In Experiment 2, signed metacognitive bias scores showed no significant group effect or linear trend, failing to replicate the Experiment 1 signed-bias finding¶

Claim · All claims
Evidence · 1 study · causal r3 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? Analysis of signed metacognitive bias scores in Experiment 2 showed no significant group effect and no significant linear trend across the four intervention groups. → Ceri Ngai and Sam J. Gilbert 2026

Evidence¶

Ceri Ngai and Sam J. Gilbert 2026¶

Ceri Ngai and Sam J. Gilbert (2026). Metacognitive training facilitates optimal cognitive offloading. Cognitive Research: Principles and Implications. https://doi.org/10.1186/s41235-026-00714-0

q3 · i? · causal · r3

Pre-registered one-way ANOVA with linear contrast on signed metacognitive bias in Experiment 2 (N = 416). The analysis "did not show a significant group effect" or linear trend; numerically, Group 4 (M = -11.6, SD = 21.1) showed greater underconfidence than Group 1 (M = -5.5, SD = 30.9).

"Analysis of signed metacognitive bias scores did not show a significant group effect (F(3, 412) = 1.58, p = .194), nor a significant linear trend (F(1, 412) = 1.58, p = .209)."

Discussion¶

Learner Variables¶

• Self-Regulation — outcome: instruction changes it

Related Claims¶

• Self Assessment Accuracy Is Low Without Training — the general claim this finding bears on, with the studies behind it
• In Experiment 2, absolute metacognitive bias decreased linearly across the four additive intervention groups, and the full intervention group was significantly less biased than control — related
• In Experiment 1, the feedback intervention reduced pre-task signed metacognitive bias (underconfidence) but the group difference did not persist at post-task — related
• In Experiment 1, a five-trial metacognitive prediction-plus-feedback intervention significantly reduced reminder bias, producing more optimal reminder-setting behaviour — related
• Overconfidence (bias) decreased from pretest to posttest across all groups including control, with no group differences — related
• Providing a rubric reduces bias in self-assessment, moving judgments closer to expert ratings — related
• In Experiment 2, making metacognitive predictions alone did not improve calibration, performance feedback was sufficient, and explicit metacognitive feedback added no significant benefit — related