Table 1
Pearson correlations coefficients between past numbers of missed school days by motives and drug use indicators in respondents (n=6007)
|
Skipped days (truancy only)
|
Legitimate absence only
|
Missed days (absence and truancy)
|
Daily tobacco use (0-1)
|
Regular alcohol use (0-1)
|
Regular cannabis use (0-1)
|
Girls
|
|
|
|
|
|
|
Skipped days (truancy only)
|
1.00
|
0.30
|
0.65
|
0.20
|
0.03
|
0.19
|
Legitimate absences only
|
0.30
|
1.00
|
0.92
|
0.11
|
0.06
|
0.07
|
Missed days (absence or truancy)
|
0.65
|
0.92
|
1.00
|
0.17
|
0.06
|
0.14
|
Boys
|
|
|
|
|
|
|
Skipped days (truancy only)
|
1.00
|
0.27
|
0.68
|
0.24
|
0.10
|
0.21
|
Legitimate absences only
|
0.27
|
1.00
|
0.89
|
0.08
|
-0.02
|
0.06
|
Missed days (absence or truancy)
|
0.68
|
0.89
|
1.00
|
0.17
|
0.03
|
0.14
|
Data are weighted by the sampling weights.
Table 2
Drug use levels in respondents according to their reports of past absences
|
Daily tobacco use (%)
|
Regular alcohol use (%)
|
Regular cannabis use (%)
|
Girls
|
23.5
|
8.9
|
5.0
|
No absence
|
15.7
|
6.4
|
2.3
|
Any type of absence
|
26.9
|
9.9
|
6.2
|
Legitimate absences only
|
19.1
|
9.3
|
3.3
|
Truancy
|
41.0
|
11.1
|
11.6
|
Boys
|
20.6
|
18.9
|
10.4
|
No absence
|
14.7
|
16.9
|
6.6
|
Any type of absence
|
24.0
|
20.1
|
12.6
|
Legitimate absences only
|
16.6
|
17.5
|
7.5
|
Truancy
|
33.8
|
23.6
|
19.3
|
Total
|
22.1
|
13.7
|
7.6
|
No absence
|
15.2
|
11.9
|
4.6
|
Any type of absence
|
25.6
|
14.6
|
9.1
|
Legitimate absences only
|
18.0
|
12.8
|
5.1
|
Truancy
|
37.3
|
17.4
|
15.6
|
Data are weighted by the sampling weights.
Table 3
Levels of drug use and correlations between truancy and drug use by stratum (n=6,007)
|
Stratum
|
|
Missingness pattern at data collection
|
Level of use in respondents(a)
|
Pearson correlation coefficients with reported truancy(a)
|
N
|
Educational field
|
Sector
|
Missing pupils
|
Discarded pupils
|
Truancy in missing
|
Daily tobacco use
|
Regular alcohol use
|
Regular cannabis use
|
Daily tobacco use
|
Regular alcohol use
|
Regular cannabis use
|
520
|
1st year Highschool
|
Private
|
6.2%
|
1.7%
|
15.6%
|
13.8%
|
10.6%
|
3.5%
|
0.13
|
0.04
|
0.23
|
1887
|
|
Public
|
10.6%
|
1.1%
|
19.5%
|
13.7%
|
7.0%
|
4.9%
|
0.18
|
0.03
|
0.19
|
368
|
2d year Highschool
|
Private
|
8.4%
|
2.7%
|
19.4%
|
18.8%
|
19.2%
|
5.5%
|
0.10
|
0.17
|
0.04
|
996
|
|
Public
|
13.3%
|
1.8%
|
27.8%
|
17.9%
|
12.4%
|
8.5%
|
0.29
|
0.06
|
0.21
|
182
|
3rd year Highschool
|
Private
|
10.4%
|
3.8%
|
26.3%
|
29.2%
|
24.3%
|
6.4%
|
0.13
|
-0.01
|
0.12
|
810
|
|
Public
|
14.9%
|
1.4%
|
42.1%
|
22.6%
|
20.6%
|
8.7%
|
0.26
|
0.09
|
0.25
|
159
|
1st year short vocational
|
All
|
21.3%
|
3.8%
|
17.6%
|
34.5%
|
15.9%
|
3.9%
|
0.31
|
0.08
|
0.33
|
125
|
2d year short vocational
|
All
|
22.4%
|
1.6%
|
82.1%
|
38.3%
|
14.4%
|
9.6%
|
0.13
|
-0.18
|
0.11
|
158
|
1st year long vocational
|
Private
|
9.4%
|
1.9%
|
26.7%
|
36.7%
|
17.5%
|
11.8%
|
0.27
|
-0.01
|
0.35
|
912
|
|
Public
|
16.4%
|
4.6%
|
54.0%
|
27.7%
|
11.4%
|
8.3%
|
0.14
|
0.16
|
0.23
|
112
|
2d year long vocational
|
Private
|
7.1%
|
3.6%
|
0.0%
|
36.5%
|
7.5%
|
9.2%
|
0.15
|
0.01
|
0.01
|
405
|
|
Public
|
16.0%
|
5.2%
|
56.9%
|
26.3%
|
11.9%
|
10.3%
|
0.21
|
0.07
|
0.09
|
532
|
3rd year long vocational
|
All
|
21.6%
|
4.5%
|
53.0%
|
40.4%
|
20.2%
|
13.5%
|
0.20
|
0.11
|
0.19
|
(a) Weighted by the classic direct calibration
Table 4
Statistics of the different weighting schemes
|
TNRC(a) weights before calibration
|
Final weights
|
|
Without sampling weight
|
With sampling weight
|
TNRC followed by calibration
|
Method: normalisation level
|
Sum
|
Min
|
P99
|
Max
|
CV
|
Sum
|
CV
|
Min
|
P99
|
Max
|
Sum
|
CV
|
Sampling weight(b)
|
6007.0
|
0.5
|
2.0
|
2.0
|
68.7
|
-
|
-
|
0.4
|
4.4
|
5.6
|
7166.0
|
76.1
|
Method 1: original(c)
|
7998.3
|
1.0
|
5.7
|
20.0
|
111.1
|
7726.3
|
118.9
|
0.3
|
5.9
|
36.2
|
7166.0
|
118.9
|
Trimmed(d)
|
7483.2
|
1.0
|
5.7
|
5.7
|
47.9
|
7364.7
|
82.4
|
0.3
|
4.7
|
15.9
|
7166.0
|
89.7
|
Method 2: strata
|
7166.0
|
1.0
|
2.4
|
5.7
|
24.4
|
7177.4
|
75.2
|
0.3
|
4.4
|
10.8
|
7166.0
|
80.9
|
Method 3: strata
|
7166.0
|
1.0
|
2.4
|
6.8
|
25.9
|
7182.4
|
76.2
|
0.3
|
4.4
|
12.1
|
7166.0
|
81.1
|
CV: coefficient of variation; P99: 99th percentile.
(a): Total non-response correction
(b): No TNRC was computed: sampling weight is the benchmark.
(c): There is no normalisation in the original method
(d): The weights were trimmed at the 99th percentile computed in the original method.
Note: the real number of respondents is 6007 while the real number of pupils is 7166.
Table 5
Estimates of key drug use indicators according to various weighting schemes at the national level before calibration
Method
|
No treatment(a)
|
Method 1
|
Method 2
|
Method 3
|
Level(b)
|
Sample
|
Sample
|
Sample
|
Strata
|
Strata
|
Truncation
|
No
|
No
|
Yes
|
No
|
No
|
|
|
%
|
std
|
%
|
std
|
%
|
std
|
%
|
std
|
%
|
std
|
Tobacco
|
Lifetime
|
59.8
|
1.10
|
61.6
|
1.25
|
60.4
|
1.08
|
61.2
|
1.12
|
61.3
|
1.12
|
|
Daily
|
22.1
|
1.04
|
24.8
|
1.33
|
24.0
|
0.92
|
23.8
|
1.14
|
23.9
|
1.15
|
Alcohol
|
Lifetime
|
85.9
|
0.98
|
84.5
|
1.18
|
83.4
|
1.09
|
85.4
|
0.99
|
85.5
|
0.99
|
|
Month
|
62.2
|
1.40
|
61.0
|
1.63
|
59.6
|
1.32
|
62.2
|
1.43
|
62.2
|
1.44
|
|
Regular
|
13.7
|
0.80
|
14.5
|
0.90
|
13.3
|
0.60
|
14.3
|
0.85
|
14.4
|
0.85
|
Cannabis
|
Lifetime
|
43.3
|
1.15
|
45.4
|
1.33
|
43.3
|
1.07
|
45.2
|
1.20
|
45.3
|
1.20
|
|
Month
|
22.1
|
0.98
|
24.1
|
1.16
|
23.2
|
0.93
|
23.7
|
1.10
|
23.7
|
1.10
|
|
Regular
|
7.6
|
0.63
|
9.5
|
0.86
|
8.8
|
0.61
|
8.6
|
0.74
|
8.7
|
0.74
|
(a): sampling weight only.
(b): the level within which the total non-response correction is done: sample, or strata within sample.
%=point estimate; std=standard error;
Table 6
Estimates of key drug use indicators according to various weighting schemes at the national level after calibration
Method
|
No treatment(a)
|
Method 1
|
Method 2
|
Method 3
|
Level(b)
|
Sample
|
Sample
|
Sample
|
Strata
|
Strata
|
Truncation
|
No
|
No
|
Yes
|
No
|
No
|
|
|
%
|
std
|
%
|
std
|
%
|
std
|
%
|
std
|
%
|
std
|
Tobacco
|
Lifetime
|
60.4
|
1.15
|
62.4
|
1.29
|
62.1
|
1.19
|
61.6
|
1.18
|
61.6
|
1.18
|
|
Daily
|
22.6
|
1.17
|
25.2
|
1.42
|
24.6
|
1.32
|
24.0
|
1.27
|
24.1
|
1.28
|
Alcohol
|
Lifetime
|
86.3
|
1.00
|
85.7
|
1.12
|
85.8
|
1.05
|
86.0
|
1.02
|
86.0
|
1.02
|
|
Month
|
63.3
|
1.50
|
63.1
|
1.67
|
63.2
|
1.58
|
63.5
|
1.52
|
63.4
|
1.53
|
|
Regular
|
14.2
|
1.01
|
15.1
|
1.11
|
14.6
|
1.06
|
14.6
|
1.09
|
14.6
|
1.09
|
Cannabis
|
Lifetime
|
44.0
|
1.19
|
46.5
|
1.36
|
46.1
|
1.27
|
45.5
|
1.23
|
45.6
|
1.23
|
|
Month
|
22.6
|
1.03
|
24.6
|
1.19
|
24.4
|
1.12
|
23.9
|
1.09
|
23.9
|
1.08
|
|
Regular
|
7.7
|
0.67
|
9.4
|
0.86
|
9.0
|
0.78
|
8.5
|
0.76
|
8.6
|
0.75
|
(a): classical weighting scheme used as benchmark: direct calibration of the sampling-weighted respondent sample
(b): the level within which the total non-response correction is done: sample, or strata within sample.
%=point estimate; std=standard error;