Table 4.

Predictors of over/underestimation in favor of H1, world regions

PredictorBehavioral, all (k = 42, n = 608)Biobehavioral (k = 20, n = 308)Behavioral (k = 22, n = 300)
(Intercept)−0.15 [−0.20, −0.08]−0.08 [−0.19, 0.03]−0.17 [−0.24, −0.07]
Asia0.13 [−0.03, 0.30]0.18 [0.00, 0.37]−0.03 [−0.32, 0.29]
EU150.09 [0.02, 0.16]0.09 [−0.05, 0.23]0.08 [0.00, 0.17]
Other0.22 [0.07, 0.33]0.02 [−0.15, 0.19]0.35 [0.13, 0.50]
Expectation0.08 [0.00, 0.14]0.05 [−0.06, 0.16]0.08 [−0.02, 0.18]
Asia*expectation−0.20 [−0.38, 0.01]−0.28 [−0.50, −0.06]0.01 [−0.34, 0.37]
EU15*expectation−0.12 [−0.20, −0.03]−0.16 [−0.32, −0.01]−0.09 [−0.19, 0.04]
Other*expectation−0.24 [−0.33, −0.05]−0.03 [−0.22, 0.16]−0.35 [−0.50, −0.09]
SE0.21 [0.07, 0.30]0.15 [0.00, 0.31]0.24 [−0.01, 0.38]
Pub. order0.01 [−0.01, 0.03]−0.02 [−0.06, 0.00]0.03 [0.00, 0.06]
  • Likelihood of a primary study within a meta-analysis to deviate from the summary effect size (deviation score, i.e., double square root-transformed absolute value of deviation values in Fig. 1), depending on the primary study’s size (SE), its chronological order of appearance (pub. order) within the meta-analysis (z-scaled by meta-analysis), the geographical origin of its corresponding author (United States is the reference category), and whether the experimental hypothesis tested in the meta-analysis predicted a protective effect (OR<1) or not (an “expectation” dummy variable). A positive interaction term with this variable indicates a greater likelihood to deviate from the summary effect size in the direction predicted by the experimental hypothesis. Studies are classified as in Fig. 1.