Sensitivity and publication-bias analyses indicate the pooled CBL effect is stable despite slight publication bias¶

Claim · All claims
Evidence · 1 study (2 entries) · quant-synthesis r2 · q4
Currency · newest study 2023 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q4 i? Leave-one-out sensitivity analysis showed stable point estimates, and Egger's test indicated slight publication bias (t = 2.21, p = 0.046) that trim-and-fill showed did not overturn significance. → Wu Fanghui 2023

Evidence¶

Wu Fanghui 2023¶

Wu Fanghui, Wang Tao, Yin Danxu, Xu Xiaoxiao, Jin Cancan, Mu Nan, Tan Qingrong. (2023). Application of case-based learning in psychology teaching: a meta-analysis. BMC Medical Education. https://doi.org/10.1186/s12909-023-04525-5

q4 · i? · quant-synthesis · r2

Sensitivity analysis omitting studies one-by-one; the article reports point estimates stayed within the confidence interval. No effect size is printed for this stability finding.

"the point estimate of pooled effects after removing each study always stayed within the confidence interval, suggesting that the results were stable."

Wu Fanghui 2023 (2)¶

Wu Fanghui, Wang Tao, Yin Danxu, Xu Xiaoxiao, Jin Cancan, Mu Nan, Tan Qingrong. (2023). Application of case-based learning in psychology teaching: a meta-analysis. BMC Medical Education. https://doi.org/10.1186/s12909-023-04525-5

q4 · i? · quant-synthesis · r2

Funnel plot and Egger's test on the 15 studies indicated slight publication bias; trim-and-fill hypothetically filling five missing studies still yielded a significant pooled effect (Hedges' g = 1.69, 95%CI 1.41–2.04).

"Egger’s test was also performed as a reference, and the findings suggested that there was slight publication bias in the meta-analysis (t = 2.21; p = 0.046)."

Discussion¶

Related Claims¶

• The CBL exam-score advantage is robust to subgroup and leave-one-out sensitivity analyses but shows potential publication bias — related