0.75, 1.05]
Year 9	1969 (21.3)	1360 (21.1)	0.63 [0.54, 0.73]		0.62 [0.53, 0.73]
Year 10	1696 (18.4)	1023 (15.9)	0.43 [0.37, 0.5]		0.43 [0.36, 0.50]
Year 11	1462 (15.8)	912 (14.2)	0.41 [0.35, 0.48]	X²(4) = 212.34	0.39 [0.33, 0.47]	X²(4) = 189.57
SWEMWBS (sd.)	22.449 (4.24)	22.651 (4.19)	1.04 [1.03, 1.05]		1.03 [1.01, 1.04]
FAS (sd.)	9.484 (2.27)	9.504 (2.24)	1.00 [0.98, 1.03]		1.00 [0.97, 1.02]
School-level
Mean FAS (sd.)	9.484 (0.51)	9.499 (0.52)	1.11 [0.75, 1.63]		1.15 [0.76, 1.76]
ICC – constant only	0.119
ICC – Student level	0.128
ICC – Level 1 & 2 variables	0.126

n = 9232 (adjusted model n = 8401); schools = 39; observations per school: min = 41; max = 817; P < 0.05 Intraclass correlation coefficient (ICC)

ISSN: 1471-2288