Table 2.

A summary of obtained effect sizes for differences between blind and sighted individuals in threshold, discrimination, cued identification and free identification

Estimate type	Effect size (Hedges g)	95% CI	p
Threshold (k = 18, N = 1227, Nblind = 590, Nsighted = 637)
Uncorrected estimates	0.107	− 0.218 to 0.433	0.43
Publication bias test
Egger’s test	z = 0.717, p = 0.47
Rank-correlation test	τ = 0.085, p = 0.65
Discrimination (k = 14, N = 940, Nblind = 455, Nsighted = 485)
Uncorrected estimates	0.413	0.064 to 0.763	0.021
Publication bias test
Egger’s test	z = 3.282, p < 0.001
Rank-correlation test	τ = 0.407, p = 0.047
Publication-bias corrected estimates
Trim and fill	0.111	− 0.165 to 0.387	0.43
PET-PEESE	− 0.31	− 0.82 to 0.20	0.21
Cued identification (k = 14, N = 968, Nblind = 468, Nsighted = 500)
Uncorrected estimates	− 0.131	− 0.378 to 0.116	0.30
Publication bias test
Egger’s test	z = 1.83, p = 0.068
Rank-correlation test	τ = 0.187, p = 0.39
Free Identification (k = 7, N = 443, Nblind = 224, Nsighted = 219)
Uncorrected estimates	1.20	0.072 to 2.326	0.037
Publication bias test
Egger’s test	z = 7.54, p < 0.001
Rank-correlation test	τ = 0.91, p = 0.003
Publication-bias corrected estimates
Trim and fill	0.084	− 0.409 to 0.577	0.738
PET-PEESE	0.02	− 0.26 to 0.30	0.88

95% CI 95% confidence intervals, p p value, k the number of studies, N total number of participants, PET-PEESE precision-effect testing–precision-effect-estimate with standard error meta-analysis