Table 5 Nonparametric trim-and-fill analysis of publication bias, linear estimator, imputing on the left: random effects model
Number of studies | Observed | Imputed |
|---|---|---|
40 | 33 | 7 |
Studies | Effect size | 95% Conf. interval |
Observed | −0.053 | (−0.085, −0.022 ) |
Observed + imputed | −0.076 | (−0.106, −0.045 ) |