Table 3

Area and individual level effects on fruit purchasing (random intercept models)*

Areas = 50
 Individuals = 970Model 1 (null model)Model 2 (plus age, sex, and household income)Model 3 (plus area disadvantage)
*The fruit purchasing index ranged from 0–100, with higher scores indicating a wider variety and greater regularity of fruit purchase. †Expressed as a two standard deviation effect size. ‡The proportion of the total variance in fruit purchasing behaviour that is between the census collectors districts (small urban areas).
Constant50.4640.4329.55
Fixed effects Est95%CIEst95% CI
Age of main food purchaser0.340.27 to 0.410.330.26 to 0.40
Sex of main food purchaser (male)−10.74−13.25 to −8.23−10.69−13.20 to −8.18
Household income
$52000 or more
$36400–51999−0.68−3.62 to 2.24−0.44−3.39 to 2.50
$20800–36399−5.76−8.55 to −2.98−5.29−8.14 to −2.46
⩽$20799−8.60−11.84 to −5.36−7.85−11.23 to −4.48
Area socioeconomic disadvantage†2.01−0.49 to 4.50
Random effects variance Est.SEEst.SEEst.SE
Level 2 (areas)12.945.815.583.854.693.65
Level 1 (individual)304.0314.16262.7912.24262.7912.23
Deviance8328.028174.048171.61
Intraclass correlation (%)‡4.082.081.74