Table 7.
Characteristics of studies included in meta‐analysis
|
Study characteristics |
|
|
|---|---|---|
|
Publication year (range) |
1980 ‐2016 |
|
|
|
|
|
|
Type of publication |
n |
% |
|
Journal articles (published) |
19 |
51.3 |
|
PhD thesis, technical report (unpublished) |
18 |
48.6 |
|
|
|
|
|
Language |
n |
% |
|
English |
37 |
100 |
|
|
|
|
|
Country of the sample |
n |
% |
|
United States |
33 |
89.1 |
|
United Kingdom |
3 |
8.1 |
|
Unclear |
1 |
2.7 |
|
|
|
|
|
Author's main discipline |
n |
% |
|
Education |
13 |
35.3 |
|
Social work |
1 |
2.7 |
|
Psychology |
12 |
32.3 |
|
Criminal justice |
1 |
2.7 |
|
Psychiatry/medicine |
6 |
16.2 |
|
Econometrist‐economics |
2 |
5.4 |
|
Not reported |
2 |
5.4 |
|
|
|
|
|
Declared conflict of Interest in published studies (N=19) |
n |
% |
|
Yes |
7 |
36.8 |
|
No |
12 |
63.2 |
|
|
|
|
|
Conflict of financial Interest |
n |
% |
|
Unlikely |
18 |
48.6 |
|
Possibly |
13 |
35.1 |
|
Likely |
2 |
5.4 |
|
Not enough information |
4 |
10.8 |
|
|
|
|
|
Unit of randomisation |
n |
% |
|
Individuals |
26 |
70.3 |
|
School clusters |
8 |
21.6 |
|
Classroom clusters |
3 |
8.1 |
|
|
|
|
|
Statistical analysis |
n |
% |
|
Multilevel modelling |
4 |
10.8 |
|
Differences in means (MANOVA, X2 , ANOVA, ANCOVA) |
22 |
59.4 |
|
Regression |
8 |
21.6 |
|
Frequencies |
3 |
8.1 |
|
|
|
|
|
Exclusion measurement |
n |
% |
|
Self‐report |
2 |
5.4 |
|
Teacher‐report |
3 |
8.1 |
|
School records (official records) |
30 |
81.1 |
|
Unknown |
2 |
5.4 |
|
|
|
|
|
Effective28 sample size |
n |
% |
|
< 300 |
24 |
64.8 |
|
Between 300 and 800 |
7 |
18.9 |
|
> 900 |
6 |
16.2 |