How Confidence Influences Information Search Strategies

点击翻译按钮获取中文摘要

Uncertainty is a core challenge when making decisions in the real world. To reduce uncertainty, individuals actively gather additional information until a decision can be made, trading off betweenthe cost of seeking further evidence and the benefit of being more likely to make the optimal decision. However, it has repeatedly been shown that people exhibit a systematic confirmationbias in information selection. That is, humans tend to seek, interpret, and recall information that favour their beliefs or support personal pre-existing hypotheses. Through contributing tooverconfidence, this confirmation bias can lead to systematic errors in reasoning, belief perseverance even in light of contrary evidence, and attitude polarization. In my doctoral research,I would like to investigate the influences of cognitive biases such as confirmation and overconfidence on decision-making in uncertain environments. When we are less confident, arewe even more prone to seeking information favouring our initial position to gain more confidence, or do lower levels of confidence make us more open and tolerant towards potentially disagreeinginformation?Gaining an advanced understanding of how our meta-attitudes influence our tolerance towards divergent opinions could inform interventions to counteract phenomena like attitude polarizationand belief perseverance. Cooperation requires that people are tolerant towards opposing opinions and open to be influenced in their attitudes and beliefs. Possible interventions could seek toprovide additional information on the topic as such. However, considering that meta-attitudes might influence the intake of information, instead targeting the meta-level first might be usefulwhen trying to bring about changes in opinions. Building on this line of research, it could further be investigated to what extend the confirmation bias serves a goal, and whether by changingpeople's goal, this bias can be changed. Furthermore, it would be intriguing to dissociate individuals' biases in information gathering from biases in belief updating. Even when informationsampling is biased, belief updating may still occur normatively. Another interesting avenue to pursue would be to study the integration of inconsistent evidence from different sources. Aversiontowards conflicting information may lead people to refrain from getting further advice even though this might not be an optimal strategy in gathering information to reduce uncertainty. Exploring therelation of confidence to cognitive biases in information seeking and decision-making offers lots of scope for intriguing research. Findings may be useful in developing decision-aids,complementing research on artificial intelligence, constructing interventions to counteract attitude polarization, or informing social policies