Overconfidence across the psychosis continuum - a calibration approach

TY - JOUR

T1 - Overconfidence across the psychosis continuum - a calibration approach

AU - Balzan, Ryan P

AU - Woodward, Todd S

AU - Delfabbro, Paul

AU - Moritz, Steffen

PY - 2016/10/6

Y1 - 2016/10/6

N2 - INTRODUCTION: An 'overconfidence in errors' bias has been consistently observed in people with schizophrenia relative to healthy controls, however, the bias is seldom found to be associated with delusional ideation. Using a more precise confidence-accuracy calibration measure of overconfidence, the present study aimed to explore whether the overconfidence bias is greater in people with higher delusional ideation.METHODS: A sample of 25 participants with schizophrenia and 50 non-clinical controls (25 high- and 25 low-delusion-prone) completed 30 difficult trivia questions (accuracy <75%); 15 'half-scale' items required participants to indicate their level of confidence for accuracy, and the remaining 'confidence-range' items asked participants to provide lower/upper bounds in which they were 80% confident the true answer lay within.RESULTS: There was a trend towards higher overconfidence for half-scale items in the schizophrenia and high-delusion-prone groups, which reached statistical significance for confidence-range items. However, accuracy was particularly low in the two delusional groups and a significant negative correlation between clinical delusional scores and overconfidence was observed for half-scale items within the schizophrenia group. Evidence in support of an association between overconfidence and delusional ideation was therefore mixed.CONCLUSIONS: Inflated confidence-accuracy miscalibration for the two delusional groups may be better explained by their greater unawareness of their underperformance, rather than representing genuinely inflated overconfidence in errors.

AB - INTRODUCTION: An 'overconfidence in errors' bias has been consistently observed in people with schizophrenia relative to healthy controls, however, the bias is seldom found to be associated with delusional ideation. Using a more precise confidence-accuracy calibration measure of overconfidence, the present study aimed to explore whether the overconfidence bias is greater in people with higher delusional ideation.METHODS: A sample of 25 participants with schizophrenia and 50 non-clinical controls (25 high- and 25 low-delusion-prone) completed 30 difficult trivia questions (accuracy <75%); 15 'half-scale' items required participants to indicate their level of confidence for accuracy, and the remaining 'confidence-range' items asked participants to provide lower/upper bounds in which they were 80% confident the true answer lay within.RESULTS: There was a trend towards higher overconfidence for half-scale items in the schizophrenia and high-delusion-prone groups, which reached statistical significance for confidence-range items. However, accuracy was particularly low in the two delusional groups and a significant negative correlation between clinical delusional scores and overconfidence was observed for half-scale items within the schizophrenia group. Evidence in support of an association between overconfidence and delusional ideation was therefore mixed.CONCLUSIONS: Inflated confidence-accuracy miscalibration for the two delusional groups may be better explained by their greater unawareness of their underperformance, rather than representing genuinely inflated overconfidence in errors.

U2 - 10.1080/13546805.2016.1240072

DO - 10.1080/13546805.2016.1240072

M3 - SCORING: Journal article

C2 - 27710199

VL - 21

SP - 510

EP - 524

JO - COGN NEUROPSYCHIATRY

JF - COGN NEUROPSYCHIATRY

SN - 1354-6805

IS - 6

ER -