Skip to main content

Table 2 Effects of moderator variables on effect size in the random-effect model

From: Effects of digital game-based STEM education on students’ learning achievement: a meta-analysis

Moderator variables

N

ES

SE

95% CI

Qb

Lower

Upper

Control treatment

3.506

 Traditional

22

0.558***

0.095

0.373

0.744

 Multimedia

14

0.848***

0.122

0.608

1.088

Subject discipline

2.188

 Science

19

0.750***

0.105

0.544

0.956

 Mathematics

14

0.629***

0.119

0.395

0.863

 Technology/engineering

3

0.367

0.249

-0.121

0.854

Education level

5.184

 Primary education

20

0.835***

0.106

0.628

1.042

 Secondary education

12

0.487***

0.127

0.238

0.736

 Higher education

4

0.492*

0.218

0.064

0.920

Game type

11.126**

 Immersive games

12

0.583***

0.125

0.339

0.828

 Tutorial games

20

0.593***

0.097

0.403

0.784

 Board games

4

1.455***

0.248

0.969

1.941

Gaming platform

0.752

 Computer

24

0.625***

0.091

0.447

0.802

 Mobile

12

0.768***

0.138

0.497

1.038

Intervention duration

5.070

  < 1 week

10

0.953***

0.154

0.651

1.255

 1 week–1 month

6

0.464*

0.190

0.091

0.837

 1 month–3 months

9

0.579***

0.157

0.272

0.886

  ≥ 3 months

7

0.597***

0.166

0.272

0.923

 Not specified

4

0.676**

0.230

0.224

1.127

  1. N, the number of effect sizes; ES, effect size; SE, standard error; Qb, Q value of the heterogeneity test between the subgroups; CI, confidence interval; *p < 0.05; **p < 0.01; ***p < 0.001