Table 2 Homogeneity measure, mean effect size and 95% CI for the comparable cognitive parameters
Cognitive parameter/studiesHomogeneity (χ2)Effect size† (Δ̄)95% CI
Simple Reaction Task/4 8134.84−0.05−0.25 to 0.15
Choice Reaction Task‡/1 4 8134.27−0.03−0.20 to 0.14
Vigilance§/4 10 11 136.71*0.01−0.22 to 0.23
Subtraction
    Correct/4 8 10 110.51−0.05−0.18 to 0.07
    Time/4 8 10 11 130.40−0.09−0.16 to −0.02
Sentence Verification/4 10 11
    Correct0.490.01−0.12 to 0.14
    Time0.460.01−0.12 to 0.13
N-back test/5 10 11
    Targets
    0-back
        Time0.01−0.04−0.06 to −0.02
        Errors00NA
    1-back
        Time0.220.02−0.06 to 0.11
        Errors1.700.01−0.23 to 0.26
    2-back
        Time2.540.100.098 to 0.99
        Errors0.450.01−0.12 to 0.13
    3-back
        Time0.59−0.04−0.18 to 0.11
        Errors0.600.07−0.08–0.21
    Non-targets
    0-back
        Time0.280.003−0.09 to 0.10
        Errors3.13−0.14−0.46 to 0.18
    1-back
        Time0.250.002−0.09 to 0.10
        Errors0.19−0.04−0.12 to 0.05
    2-back
        Time0.110.04−0.02 to 0.10
        Errors0.0090.050.03 to 0.07
    3-back
        Time0.020.0004−0.03 to 0.03
        Errors0.020.01−0.01 to 0.04
Trail Making Test
    TMT A/1 90.41−0.04−0.13 to 0.05
    TMT B/1 90.540.05−0.05 to 0.15
Digit span forward§/6 9 1210.95**0.08−0.27 to 0.43
Digit span backward/9 120.42−0.09−0.21 to 0.03
Spatial span forward/6 9 123.380.29−0.05 to 0.63
Spatial span backward/6 123.240.15−0.21 to 0.51
  • *p⩽0.05; **p⩽0.01.

  • †A positive effect size indicates higher values under exposure; a negative effect size indicates lower values under exposure.

  • ‡The studies of Besset et al12 and Koivisto et al4 used two forms measuring choice reaction time. In both cases the average delta is used as the reference effect size.

  • §The assumption of a common effect size is violated, computations follow a random effects model.