A dilution effect without dilution: When missing evidence, not non-diagnostic evidence, is judged inaccurately

Categorical Updating in a Bayesian Propensity Problem

We present three experiments using a novel problem in which participants update their estimates of propensities when faced with an uncertain new instance. We examine this using two different causal structures (common cause/common effect) and two different scenarios (agent-based/mechanical). In the first, participants must update their estimate of the propensity for two warring nations to successfully explode missiles after being told of a new explosion on the border between both nations. In the second, participants must update their estimate of the accuracy of two early warning tests for cancer when they produce conflicting reports about a patient. Across both experiments, we find two modal responses, representing around one-third of participants each. In the first, "Categorical" response, participants update propensity estimates as if they were certain about the single event, for example, certain that one of the nations was responsible for the latest explosion, or certain about which of the two tests is correct. In the second, "No change" response, participants make no update to their propensity estimates at all. Across the three experiments, the theory is developed and tested that these two responses in fact have a single representation of the problem: because the actual outcome is binary (only one of the nations could have launched the missile; the patient either has cancer or not), these participants believe it is incorrect to update propensities in a graded manner. They therefore operate on a "certainty threshold" basis, whereby, if they are certain enough about the single event, they will make the "Categorical" response, and if they are below this threshold, they will make the "No change" response. Ramifications are considered for the "categorical" response in particular, as this approach produces a positive-feedback dynamic similar to that seen in the belief polarization/confirmation bias literature.

Sensory and multisensory reasoning: Is Bayesian updating modality-dependent?

Bayesianism assumes that probabilistic updating does not depend on the sensory modality by which information is processed. In this study, we investigate whether probability judgments based on visual and auditory information conform to this assumption. In a series of five experiments, we found that this is indeed the case when information is acquired through a single modality (i.e., only auditory or only visual) but not necessarily so when it comes from multiple modalities (i.e., audio-visual). In the latter case, judgments prove more accurate when both visual and auditory information individually support (i.e., increase the probability of) the hypothesis they also jointly support (synergy condition) than when either visual or auditory information support one hypothesis that is not the one they jointly support (contrast condition). In the extreme case in which both visual and auditory information individually support an alternative hypothesis to the one they jointly support (i.e., double-contrast condition), participants' accuracy is not only lower than in the synergy condition but near chance. This synergy-contrast effect represents a violation of the assumption that information modality is irrelevant for Bayesian updating and indicates an important limitation of multisensory integration, one which has not been previously documented.

What Proportion of Systematic Reviews and Meta-Analyses Published in the Annals of Surgery Provide Definitive Conclusions—A Systematic Review and Bibliometric Analysis

Objective: To perform a systematic review and bibliometric analysis of systematic reviews and meta-analyses published in the Annals of Surgery during a 10-year eligibility period and determine the unambiguity of concluding statements of these reviews published in the journal. Background: Systematic reviews and meta-analyses integrate clinically pertinent results from several studies to replicate large-volume, ‘real world’ scenarios. While the assimilation of results from multiple high-quality trials are at the summit of the evidence-base, the increasing prevalence of reviews using low-to-moderate levels of evidence (LOE) limit the ability to make evidence-based conclusions. In surgery, increasing LOE are typically associated with publication in the highest impact surgical journals (e.g., Annals of Surgery). Methods: A systematic review was performed as per PRISMA guidelines. An electronic search of the Annals of Surgery for articles published between 2011 and 2020 was conducted. Descriptive statistics were used. Results: In total, 186 systematic reviews (with or without meta-analyses) were published in the Annals of Surgery between 2011 and 2020 (131 systematic reviews with meta-analyses (70.4%) and 55 without meta-analyses (29.6%)). Study data were from 22,656,192 subjects. In total, 94 studies were from European research institutes (50.5%) and 58 were from North American institutes (31.2%). Overall, 75.3% of studies provided conclusive statements (140/186). Year of publication (P = 0.969), country of publication (P = 0.971), region of publication (P = 0.416), LOE (P = 0.342), surgery performed (P = 0.736), and two-year impact factor (IF) (P = 0.251) failed to correlate with conclusive statements. Of note, 80.9% (106/131) of meta-analyses and 61.8% of systematic reviews (34/55) provided conclusive statements (P = 0.009, †). Conclusions: Over 75% of systematic reviews published in the Annals of Surgery culminated in conclusive statements. Interestingly, meta-analyses were more likely to provide conclusive statements than systematic reviews, while LOE and IF failed to do so.