Skip to main content

Table 2 Multilevel intraclass correlations in secondary schools for fruit and vegetable consumption

From: School health promotion and fruit and vegetable consumption in secondary schools: a repeated cross-sectional multilevel study

 

Fruit¹ (N = 576²)

 

Vegetables¹ (N = 576²)

 

ICCschool-level (%)

ICC school year-level (%)

ICC school-level (%)

ICC school year-level (%)

0 model

2.87

0.78

5.57

0.77

School health promotion

Healthy School

2.90

0.69*

5.61

0.67*

Healthy School ever

2.88

0.78

5.58

0.77

Number of years Healthy School

2.94

0.61*

5.63

0.63*

Nutrition

2.94

0.71

5.60

0.70

General school characteristics

School size

2.58

0.83

5.11

0.83

Urbanicity school area

2.85

0.75

5.65

0.72

School type

2.83

0.78

5.56

0.77

School population characteristics

Poverty level

2.82

0.77

5.10

0.77

High parental educational attainment

0.63*

0.52*

1.84*

0.45*

Household income³

1.00*

0.79

2.96*

0.84

Migration background³

2.72

0.78

4.93*

0.79

Age3 4

2.27*

0.64*

5.11

0.73

Grade4

2.85

0.73

5.57

0.77

Educational track3 4;

0.99*

0.73

2.53*

0.66*

Self-rated general health4

2.46*

0.71

5.01*

0.63*

Psychosocial health4

2.54*

0.87

5.21

0.83

Bullied at school4

2.25*

0.57*

5.00*

0.46*

Cyberbullied4

2.38*

0.66*

4.95*

0.46*

Truancy4

2.88

0.79

5.60

0.78

Sickness4 5

3.22

0.67*

6.09

0.66*

School experience3 4

1.86*

0.80

4.34*

0.72

Urbanicity home area3 4

2.92

0.75

5.65

0.76

  1. Note All analyses were controlled for the season of administration and whether the survey was filled out anonymously or not. 1 Included as number of days per week consumed. 2 Number of school x school year combinations. N (schools) = 256, N (students) = 168,127. * Adding the characteristic to the 0-model reduced the ICC by ≥ 10%. 3 The school population characteristic consists of two variables. 4 The analysis was controlled for the individual characteristic. 5 Due to convergence warnings, we changed the optimizer method for the model estimation