TABLE 4.

Mean weighted effect sizes ($\overline{\mathrm{ß}}$) for epistemic emotions in Model 1 across three studies.

	Surprise				Curiosity				Confusion
Predictor	$\overline{\mathrm{ß}}$	p	95% CI	τ2	$\overline{\mathrm{ß}}$	p	95% CI	τ2	$\overline{\mathrm{ß}}$	p	95% CI	τ2
Accuracy	−0.448	<0.001	[−0.518; −0.377]	0.003	−0.424	<0.001	[−0.449; −0.398]	0.000	−0.467	<0.001	[−0.502; −0.433]	0.001
Confidence	0.142	<0.001	[0.094; 0.191]	0.002	−0.043	0.112	[−0.096; 0.010]	0.002	0.240	<0.001	[0.201; 0.280]	0.001
Accuracy × Confidence	−0.515	<0.001	[−0.536; −0.494]	0.000	−0.340	<0.001	[−0.381; −0.298]	0.001	−0.337	<0.001	[−0.357; −0.317]	0.000
Confidence in incorrect answers	0.651	<0.001	[0.621; 0.681]	0.000	0.365	<0.001	[0.315; 0.416]	0.001	0.508	<0.001	[0.436; 0.580]	0.003
Confidence in correct answers	−0.514	<0.001	[−0.550; −0.479]	0.000	−0.409	<0.001	[−0.446; −0.372]	0.000	−0.185	<0.001	[−0.236; −0.134]	0.001

$\overline{\mathrm{ß}}$, mean-weighted standardized path coefficient; p, p-value; CI, confidence interval; τ², between-study heterogeneity of effect sizes.