Skip to main content

Table 3 Combined effect sizes and moderator analysesa for intervention effects

From: A gloomy picture: a meta-analysis of randomized controlled trials reveals disappointing effectiveness of programs aiming at preventing child maltreatment

 

K b

N

d

95 % CI

Q homogeneity

Q c,d contrast

Total effects on maltreatment

27

4883

0.13**

0.05 ~ 0.21

56.06**

 

Intervention characteristics

      

Focus

      

 Support

13

3742

0.03

−0.04 ~ 0.11

13.28

15.85**

 Parent training

4

421

0.37**

0.15 ~ 0.59

12.45**

 

 Parent training and support

6

362

0.37**

0.16 ~ 0.59

5.21

 

 Other

4

358

0.17

−0.09 ~ 0.44

2.39

 

Location of delivery

     

1.97

 At participants’ home

17

3731

0.10

−0.00 ~ 0.19

15.19

 

 In center

6

628

0.26*

0.05 ~ 0.46

32.16***

 

 Both

3

462

0.10

−0.15 ~ 0.36

4.47

 

Delivery format

     

0.16

 Individual

22

4133

0.14**

0.05 ~ 0.23

30.19

 

 Individual and group

5

750

0.09

−0.11 ~ 0.29

24.51***

 

Duration

     

6.04*

 < 6 months

5

452

0.22

−0.01 ~ 0.45

4.99

 

 6-12 months

11

1204

0.23**

0.10 ~ 0.37

25.01**

 

 > 12 months

10

3122

0.04

−0.06 ~ 0.13

12.55

 

Sessions

     

9.65**

 < 16

4

845

0.05

−0.12 ~ 0.23

14.57**

 

 16-30

11

726

0.37***

0.19 ~ 0.56

15.47

 

 > 30

7

1998

0.03

−0.10 ~ 0.16

3.51

 

Sample characteristics

      

Country of origin

     

0.66

 USA

23

 

0.11*

0.02 ~ 0.20

45.46**

 

 Other

4

 

0.19*

0.01 ~ 0.37

5.30

 

Type of sample

     

9.31**

 At risk

20

4236

0.05

−0.02 ~ 0.12

21.85

 

 Maltreating

6

423

0.35***

0.17 ~ 0.53

13.18*

 

 General population

1

224

0.44**

0.13 ~ 0.75

  

Design characteristics

      

Intent to treat

     

1.86

 Yes

20

3349

0.17**

0.07 ~ 0.27

37.19**

 

 No

7

1534

0.05

−0.09 ~ 0.19

16.23*

 

Blind assessment

     

2.33

 No

18

3369

0.19**

0.08 ~ 0.29

42.50**

 

 Yes

9

1514

0.06

−0.07 ~ 0.18

12.33

 

Pre-test

     

1.55

 No

18

3945

0.09*

0.01 ~ 0.18

38.87**

 

 Yes

9

938

0.20**

0.06 ~ 0.35

9.99

 

Control condition

     

1.85

 Other

16

2661

0.18**

0.05 ~ 0.30

30.74*

 

 Treatment as usual

7

1011

0.05

−0.12 ~ 0.21

11.36

 

Type of measure

     

3.08

 Self-report

9

1085

0.31**

0.12 ~ 0.49

6.13

 

 Other report

13

1764

0.11

−0.02 ~ 0.24

30.36**

 

Background characteristics

      

Type of publication

      

 Journal article

25

4603

0.14**

0.06 ~ 0.23

54.53***

 

 Dissertation

2

280

−0.10

−0.42 ~ 0.23

0.08

 
  1. k number of study outcomes, N total sample size, d effect size (Cohen’s d), 95 % CI 95 % confidence interval around the point estimate of the effect size, Qhomogeneity homogeneity statistic, Qcontrast moderation statistic
  2. *p < .05, **p < .01, ***p < .001
  3. aOnly categorical moderators are included in the table; the effects of continuous moderators are described in the text
  4. bMissings were excluded from moderator analyses. Therefore, sample sizes range from 22 to 27
  5. cSubgroups with k < 4 excluded from contrast
  6. dAfter controlling for year of publication and sample size, only type of intervention remained a significant moderator
Back to article page

BMC Public Health

ISSN: 1471-2458

Contact us