Table 3.

Results from sensitivity analyses with data from studies including performance as the primary outcome

Sensitivity analysis	Analysis details	Effect size	Between study variance	Intraclass correlation	Probability of small effect
Between group: oral contraceptive pill withdrawal versus the early follicular phase of the menstrual cycle	34 effect sizes from 11 studies (combined quality = H/M/L; 27% H; 27% M; 27% L; 18% VL)	0.14 [− 0.14–0.38]	0.20 [0.01–0.59]	0.28 [0.0–0.82]	(d ≥ 0.2; p = 0.323; d ≤  − 0.2; p = 0.014)
Between group: oral contraceptive pill consumption versus all phases of the menstrual cycle except the early follicular phase	57 effect sizes from 16 studies (combined quality = M; 26.7% H; 33.3% M; 26.7% L; 13.3% VL)	0.14 [− 0.03–0.31]	0.10 [0.0–0.40]	0.42 [0.0–0.86]	(d ≥ 0.2; p = 0.257; d ≤  − 0.2; p = 0.001)
Within group: oral contraceptive pill consumption with oral contraceptive pill withdrawal	141 effect sizes from 21 studies (combined quality = H/M; 33.3% H; 33.3% M; 19.1% L; 14.3% VL)	0.05 [− 0.03–0.11]	0.06 [0.0–0.17]	0.19 [0.0–0.66]	(|d|≥ 0.2; p < 0.001)

Results are from multilevel random effects models with median parameter estimates and 95% credible intervals (95% CrI)

H high, M moderate, L low, VL very low