Measures of metacognitive efficiency across cognitive models of decision confidence

Abstract

Meta-d’/d’ has become the quasi-gold standard to quantify metacognitive efficiency because it has been assumed that meta-d’/d’ provides control for discrimination performance, discrimination criteria, and confidence criteria even without the assumption of a specific generative model underlying confidence judgments. Here, we investigate the robustness of meta-d’/d’ using simulated data from a variety of different generative models of confidence. For most models of confidence, meta-d’/d’ is affected by discrimination performance, discrimination task criteria, and confidence criteria. The single exception was a generative model of confidence according to which the evidence underlying confidence judgements is sampled independently from the evidence utilized in the choice process from a Gaussian distribution truncated at the discrimination criterion. According to all alternative generative models of confidence examined in the present study, there exist at least some combination of parameters where meta-d’/d’ is affected by discrimination performance, discrimination criteria and confidence criteria. These simulations imply that a difference in meta-d’/d’ between conditions does not necessarily reflect a difference in metacognitive efficiency but might as well be caused by a difference in discrimination performance, discrimination task criterion, or confidence criteria.

Unique effects of sedatives, dissociatives, psychedelics, stimulants, and cannabinoids on episodic memory: A review and reanalysis of acute drug effects on recollection, familiarity, and metamemory.

“…Considering that the metacognitive decisions of interest happened during retrieval, drug effects isolated to encoding and consolidation provided a test for the separability of memory and metamemory. Consistent with work highlighting blurred boundaries between cognition and metacognition in perception (Fetsch et al, 2014;Peters et al, 2017;Rausch et al, 2023;Zheng et al, 2023), we found that drug manipulations of encoding and consolidation could indeed impact metamemory decisions at retrieval. This nonindependence between memory and metamemory is inherent in the meta-d′ model, which assumes the information available for the Type 1 task is exhaustive of the information available for the Type 2 task (Maniscalco & Lau, 2012;cf.…”

“…This nonindependence between memory and metamemory is inherent in the meta- d ′ model, which assumes the information available for the Type 1 task is exhaustive of the information available for the Type 2 task (Maniscalco & Lau, 2012; cf. Rausch et al, 2023). Nevertheless, unlike metacognitive measures that largely scale with cognitive performance, there was clearly not a one-to-one mapping between memory and metamemory (i.e., high or low memory accuracy did not necessarily entail high or low metamemory, respectively).…”

Domain Generality in Metacognitive Ability: A Confirmatory Study Across Visual Perception, Memory, and General Knowledge

“…This is a general challenge for all research within the field of metamemory, and new tasks that minimize this limitation should be developed. Further, there was a strong correlation in metacognitive bias between all domains, which may confound the estimated correlations in metacognitive efficiency (Guggenmos, 2021;Rausch et al, 2023;Shekhar & Rahnev, 2021a). Future studies may benefit from manipulations that can decorrelate confidence bias across tasks, such as altering metacognitive criterion through reward or task instructions.…”

Contributions of tactile information to the sense of agency and its metacognitive representations

“…First, is it possible that this is caused by an estimation error: Despite the online adaptive staircases, discrimination performance (d’) was consistently lower in the spatial condition than in the other two conditions. Mratio is, in theory, independent of discrimination performance (Maniscalco & Lau, 2012), but in practice a dependency exists (Guggenmos, 2021; Rausch et al, 2023). Because Mratio increases slightly with decreasing d’, this could explain (partially, at least) both the consistently higher Mratio values in the spatial condition and the absence of correlations with Mratio in the other conditions.…”