Decision-making is a crucial process in which one’s choices impact their quality of life. An important factor in making decisions is confidence, the subjective experience of being correct, and appraising one’s beliefs in light of evidence and past experience (Baron, 2000; Kleitman, 2008). In fact, some researchers claim that ‘confidence controls action’ (Gilovich, Griffin, & Kahneman, 2002, p.248). A key question of this research is whether confidence levels people assign to their on-task performance reflect their competence or reflect what might be arrogance. Ideally, confidence should arise from the accuracy of performance, and in more general terms, from intelligence, meaning more intelligent individuals should be more confident. This is not always the case, with ample examples that the confidence people express do not correspond (or are not justified) by their levels of performance and/or intelligence. This research aims to situate on-task confidence in the competence-arrogance taxonomy.
Past research has found positive correlations between cognitive ability and resistance to decision-making biases, such as overconfidence, framing errors, conjunction errors and statistical reasoning errors (Stanovich & West, 1998, 2000). This supports the notion that competent decision-making is due, in part, to intelligence. There is a consensus that decision-making competence depends on four underlying skills (Bruine de Bruin, Parker, & Fischhoff, 2007; Parker & Fischhoff, 2005). These are:
(1) belief assessment, which involves judging the likelihood of an outcome,
(2) value assessment, which requires evaluating that outcome,
(3) integration, combining the relevant beliefs and values in the decision-making process,
(4) metacognition, self-awareness of one’s abilities.
According to Parker & Fischhoff (2005), performance is determined through accuracy and internal consistency. Accuracy is measured relative to an external criterion (whether a judgement matches an actual estimate), while internal consistency focuses on the correspondence of between-question judgements (Parker & Fischhoff, 2005). The quality of decision-making depends on the process, emphasising the way the decision is reached, as well as the outcome. Underpinning the model is the assumption that an individual with better decision-making skills is more likely to produce good (e.g., more effective, accurate, more optimal) decision outcomes (Bruine de Bruin et al., 2007). These optimal decision outcomes should yield positive feedback for the individual, contributing to the individual’s experience and beliefs. Consequently, individuals with higher cognitive and decision-making abilities should exhibit higher confidence in their decision outcomes.
However, only a moderate relationship between confidence and intelligence has been found, with correlations averaging between .40 and .50 (Kleitman & Stankov, 2007; Kleitman, 2008). Therefore, this discrepancy in confidence and intelligence may be attributed to incompetence, defined as a lack of skill and poor performance. Ehrlinger and colleagues observed that some of those who are most confident in their abilities were not necessarily the most competent (Ehrlinger, Johnson, Banner, Dunning, & Kruger, 2008). That is, incompetent individuals were ‘overly optimistic’ when estimating their own performance (Ehrlinger et al., 2008, p. 119). This was attributed to a lack of awareness of errors, which suggests that those who are incompetent have such poor metacognitive skills that they lack the insight to recognise their errors in decision-making (Ehrlinger et al., 2008). A key research question is whether this failure to recognise errors is at least partly attributed to arrogance.
There has been minimal research into the domain of arrogance, however, with a majority considering it in relation to other constructs, such as narcissism and agreeableness. Arrogance has been found to correlate with some scales in narcissism, especially with entitlement (Johnson et al., 2010). In regards to making judgements, Farwell and Wohlwend-Lloyd (1998) found that narcissists tended to overestimate their academic grades, and held optimistic expectations for their performance. Additionally, the NEO-PI-R modesty subscale in Agreeableness was found to correlate negatively with arrogance, with a low score on the subscale suggesting that (1) the individual believed they were superior to others, and (2) perceived by others as displaying arrogance (Rowatt et al., 2006). This was supported by Silverman, Johnson, McConnell and Carr (2007), who found that higher arrogance was correlated with lower agreeableness and higher self-centrism. Therefore measuring aspects of arrogance through scales such as narcissism and agreeableness should offer insight into the effect of arrogance on incompetent decision-making.
Closely related to arrogance is dogmatism, a ‘resistance to new information, ignorance of one’s own limits, or arrogance about what one claims to know’ (Rowatt et al., 2006, p. 207). Parker and Fischhoff (2005) found negative correlations between decision-making competence and polarised thinking (thinking in terms of black-and-white), which is similar to dogmatic or rigid thinking. This has been further substantiated with evidence that individuals with a tendency towards rigid thinking have poorer outcomes in problem solving and general reasoning (Kleitman, 2008; Rowatt et al., 2006).
Aims and hypotheses
Few studies have examined the relationship between confidence and arrogance/dogmatism. The proposed study will model confidence, and place it within the competence and arrogance taxonomy. The overarching goal of the study is to extend previous work on competence and confidence in decision-making to include measures developed to capture different aspects of arrogance. Extending previous research, in addition to assessing competence in test-taking, competence will be also defined by the outcomes of decision-making tasks designed to capture optimality of decision-making and their errors. The study will also examine the factorial structure of measures employed to capture competence, arrogance and dogmatism. Several theoretically-driven models (see XX 1) will be compared and contrasted.
It is hypothesised that competence and arrogance/dogmatism will separate into two factors and predict confidence independently. While high levels of confidence and decisiveness are expected in both competent and arrogant individuals, it is postulated that arrogance will lead to higher levels of error. It is also predicted that arrogance and dogmatism will not separate into distinct factors, and thus predict confidence in the same way