Table 3.

Moderator analyses on young children's touchscreen learning.

Variables	N	k	Effect size					Homogeneity test between subgroups
			Cohen's d	p	95% CI	z	80% CV	QB (df)	p
Learning material domain								8.23(1)	0.004
STEM	1,907	37	0.65***	< 0.001	[0.47, 0.82]	7.20	[−0.05, 1.35]
Non-STEM	2,299	42	0.31***	< 0.001	[0.16, 0.46]	4.04	[−0.32, 0.94]
Comparison group								39.47(5)	< 0.001
Baseline	1,596	34	0.77***	< 0.001	[0.62, 0.93]	9.99	[0.19, 1.36]
Traditional classroom teaching	2,215	12	0.46***	< 0.001	[0.20, 0.71]	3.53	[0.01, 0.90]
Mouse-based computers	256	1	0.34**	0.006	[0.10, 0.59]	2.73	[0.18, 0.51]
Paper	658	13	0.11	0.537	[−0.24, 0.46]	0.62	[−0.71, 0.93]
Physical objects	99	3	0.10	0.600	[−0.27, 0.46]	0.52	[−0.31, 0.51]
Watch on touchscreen	754	16	0.07	0.502	[−0.13, 0.27]	0.67	[−0.45, 0.59]
Test media								6.62(3)	0.085
Touchscreen devices	977	13	0.73***	< 0.001	[0.43, 1.04]	4.69	[0.01, 1.45]
Paper	2,567	39	0.48***	< 0.001	[0.34, 0.62]	6.77	[−0.09, 1.05]
Physical objects	339	15	0.41*	0.015	[0.08, 0.73]	2.42	[−0.43, 1.24]
Oral	368	9	0.06	0.765	[−0.35, 0.48]	0.30	[−0.75, 0.88]
Experimental environment								10.24(2)	0.006
Classroom	2,659	27	0.55***	< 0.001	[0.38, 0.73]	6.12	[−0.05, 1.15]
Laboratory	675	21	0.20*	0.027	[0.02, 0.37]	2.21	[−0.33, 0.73]
Other	872	31	0.55***	< 0.001	[0.37, 0.74]	5.88	[−0.12, 1.22]

^*p < 0.05;

^**p < 0.01;

^***p < 0.001.