The nature of biases.

Biases arise from cognitive shortcuts, or heuristics, that are deeply ingrained in our decision-making processes and often lead to systematic errors and irrational behaviors, significantly impacting our interactions and perceptions of others [1–3]. The study of biases has long intrigued social psychologists, with pioneering work by Daniel Kahneman and Amos Tversky in the early 1970s laying the foundation for understanding the pervasive impact of heuristics on human cognition. Tversky and Kahneman [4] and Chen and colleagues [5] revealed that biases result from the fast, automatic, and intuitive thinking. Therefore, individuals are not necessarily aware of their biases.

Individuals who focus more on positive than negative information about options, for example, are biased (e.g., optimistic bias; [6]). Since biases are cognitive failures [7,8] they influence individual behavior but also have broader implications for societal outcomes. For example, biases can affect hiring decisions, interpersonal relationships, and even policy-making [9]. When individuals become aware of their own biases, this self-awareness can affect their self-regulation [10], self-concept [11,12], and social interactions [13]. However, different types of biases may affect individuals differently.

Thus, a separate problem is the issue of bias types itself. Research is frequently limited to one or several biases. To our best knowledge, the typology of biases is understudied [14]. Several studies focus on positive versus negative interpretation biases. Baumgardner and colleagues [15] created a measure for positive and negative interpretation biases. Hannuschke and colleagues [16] define personality-congruent perception bias as positive when it reflects a perception bias: neuroticism predicted participants’ ratings of their interaction partners’ sociability and warmth indicating a positive but not a negative bias in interpersonal perception. Positive perception bias may result from a personality-congruent contrast effect: neurotic people see themselves as less sociable than others. Authors did not find a personality-congruent selection effect (negative biases). Insufficient psychological literature on negative versus positive roles of biases in many contexts of human behavior make us enlarge the potential theoretical basis.

Mentioned above there are a few approaches that formulate distinct types of biases [15,16]. These approaches primarily classified biases based on the mechanisms that give rise to them and their societal consequences. In our research, we propose a new typology by classifying biases based on their effects on individuals once they become aware of their biases. We identified two types of biases: ego-nonthreatening biases, conceptualized as the Metacognitive Self, and ego-threatening biases. We focus on whether awareness of the bias is positive or negative for the individual and how this awareness influences self-regulation. Positive biases can enhance social interactions and well-being as, for example, in positivity bias. Conversely, negative biases can diminish social interactions and lead to behaviors such as self-handicapping or behavioral consequences of dehumanization. Awareness of negative biases can result in self-threats and lower self-regulation, whereas awareness of positive biases fosters positive emotions, well-being, and self-regulation. We term this awareness of positive biases as the metacognitive self, based on previous literature, while our exploration of the consequences of awareness of negative biases is novel.

Thus, in our research, we propose a new typology by classifying biases based on their effects on individuals once they become aware of their biases. We identified two types of biases: ego-non threatening biases, conceptualized as the Metacognitive Self, and ego-threatening biases.

Self-awareness of ego-nonthreatening biases (metacognitive self) and ego-threatening biases.

Self-awareness of one’s emotions, thoughts, and decisions imply metacognition, which is the understanding and regulation of one’s own cognitive processes. This concept dates back to Dewey [17] and has been expanded by scholars like Nelson and Narens [18]. Carden and colleagues [19] provided a comprehensive definition of self-awareness based on a systematic literature review: “Self-awareness consists of a range of components, which can be developed through focus, evaluation, and feedback, and provides individuals with an awareness of their internal state (emotions, cognitions, physiological responses), that drives their behaviors (beliefs, values, and motivations) and considers how these impact and influence others.”

Building on the foundation of self-awareness, the concept of the Metacognitive Self (MCS) has been proposed. It reflects an individual’s ability to understand and manage their cognitive processes and biases, promoting better self-knowledge and self-regulation. The Metacognitive Self is consistent with Taylor and Brown’s [20] idea that some biases can be motivationally adaptive, while also acknowledging Block and Colvin’s [21] limitations of this theory. It embraces the Aristotelian idea of the golden mean, balancing the adaptive role of psychological rules with accurate self-reflection.

Individuals with a high Metacognitive Self can better regulate their behavior by recognizing and managing their biases, leading to more consistent alignment with personal and social standards. High Metacognitive Self promotes deeper self-reflection and accurate assessment of one’s actions, which is crucial for personal growth and development. The balance achieved through Metacognitive Self helps individuals maintain motivation without falling into the traps of overconfidence or excessive self-criticism. With high Metacognitive Self, individuals can internalize self-knowledge, making it readily accessible and applicable in various situations, thus enhancing decision-making and behavior [22,23].

Kleka and colleagues [24] expanded on the role of metacognitive self and intrinsic epistemic motivation in shaping an individual’s awareness of biases. They suggest that feedback from emotional responses and interpersonal relationships helps individuals recognize biases within their own behavior. Similarly, Brycz and colleagues [25] found that strong self-awareness of positive biases (a high Metacognitive Self) fostered participants’ willingness to learn social adaptive behavior even in the face of failure in contrast to participants with a low Metacognitive Self.

Although the psychological literature on the causes and nature of biases is rich (e.g., [19,26–28], research has mainly focused on self-awareness of biases that do not threaten individuals’ ego (e.g., overgeneralizing positive impressions of others). Many findings show that higher self-awareness of ego-nonthreatening biases is associated with self-regulatory outcomes [29,30]. In contrast, individuals may also become aware of biases that threaten their self-esteem and moral self-views (e.g., dehumanizing the victim, prejudice). Preliminary findings indicate that self-awareness of ego-threatening biases is associated with negative outcomes for the individual [31] and teaching the nature of prejudice results in no outcome [32]. However, self-awareness of ego-threatening biases has rarely been investigated and systematically distinguished from self-awareness of ego-nonthreatening biases so far.

To close this gap, we developed and validated the self-awareness of ego-threatening biases questionnaire (SAETBQ). We predicted that self-awareness of ego-threatening biases is associated with defense mechanisms, a disjointed self-concept, and unfavorable social interactions, whereas self-awareness of ego-nonthreatening biases as measured by the Metacognitive Self (MSC) questionnaire is associated with self-regulation, an integrative self-concept, and favorable social interactions. Furthermore, we predicted Dark Triad traits (e.g., narcissism) to be positively and Light Triad traits (e.g., faith in humanity) to be negatively associated with self-awareness of ego-threatening biases.

The importance of self-awareness of biases – metacognitive approach.

Self-awareness implies metacognition, which is the understanding of one’s own cognitive processes [17,18]. Thus, metacognition requires not the automatic but the slow, deliberate, and rational thinking [4,5]. Previous research has assumed that improving rational awareness of personal biases will mitigate the effects of automatic biases. But so far research revealed mixed results. Whereas Azad and colleagues [33] showed that enhanced self-awareness through feedback helps individuals recognize and understand the automatic processes that form the basis of implicit biases, Quillin and colleagues [34] found only small effects of prior cultural competency training on prejudice towards Black individuals.

To better understand how self-awareness of personal biases affects individuals, self-awareness theory offers valuable insights. The theory suggests that self-awareness initiates self-regulation [35]. Accordingly, when individuals recognize and become aware of the cognitive processes underlying their biases, they actively engage in self-regulation. Brycz and colleagues [25] demonstrate that self-awareness of personal ego-nonthreatening biases enhances self-improvement through self-regulation. Szczepanik and colleagues [36] showed smooth brain emotion regulation increasing with self-awareness of biases (i.e., metacognitive self). However, self-awareness theory does not specify the various strategies individuals may use. We assume that awareness of different types of biases initiate the use of distinct strategies.

Effects of bias types on individuals.

The typology of biases, as we mentioned above, is understudied. Single biases are in the eye of investigators. For example, the Halo Effect describes the cognitive error that when someone is perceived as warm, they appear more approachable, leading to more favorable interactions. Awareness of the Halo Effect bias allows individuals to engage positively with those who seem warm and to consciously adjust their reactions in a more positive manner to those who appear cold [37]. Another example is the fundamental attribution error, which attributes others’ behaviors to personality traits rather than situational factors. Awareness of this bias encourages individuals to consider external factors influencing others’ behaviors, fostering empathy and patience in social interactions [38].

Previous literature on self-awareness of biases (e.g., Metacognitive Self) has focused mostly on positive biases. These biases do not pose a threat to the ego because they maintain or enhance self-regulation. We propose that there are biases that pose a threat to the ego when individuals become aware of them and introduce them as ego-threatening biases.

Self-awareness of ego-threatening biases.

Previous literature shows that biases can have a negative impact on individuals and pose a threat to their ego [39]. These biases include negative rumination bias, which involves persistently thinking about negative feelings and emotions [40–43] and self-handicapping, where individuals create obstacles to their success to protect their self-esteem in the event of failure [44]. Another example is defensive projection [45,46] which involves attributing one’s own unacceptable traits or choices to others, which can lead to interpersonal conflicts and hinder self-improvement. Thus, biases can pose a threat to individuals’ ego.

Few studies have shown that biases, especially those that individuals have become aware of, have a negative impact on individuals. Pronin and Kugler [47] found that heightened awareness of one’s own biases can lead to increased stress and discomfort, as individuals struggle to reconcile their biased behaviors with their self-perception as unbiased. Similarly, Wells and Matthews [42] discuss how increased self-awareness of negative biases can exacerbate emotional dysfunctions, such as anxiety and depression, by fostering maladaptive rumination and self-criticism. Therefore, empirical research indicates that individuals who are aware of personal biases that pose a threat to their ego may prefer strategies such as defense mechanisms over self-regulation.

The negative role of self-awareness of ego-threatening biases on human functioning is supported by theories of cognitive dissonance [48] and self-enhancement motives [49]. High self-awareness of ego-threatening biases leads to a cognitive dissonance between negative information about oneself in the form of ego-threatening biases and self-enhancement motives. This forces individuals to balance the negative insights about themselves with their need for positive self-esteem. One way to maintain positive self-esteem while being aware of ego-threatening biases is to use defense mechanisms (e.g., avoiding self-diagnostic information) that protect the ego but also promote a disjointed self-concept. For example, Forster and colleagues [50] showed that for people with a prevention focus, stereotype disconfirmation is a threat to efficient and effective self-regulation when the disconfirmation is discrepant from stereotypic beliefs they endorse. This produces negative emotions, vigilance motivation, and higher attention to both the disconfirming target and its background location.

Furthermore, individuals who use defense mechanisms are more likely to neglect moral standards which foster unfavorable social interactions characteristic of the dark triad ( [51,52]. The dark triad describes individuals that have an inflated sense of self-importance and a lack of empathy for others (Narcissism), use manipulative strategies and a focus on personal gain at the expense of others (Machiavellianism), and act impulsively and lack remorse or guilt (Psychopathy) [53,54]. Because the dark triad and self-awareness of ego-threatening biases both lead to more destructive and antisocial behaviors, they are likely to show positive correlations. The traits from the light triad, in contrast, emphasize faith in humanity, humanistic ideals, and Kantian moral standards and are likely to show negative correlations with self-awareness of ego-threatening biases.

Self-Awareness of Ego-Threatening Biases Questionnaire (SAETBQ).

The initial item pool consisted of 129 positive and negative biases. These items were carefully selected based on their relevance to the construct of bias awareness and were drawn from previous research conducted by Brycz [56]. The SAETBQ initial item pool reflects everyday scenarios and thoughts, such as the dehumanization of the victim, exemplified by items like: When I treat someone badly, I think they deserve it because they are worse than me. Participants rated each item on a 6-point Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree).

Metacognitive Self (MCS) questionnaire.

We used the 24-item self-report questionnaire to assess the metacognitive self (MCS-24; [22]). Each item describes a behavioral aspect related to biases that foster self-regulatory functions and provide positive assertions about the self. An example item for positivity bias is: I tend to assess others and myself higher than we really are. Participants rated each item on a 6-point Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree). The scale showed good reliability in this study with Cronbach’s .

Self-diagnostic motive scale.

We used the self-diagnostic motive scale [25] to measure the drive to seek diagnostic information about the self. It includes 6 items across 3 subscales assessing the effects of one’s actions (e.g., What negative things does the test say about me?; Cronbach’s ), seeking self-improvement (e.g., What can I do to make myself better off in my life achievements?; ), and comparing one’s results with others (e.g., To what extent did I complete the task worse than others?; ). Responses are rated on a 6-point Likert scale from 1 (I definitely don’t want to know anything about it) to 6 (I definitely want to know about it). In the present study, internal consistency of the total scale was Cronbach’s .

Integrative self-knowledge scale.

We used the Polish adaptation [57] of the integrative self-knowledge scale [58] to measure an individual’s tendency to unify self-experiences over time. It includes 12 items covering past experiences (e.g., What I have learned about myself in the past has helped me to respond better to difficult situations), present awareness (e.g., Often, I am unaware of my thoughts and feelings as they are happening, and only later get some idea about what I may really have been experiencing), and future goals (e.g., By thinking deeply about myself, I can discover what I really want in life and how I might get it). Each item is rated on a 5-point Likert scale from 0 (largely untrue) to 4 (largely true). In the present study, internal consistency was Cronbach’s .

Moral disengagement.

We used the 8-item moral disengagement scale [59] to capture cognitive mechanisms that facilitate unethical behavior, such as moral justification (e.g., It is okay to spread rumors to defend those you care about) and displacement of responsibility (e.g., People shouldn’t be held accountable for doing questionable things when they were just doing what an authority figure told them to do). Items are rated on a 7-point Likert scale from 1 (strongly disagree) to 7 (strongly agree). In the present study, internal consistency was Cronbach’s

Selection of biases.

The first stage of our investigation involved a selection of negative biases from the initial item pool based on the dimension of ego-threat: how much does being aware of a given bias threaten an individual’s self-image? The initial item pool consisted of 129 well-known and popular biases (attribution biases, decision-making biases, and logical thinking biases like base rate neglect, and so forth) which 11 trained psychologists evaluated on mentioned dimension using a scale from −5 (“extremely ego-threatening”) to +5 (“not at all ego-threatening”). This evaluation process had already been carried out in a previous study by Brycz [56], where the biases were scored on the dimensions but not selected for the SAETBQ. The evaluation judgments were highly consistent, as indicated by a strong Kendall’s tau coefficient [56]. Out of the original 129 biases, we identified 55 that had a high potential to threaten one’s ego. We suggest that a bias should have a score of at least −1 from the judges to be considered as ego-threatening. These biases were likely to damage one’s self-image if admitted. We turned each of them into an item that reflected daily situations, thoughts, or feelings.

Psychometric item selection.

In the second stage, we subjected a set of 55 biases identified in the previous stage to further analysis. We divided our sample of 1,609 participants into a training set (80% − 1,287 participants) and a test set (20% − 322 participants) [60,61]. The psychometric item selection was conducted exclusively on the training dataset using an exploratory factor analysis (EFA) with repeated data splits [62]. In each random data split, 60% of the training data was selected to assess the factor loadings of the 55 items on a single factor. Items with loadings meeting or exceeding 0.45 were retained. The random data splits were repeated in 1000 iterations. Items that met the 0.45 loading threshold in at least 80% of iterations were deemed most robust and retained. Next, we conducted confirmatory factor analysis (CFA) on the training dataset to validate the selected items on independent data and evaluate its fit using common model evaluation measures [63–65]. We also conducted additional analyses including bootstrap EFA and exploratory analysis of multifactor solutions. As these are additional exploratory analyses, their details and discussions are provided in the Supplementary materials.

Construct validity.

The final stage involved conducting a correlation analysis to examine the relationships between the SAETBQ and various psychological constructs. We expected a positive correlation between self-awareness of ego-threatening biases, as measured by SAETBQ, and Moral Disengagement. This correlation would indicate that individuals who rationalize unethical behavior to protect their self-image are more likely to be aware of their negative biases (H1a). Conversely, we hypothesized negative correlations between self-awareness of ego-threatening biases and self-regulation indicators such as the Self-Diagnostic Motive Scale and its subscales (H1b). Self-awareness of ego-threatening biases may undermine self-regulation, as individuals struggle to maintain a positive self-view. Specifically, we expected that individuals with high self-awareness of ego-threatening biases would score lower on self-diagnostic motives, indicating a difficulty in integrating this awareness into a coherent self-concept (H1c). Additionally, we expected self-awareness of ego-nonthreatening biases, as measured by the Metacognitive Self questionnaire (MCS-24), to show negative correlations with moral disengagement (H2a) and positive correlations the self-diagnostic motive (H2b) and the integrative self-concept (H2c). These correlations would suggest that individuals who are aware of their positive biases are better able to self-regulate their behavior and maintain a coherent self-concept.

Positive correlations.

As expected, the SAETBQ demonstrated significant positive correlations with constructs related to moral disengagement. As listed in Table 3, we found a significant positive correlation between SAETBQ and the Moral Disengagement scale (r = .54, p < .01), indicating that individuals who score higher on SAETBQ tend to engage in cognitive mechanisms that justify unethical behavior. This correlation supports our assumption that individuals who are more aware of their negative biases tend to rationalize unethical behavior (H1a).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Means, standard deviations, correlations, and 95% confidence intervals for positive and negative correlations in Study 1 (N = 1609).

https://doi.org/10.1371/journal.pone.0327989.t003

Negative correlations.

The analysis revealed negative correlations between SAETBQ and constructs associated with self-regulation and positive self-view (see Table 3). We found a significant negative correlation with the composite Self Motive scale (r = −.08, p < .01), suggesting that awareness of ego-threatening biases undermines the drive to seek diagnostic information about oneself. This finding is further supported by the significant negative correlations with the Self Evaluation subscale (r = −.07, p < .01) and the Self Improvement subscale (r = −.16, p < .01), indicating that awareness of ego-threatening biases was associated with a lower ability to assess one’s actions and a reduced drive for self-improvement. However, the correlation with the Self Comparison subscale was not significant (r = .01, ns), indicating that self-awareness of ego-threatening biases does not impact self-comparison. The negative correlations align with those observed for the Integrative Self-Knowledge scale, which also showed a negative correlation with the SAETBQ (r = −.56, p < .01). Overall, these results suggest that individuals who are more aware of their negative biases struggle with self-regulation and maintaining a positive self-concept (H1b and H1c).

Although these correlations are small sized effects, they contrast with the correlations found for self-awareness of nonthreatening biases as measured by the Metacognitive Self scale (MCS-24). Consistent with H2a, the Metacognitive Self scale correlated negatively with Moral Disengagement (r = −.17, p < .01). Consistent with H2b, the Metacognitive Self scale showed strong positive correlations with the composite Self Motive scale (r = .46, p < .01), as well as its subscales Self Evaluation (r = .41, p < .01), Self Improvement (r = .45, p < .01), and Self Comparison (r = .33, p < .01). Consistent with H2c, the Metacognitive Self scale showed a significant positive correlation with the Integrative Self-Knowledge scale (r = .09, p < .01). These correlations support our assumption that the self-awareness of ego-nonthreatening biases has beneficial effects on the individual.

Interestingly, awareness of ego-nonthreatening and ego-threatening biases, as measured by MCS-24 and SAETBQ, respectively, appeared to be unrelated to each other (r = −.01, p = .68). This lack of correlation suggests that awareness of positive and negative biases operates independently within individuals. While awareness of positive biases contributes to self-regulation and a positive self-view, awareness of negative biases seems to undermine these aspects without influencing the other.

Dirty dozen dark triad scale.

We used the dirty dozen scale [66] that was adapted for use in Poland [67]. It is a 12-item self-report questionnaire designed to assess Narcissism (e.g., I tend to want others to admire me; Cronbach’s α = .832), Psychopathy (e.g., I tend to lack remorse; Cronbach’s α = .628), and Machiavellianism (e.g., I have used deceit or lied to get my way; Cronbach’s α = .842) with four items each. Participants were asked to rate the extent to which they believe each statement accurately describes their own behavior, using a 5-point Likert scale from 1 (not at all true) to 5 (extremely true).

Light triad scale.

We used the light triad scale by Kaufman and colleagues [68] that was adapted by Gerymski and Krok [69] for use in Poland to measure Faith in Humanity (e.g., I tend to see the best in people; Cronbach’s α = .704), Humanism (e.g., I tend to admire others; Cronbach’s α = .617), and Kantianism (e.g., I prefer honesty over charm; Cronbach’s α = .598) with four items each. Participants rated their level of agreement with each statement on a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree).