Skip to main content
. 2013 Aug 18;17(8):1776–1785. doi: 10.1017/S1368980013002127

Table 3.

Multilevel linear models for eating behaviour outcomes (REML estimates*; se) among 15-year-old adolescents; data from the 2010 Scotland Health Behaviour in School-aged Children survey

Fruit consumption Vegetable consumption Sweets consumption Crisps consumption Chips consumption Healthy eating score
Mean se Mean se Mean se Mean se Mean se Mean se
Fixed effects
Constant/intercept 5·46 2·32 1·83 2·23 5·40 2·12 7·15 2·39 1·33 1·87 −10·23 11·05
Age −0·08 0·15 0·18 0·14 −0·04 0·14 −0·24 0·15 0·06 0·12 0·61 0·71
Sex (ref.: Male)
Female 0·42 0·09‡ 0·31 0·09‡ −0·33 0·09‡ −0·04 0·10 −0·46 0·08‡ 2·92 0·45‡
FAS (ref.: Low FAS)
Medium FAS 0·11 0·12 0·26 0·11‡ −0·07 0·11 −0·16 0·12 −0·33 0·09‡ 1·19 0·55‡
High FAS 0·23 0·12 0·24 0·12‡ 0·04 0·11 −0·14 0·12 −0·19 0·10‡ 1·18 0·57‡
Deprivation (SIMD score/10) −0·20 0·04‡ −0·26 0·04‡ 0·003 0·04 0·11 0·04‡ 0·21 0·03‡ −1·19 0·19‡
School type (ref.: State school)
Independent school 0·94 0·29‡ 0·99 0·26‡ −0·12 0·29 −0·81 0·37‡ −0·77 0·30‡ 5·78 1·53‡
Rurality (ref.: Four Cities)
Other Urban −0·20 0·16 −0·01 0·15 −0·05 0·15 0·20 0·18 0·47 0·14‡ −1·73 0·81‡
Accessible Towns −0·42 0·20‡ −0·14 0·18 −0·20 0·18 −0·13 0·21 0·11 0·16 −0·90 0·98
Remote Towns −0·13 0·21 0·16 0·19 −0·32 0·19 −0·12 0·23 0·12 0·18 0·22 1·06
Accessible Rural 0·08 0·17 0·31 0·16 −0·41 0·16‡ −0·22 0·18 0·11 0·14 1·02 0·85
Remote Rural 0·13 0·17 0·55 0·16‡ −0·65 0·16‡ −0·67 0·19‡ −0·08 0·15 2·90 0·88‡
Random effects
Level 1 (child) variance 5·645 0·160 5·266 0·149 4·759 0·135 5·979 0·169 3·662 0·103 123·100 3·530
Level 2 (school) variance 0·173 0·058 0·091 0·046 0·046 0·041 0·084 0·055 0·042 0·033 4·836 1·547
Level 3 (education authority) variance 0·000 0·000 0·000 0·000 0·026 0·024 0·058 0·039 0·042 0·025 0·293 0·687
−2 log likelihood 12 200·8 11 955·1 11 693·6 12 272·0 11 030·4 19 885·5
No. of parameters 15 15 15 15 15 15
AIC† 12 230·8 11 985·1 11 723·6 12 302·0 11 060·4 19 915·5
AIC† for model without rurality 12 231·7 11 996·3 11 733·5 12 315·3 11 068·9 19 934·9

REML, restricted maximum likelihood; ref., referent category; FAS, Family Affluence Scale; SIMD, 2006 Scottish Index of Multiple Deprivation.

*Via restricted iterative generalised least squares (RIGLS).

†AIC is the Akaike Information Criterion; the larger this is, the worse the model fit.

‡95 % CI is above or below 0.