Skip to main content

Table 16 Homogeneity results for different potential risk of bias variables on four outcome types

From: The effectiveness of community engagement in public health interventions for disadvantaged groups: a meta-analysis

Outcome

Risk of bias variable

 

Model homogeneity

k

Between groups Q

Within groups Q

(p-value)

(p-value)

Health behaviours

Comparison group type

7

7.71 (p = .26)

97.14 (p = .51)

Random allocation

3

.14 (p = .93)

95.60 (p = .63)

Overall low risk of bias

2

1.27 (p = .26)

97.45 (p = .64)

Health consequences

Comparison group type

-

Insufficient data

Random allocation

-

Insufficient data

Overall low risk of bias

2

.18 (p = .67)

36.66 (p = .44)

Participant self-efficacy

Comparison group type

-

Insufficient data

Random allocation

-

Insufficient data

Overall low risk of bias

2

1.68 (p = .19)

20.33 (p = .31)

Social support

Comparison group type

-

Insufficient data

Random allocation

-

Insufficient data

Overall low risk of bias

2

.04 (p = .85)

7.19 (p = .21)

  1. Note. k = number of categories in the moderator variable; Between groups Q indicates the extent to which the categories of studies differ from each other; and within groups Q indicates the extent to which the effect size estimates within a category differ from each other.