‘Jumping to conclusions’ data-gathering bias in psychosis and other psychiatric disorders — Two meta-analyses of comparisons between patients and healthy individuals

Abstract

There has been an increase in attention to studying shared mechanisms underlying psychiatric disorders. The 'Jumping to conclusions' (JTC(1)) bias, a tendency to make decisions with certainty based on insufficient information, has been reported in patients with psychosis, and process-based treatment protocols targeting this bias have recently been developed. This review aimed to investigate to what extent the JTC bias, measured by various tasks, is associated with psychotic disorders and other psychiatric disorders using a meta-analytic approach. We examined 6864 articles published between 1990 and 2015, and meta-analysed 46 studies. The first meta-analysis included 40 effect sizes comparing patients with schizophrenia spectrum or other psychotic disorders and healthy controls. There was a hastier data-gathering style in patients with psychosis than healthy individuals, with a moderate aggregated effect size. The second meta-analysis included 18 effect sizes comparing patients with non-psychotic disorders and healthy controls. There was marked heterogeneity in effect sizes and evidence for publication bias. After removal of outliers, the aggregated effect size for JTC was not statistically significant. A planned subgroup analysis showed no significant effect of JTC in depression. Other diagnostic subgroups yielded small non-significant results. Therefore, our findings do not support the suggestion that JTC is a transdiagnostic phenomenon beyond psychosis.

Analytic cognitive style, not delusional ideation, predicts data gathering in a large beads task study

“…Participants diagnosed with clinical delusions requested significantly fewer beads than psychiatric and healthy controls. Dozens of beads task studies have subsequently been published, many of which have replicated this result (Dudley, Taylor, Wickham, & Hutton, 2016;Garety & Freeman, 1999, 2013McLean, Mattiske, & Balzan, 2016;So, Siu, Wong, Chan, & Garety, 2016). Furthermore, limited data gathering in the beads task is associated with higher delusional ideation in nonclinical populations (Ross, McKay, Coltheart, & Langdon, 2015), which is consistent with theories that propose that clinical delusions lie on a continuum with normal belief (Heriot-Maitland & Peters, 2015;Larøi, Raballo, & Bell, 2015;Linscott & van Os, 2013).…”

“…1 Such proposals are consistent with recent meta-analyses that have found that reduced data gathering is associated with increased self-reported delusional ideation (Ross et al, 2015) and psychotic diagnoses (Dudley et al, 2016;McLean et al, 2016;So et al, 2016). In addition, recent studies have reported that data gathering in the beads task is associated with performance in an analytic reasoning task (Brosnan, Hollinworth, Antoniadou, & Lewton, 2014) and self-reported "systemising" (Brosnan, Ashwin, & Gamble, 2013).…”

Psychotic Experiences and Overhasty Inferences Are Related to Maladaptive Learning

“…In line with the idea that overhasty and erroneous perceptual inferences from irrelevant noise information are implicated in hallucinations and hallucination-proneness, hallucinatory experiences have been repeatedly associated with a greater tendency to perceive illusory contents in auditory noise [9,10]. Moreover, delusional ideation has been consistently linked to “jumping to conclusions” (JTC, see [11–13] for detailed meta-analyses), a cognitive reasoning bias that leads to a rash acceptance of hypotheses based on little evidence. However, it is a matter of ongoing debate, which particular cognitive alteration predisposes delusional and delusion-prone individuals for an overhasty acceptance of possible hypotheses [14–17].…”

Studying Delusions Within Research Domain Criteria: The Challenge of Configural Traits When Building a Mechanistic Foundation for Abnormal Beliefs

“…Garety et al have consistently argued that abnormal beliefs derive from a "jumping to conclusions bias." 17,18 Although reasonable people can disagree, this bias may be analogous to failing to inhibit or suppress the discomfort of not knowing. It may alternatively relate to failures in response selection (selecting a nondominant response), performance monitoring, or it may relate to something else altogether (these researchers themselves also describe it variously as a data-gathering or reasoning bias).…”

“…Much of this work is narrowly balanced on a single task (the beads task), performance on which correlates strongly with working memory (another RDoC construct) and IQ. 19 Differences between psychotic patients and controls on the task are sizeable, Cohen's d = 0.61, 18 but modest compared to d ≈ 1.00 for many other neuropsychological main effects. Risen's context for thinking about how failures in any of several cold cognitive processes might allow for the formation and maintenance of abnormal beliefs may profitably be applied to expanding our thinking about the cognitive neuroscience constructs relevant to delusions.…”