Skip to main content

Table 4 Factors associated with the heterogeneity of drinking rate (last 30 days) estimates among Chinese middle and high school students using meta-regression for four outcomes reported in 10 or more studies from 2007–2015

From: Estimate of adolescent alcohol use in China: a meta-analysis

Outcome variables

Moderator

Bivariate b

R 2%

Hieratical b

R 2%

Middle school students’ drinking rate

 Male

Location (north/south)

.368*

15

  

 Female

Location (east to middle)

.741**

28

.589*

50

Location (west to middle)

.776*

 

.804*

 

Collector (Yes)

.790**

28

.727**

 

High school students’ drinking rate

 Male

Location (east to middle)

.422

49

-.106

80

Location (west to middle)

1.29**

 

1.05***

 

Definition (Yes)

-.675*

28

  

Collector (Yes)

.729*

19

1.117***

 

 Female

Location (east to middle)

.779*

34

.474

80

Location (west to middle)

.805*

 

.867**

 

Collector (Yes)

1.043*

33

.959**

 

Definition (Yes)

-.794*

35

-.588*

 
  1. N = 20 for middle school males and females, N = 12 for high school males and females, N = 6 for vocational high school males and females. ***p < .001, **p < .01, *p < .1. The moderator of authorship for high school female students was significantly associated with drinking rate, it also has a strong correlation (r = −.775, p = .003) with the moderator of trained data collector. To deal with the collinearity issue, the moderator of authorship was removed in the hieratical meta-regression analysis. Bonferroni correction applied on the significance level of .1: the significant level changed to .05 for 2 moderators, .033 for 3 moderators; Bonferroni correction applied on the significant level of .01, the significant level changed to .005 for 2 moderators, .003 for 3 moderators; Bonferroni correction applied on the significant level of .001, the significant level changed to .0005 for 2 moderators, and .0003 for 3 moderators