Table 2 Risk of being a strengths and difficulties questionnaire case in the second phase, in relation to cultural identity in the first phase, stratified by ethnic group
Ethnic groupCultural identityUnadjustedAdjusted*
OR (95% CI)p ValueN, R2OR (95% CI)p ValueN, R2
Model 1
White BritishFriendship choiceIntegrated1
Traditional1.1 (0.5 to 2.4)0.831.0 (0.4 to 2.5)0.94
Assimilated1.1 (0.6 to 1.9)0.71N  =  3511.1 (0.4 to 2.7)0.89N  =  326
Marginalised0.7 (0.3 to 1.7)0.440.7 (0.2 to 1.9)0.44
Model 2
BangladeshiFriendship choiceIntegrated1
Traditional0.8 (0.4 to 1.5)0.470.9 (0.5 to 1.7)0.77
Assimilated1.9 (0.6 to 6.0)0.31N  =  4531.6 (0.5 to 5.6)0.43N  =  413
Marginalised1.2 (0.5 to 2.6)0.711.3 (0.5 to 3.8)0.58
Model 3
White BritishClothing choiceIntegrated11
Traditional2.7 (1.1 to 7.1)0.042.9 (1 to 7.9)0.04
Assimilated6.7 (2 to 22.1)0.002N  =  3515.5 (1.7 to 18.3)0.005N  =  325
Marginalised3.3 (1.5 to 7.1)0.0033.2 (1.5 to 6.9)0.002
Model 4
BangladeshiClothing choicesIntegrated11
Traditional0.3 (0.1 to 0.9)0.020.17 (0.06 to 0.5)0.007
Assimilated1.3 (0.8 to 2.3)0.26N  =  4541.0 (0.5 to 1.9)0.67N  =  412
Marginalised0.93 (0.5 to 1.8)0.840.82 (0.4 to 1.6)0.12
  • OR, Odds ratio.

  • *Adjustment for age (continuous), gender and year group interaction, eligibility for free school meals, car ownership, number of rooms in the house, religion, place of birth, phase 1 strengths and difficulties questionnaire score as a continuous variable, social support from friends score as a continuous variable. All analyses weighted for proportion of non-responders per cluster (using pw in Stata) and standard errors adjusted for cluster sampling using cluster in Stata.