The Cognitive And Neural Mechanism Of Processing And Effectivity Of Decision Confidence

Decision Confidence is the level of subjective correctness or optimization of individual judgments or decisions(Peterson & Pitz,1988). As a senior cognitive processing involving "metacognition" and "consciousness"(S. M. Fleming & Dolan,2012; S. M. Fleming, Weil, Nagy, Dolan,& Rees,2010), decision confidence is a link between the objective reality and subjective beliefs, resulting in different decision-making behaviors among different individuals in the same decision situation. Therefore, decision confidence may be one of the most basic compositions in the decision-making process(Kepecs Uchida, Zariwala,& Mainen,2008; Kiani & Shadlen,2009). Furthermore, decision confidence can affect a series of cognitive functions after decision process: the improvements of subsequent decision, learning of experience from results and demonstration of optimality of individual decision. Decision confidence, therefore, is an important bridge of individual cognition and emotion in the decision process, which has an important role in the adjustment and improvement in the decision process(Efklides,2008; Kepecs et al.,2008; Fontana & Fernandes,2011; T. Miller,2013). Therefore, exploring the cognitive and neural mechanism of processing and effectivity of decision confidence has high scientific value and practical significance.。More ordinarily, before an organism can judge the potential threats or the possible benefits of stimuli that depend on the noisy and insufficient environmental information, the degree of that decision uncertainty must be evaluated (Barthelme & Mamassian,2010). Before individual choose to still hold or sell for benefits when facing stocks becomes higher, they also need to make risk or conservative decisions according to their own subjective assessments. The producing and processing of decision confidence is relatively complicated process in the real life, which may require the participation with multiple brain activities. Furthermore, reporting self-confidence could play an important role for improvement in one’s own performance and have beneficial effects on their motivation or involvement (Fontana & Fernandes,2011; D. Kostons, van Gog,& Paas,2008; Danny Kostons, van Gog,& Paas,2010). And estimating confidence could be used to improve metacognitive monitoring accuracy and performance predictions of decision(Miller, 2013). However, there has sevel studies about that cognitive and neural mechanism of processing and effectivity of decision confidence。Firstly, the objective uncertainty of decision is based on the probability and potential value of outcome, and the subjective confidence not only depends on the judgment of objective uncertainty also depends on the estimation of the metacognition. It is still unknown the neural mechanisms of the subjective confidence when sequential updating outcome probability about a future decision? And which brain regions can represent the individual subjective confidence. Secondly, Kepecs and colleagues (2008) described neurons in the rat orbitofrontal cortex (OFC) as encoding decision confidence could guide the decision behaviors. The other evidence from macaque single neuron recordings showed that the lateral intraparietal cortex represented the degree of decision confidence and predicted the bet decision (Kiani and Shadlen,2009). However, we still need to explore the cognitive and neural mechanisms of process and represent underlying decision confidence. Finally, the research showed that "automatic" decision confidence had a positive role in regulating the process of gambling decision making, and the estimation of decision confidence could increase the degree of cautious in decision-making process. But, it remains unclear how "controlled" decision confidence estimating affects perceptual decision process from cognitive and neural perspective, and which brain regions can represent the effectivity of decision confidence? Therefore, adopting the behavioral and function Magnetic Resonance Imaging (fMRI) measures, the present study based on the two dimensions of "automatic or controlled" and "single or post" from the following four experiments in two research, to investigate gradually the cognitive and neural mechanisms of processing and effiectivity of decision confidence.The Research 1 included two experiments using the functional magnetic resonance imaging (fMRI), aiming to investigste the neural processing and representation of decision-making confidence. The experiment 1 investigated the neural mechanisms of "controlled" subjective decision confidence by controlling environmental objective uncertainty (reward probabilities are 1/9,1/7,1/5,1/3 and 1). In the lottery game task, the participants should estimate their subjective confidence successively associated with the experience about how certainty they can gain gold and make decision. Moreover, the objective uncertainty (reward probability) was dynamically changed as information observed. The behavioral results indicated that individual subjective confidence increased significantly with the objective uncertainty decreased demonstrating a significant positive correlation (r=0.531, p<0.001). Increased brain regions activities correlated with the subjective confidence mainly included the inferior frontal gyrus (BA9/13/45/47), middle frontal gyrus (B46/6), medial frontal fyrus (BA32/6/9), precuneus and parietal lobule. More importantly, the level of estimating confidence for each objective uncertainty (i.e.1/9,1/7,1/5 and 1/3) was performed regression analysis with their brain activities (Beta) separately. The positive correlation for all objective uncertainty (i.e.1/9,1/7,1/5 and 1/3) was only observed in the middle frontal gyrus (1/9:r=0.457, p=0.049; 1/7:r=0.408, p<0.083; 1/5:r=0.443, p=0.058; 1/3:r=0.476, p=0.04). And the positive correlations for the objective uncertainty (i.e.1/9 and 1/7) was observed in the superior parietal lobule (1/9:r=0.538, p=0.018; 1/7:r=0.492, p=0.033) and inferior parietal lobule (1/9:r=0.467, p=0.044; 1/7:r=0.428, p<0.068). It demonstrated that the more activated in these regions, the high levels of subjective confidence. Therefore, the middle frontal gyrus and parietal lobule might represent the individual subjective confidence. The experiment 2 investigated the neural mechanisms of "automatic" decision confidence process by manipulating the gambling decision conditions (forced decision and optional decision). More specifically, the individual should judge the larger area of rectangular images stimuli. After that, in the optional decision condition, subjects can make decisions of accepting or waiving the gambling by estimate the degree of correctness of perceptual tasks (decision confidence level). However, subjects have to make decisions according to the computer in the forced condition. Behaviourally, compared to the forced bet decision, the accuracy of optional bet decisions increased significantly. And compared to the forced pass decision, the accuracy of optional pass decisions decreased significantly. Moreover, the accuracy of optional bet decision increased significantly than the optional pass decision. These results indicated that the optional bet decision was based on confidence estimation of perception judgment and priming automatic confidence estimation positively affected the bet decision process. In addition, compared to the forced bet decision, the reaction time of optional bet decisions prolonged significantly. And compared to the forced pass decision, the reaction time of optional pass decisions prolonged significantly. The result indicated that priming automatic confidence processing could lead to the confidence evaluation of perception judgment which consumed more resources time to cognitive process. Decision confidence process was mainly involed in the cingulate cortex (BA24/32), superior parietal lobule (BA7) and temporal cortex (BA21/40/22). And confidence level process was mainly associated with the inferior parietal lobule (BA40), lingual gyrus (BA18) and cuneus (BA18). More importantly, there was a positively correlation between the activation in the left inferior parietal and optional waiving rate, that was, the higher brain activation, the more waiving decision. The significance of the effect was a strong indication that the activation in this region could represent the decision confidence processing and predict the extent to waive the present bet decision.The Research 2 included two experiments which using behavioral and fMRI, aiming to investagete the cognitive and neural mechanism of effectivity of decision confidence. By manipulating the Report condition (estimating the confidence) and Follow condition, the experiment 3 investigated the cognitive mechanism of effectivity of "controlled" decision confidence on the different difficulty perception decision-making process. Using experimental design within subjects, we selected three difficulty levels of perception judgment task (hard, middle and easy). More specifically, in the Report condition, after each perceptual decision participants were required to report the level of confidence about correctness of their decision by sliding cursor onto an appropriate scale. Behaviorally, the reaction time of perceptual decision with estimating confidence was observed significantly longer and the accuracy of perceptual decision was observed significantly increased than the perceptual decision without estimating confidence. These results indicated that estimating confidence had a positive role in regulating the process of perception decision and made the decision performance better. The experiment 4 further investigated the neural mechanism of effectivity of "controlled" decision confidence on the perception decision-making process by manipulating the Report condition (estimating the confidence) and Follow condition (following the computer). Using mixed experimental design, we selected the difficulty perception judgment task. As in experiment 3 participants should report the level of confidence about correctness of their decision after each perceptual decision. However, in the Follow condition, they made response accooding to the computer. Behaviorally, firstly, no difference of perceptual processing was observed between subjects. The order of confidence evaluation could not affect the accuracy of perception judgment. And the participants had not practice effect. Secondly, the reaction time of perceptual decision with estimating confidence was also observed significantly longer and the accuracy of perceptual decision was also observed significantly increased than the perceptual decision without estimating confidence. These results further indicated that estimating confidence had a positive role in regulating the process of perception decision process. Brain regions involved in the perceptual decision making process with estimating confidence including the left post cingulate cortex, which showed increased activities correlated with the effectivity of decision confidence on the perception decision-making process. More importantly, there was a positive correlation between the activation of the left post cingulate cortex and the improved accuracy of perception decision. It showed that the more activated in the region, the larger increased accuracy. Therefore, the post cingulate cortex might represent the effectivity of decision confidence on the perception decision-making process. At last, the physiological and psychological analysis showed that the functional connection intensity between the post cingulate cortex and insula was higher in perception decision with estimating confidence.In summary, this study had found the following conclusions. Firstly, regardless of the automatic or controlled, priming the decision confidence estimates had some degree of positive effect on the gambling decision making or sensory judgment process. And in present study, both the automatic and controlled decision confidence processing would occupy the individual cognitive resource, which was the control processing. Secondly, by priming control confidence estimates we found that the middle frontal gyrus and the left parietal lobule might represent the individual subjective confidence, which might be the sensitive brain regions of subjective confidence estimation. And by priming automatic confidence estimates, we found that the left parietal lobule activity could well predict behavior data, which might be sensitive regions of the decision confidence representation. Furthermore, by priming control confidence estimates, we found that the post cingulate cortex might represent the effectivity of decision confidence, which might be sensitive regions of the effectivity of confidence estimation on the perception decision process. Therefore, the results had some theoretical significance for interpretation cognitive processing mechanism and improvement mental model of decision confidence. At the same time, the results had some practical significance for learning the adjustment role of decision confidence on the current and subsequent decision, and making a more scientific and rational decision-making. In addition, above fMRI results provided preliminary the neural basis of evaluation process and improving effectivity of decision confidence in the different decision situations, and investigation of the sensitive brain regions of individual subjective confidence and the effectivity of decision confidence.