CoChat
Sign Up
Cited by 2Social and Intergroup PsychologyNames, Identity, and Discrimination ResearchDecision-Making and Behavioral Economics

Everyone I Don’t Like Is Biased: Affective Evaluations and the Bias Blind Spot

Open access
Walker, Alexander C., Collins, Robert N., Walker, Heather E. K., Fugelsang, Jonathan A., Mandel, David R.·2026·Personality and Social Psychology Bulletin
DOI 10.1177/01461672261439247
Open this paper in CoChat — free account

Abstract

People commonly exhibit a bias blind spot (BBS), judging themselves as less susceptible to bias than the “average other.” However, less is known about how people attribute bias to familiar others who evoke strong affect. We examined whether attributions of bias are sensitive to affective impressions of others. In Experiment 1, participants viewed themselves as considerably less biased than the average survey respondent and a personally-known disliked other, but not less biased than a familiar individual whom they liked. Experiments 2 and 3 examined the BBS in politically polarized groups of Democrats and Republicans. While participants judged themselves as somewhat less biased than co-partisans, they viewed themselves as much less biased than their political opponents. In all experiments, the effect of other target selection on the BBS was mediated by affective evaluations. We discuss the theoretical implications of affective evaluations guiding how people attribute bias to familiar others.

AI snapshot

Written from the paper’s full text.

TL;DR

Across three preregistered experiments, people's attributions of bias to familiar others tracked their affective evaluations of those others: participants showed a bias blind spot relative to the "average survey respondent" and disliked others, but not relative to a well-liked familiar other, and showed much larger blind spots for political out-group members than co-partisans. The authors argue affective impressions serve as heuristic cues guiding bias attribution, independent of the introspection-versus-behavior strategies emphasized by the introspection illusion account.

Key findings

  • Experiment 1 (N = 633): participants rated themselves as less biased (M = 3.56, SD = 0.90) than a target other (M = 4.31, SD = 1.03), t(632) = 17.14, p < .001, d = 0.68. Less bias was attributed to liked well-known others (M = 3.50, SD = 0.93) than disliked well-known others (M = 4.69, SD = 0.89), d = 1.29, or the average survey respondent (M = 4.72, SD = 0.77), d = 1.43. Bias blind spot scores were large for the average survey respondent (M = 1.00) and disliked other (M = 1.25), but absent for the liked other (M = −0.02, p = .784).
  • Affect tracked bias attribution: across conditions, more positive affect toward the target correlated with smaller bias blind spots, r(631) = −.52, p < .001, and this held within each condition (average respondent r = −.23; liked other r = −.18; disliked other r = −.31).
  • Experiment 2 (N = 626): participants attributed more bias to the average political out-group member (M = 5.18, SD = 0.91) than in-group member (M = 4.14, SD = 0.96), t(624) = 13.91, p < .001, d = 1.11, producing larger blind spots for out-group (M = 1.59, SD = 1.38) than in-group targets (M = 0.41, SD = 0.98), d = 0.98. The effect of reference group on blind spot scores was partially mediated by affect (b = 0.62, 95% CI [0.40, 0.80]).
  • Experiment 3 (N = 636, within-subjects): blind spot was larger for out-group (M = 1.48, SD = 1.45) than in-group targets (M = 0.43, SD = 1.00), t(635) = 19.91, p < .001, d = 0.79; affect fully mediated the reference-group effect (b = −0.98, 95% CI [−1.14, −0.82]). Democrats showed larger blind spots (M = 1.08, SD = 0.97) than Republicans (M = 0.83, SD = 1.11), driven by greater bias attributed to out-group members (Democrats M = 5.38 vs. Republicans M = 4.92, d = 0.51).

Methods

  • Three preregistered online experiments (Prolific; 660 recruited per experiment, ~$4.50 USD for a 25-min questionnaire; final Ns = 633, 626, 636 after preregistered exclusions). Experiment 1 randomly assigned participants to rate themselves and either the average survey respondent, a liked well-known other, or a disliked well-known other; Experiments 2 and 3 recruited U.S. Democrats and Republicans and had them rate themselves plus the average Democrat and/or Republican.
  • Bias blind spot was measured with a task adapted from Scopelliti et al. (2015): 14 bias descriptions in Experiment 1 and 7 non-partisan items in Experiments 2–3 (selected via a pilot study, N = 100, to be rated similarly bias-like by both parties), rated 1 (Not at all) to 7 (Very much); blind spot scores = mean Other minus mean Self rating. Affect toward targets was rated −3 (Extremely negative) to +3 (Extremely positive), and bias assessment strategy (introspection, behavior observation, declarative knowledge) was assessed in Experiments 1 and 2.
  • Analyses used mixed ANOVAs with Target as a within-subjects factor and reference group/partisanship as between- or within-subjects factors, correlations, independent-samples t-tests, and mediation analyses; data and annotated R scripts are on OSF.

Limitations

  • Stated: Experiments 2 and 3 lacked a neutral control target, so the authors could not determine whether bias attributed to partisans exceeded, matched, or fell below baseline evaluations of the average person; the affect manipulation relied on naturally occurring impressions, so target others may have differed on variables other than affect (e.g., disagreement); and the U.S.-centric partisan context raises generalizability questions, though the authors expect the affect–bias link to generalize across intergroup contexts and biases.
  • Evident from the text: the design is correlational with respect to affect (affect was not directly manipulated), so causal claims about affect driving bias attributions rest on mediation analyses of measured affect; the paper also notes that it did not assess the evidentiary basis for participants' bias attributions.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1177/01461672261439247

Cites

46 of 47 references verified1 citation errors · 0 not checkable

Statistics

Claims the paper makes

6 of 61 cited sources checked · 4 supported · 0 flagged · accuracy not yet measured

  • Premisecitation
    § Conclusion

    Unfamiliar neutral targets are already perceived as substantially more biased than the self.

    “Moreover, because unfamiliar neutral targets are already perceived as substantially more biased than the self (Mandel et al., 2022; Pronin et al., 2002), both unfamiliarity and negative affect may lead to similar conclusions—that a target other is considerably more biased than oneself.”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites Mandel et al., 2022 source

      The Bias blind spot section reports that participants rated the average survey respondent as more susceptible to bias than themselves, with a medium overall effect. But the target was an average respondent, not a specifically unfamiliar, neutral individual; the Testing explanatory hypotheses section describes comparisons with specific others as a possibility for future research. The source therefore supports the self–other bias difference, but not that distinction about the target. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites Pronin et al., 2002 source
  • Premisecitation
    § Results and Discussion

    Prior research has documented individuals' bias blind spots in non-U.S. populations.

    “prior research has documented individuals’ bias blind spots (Chandrashekar et al., 2021; Niszczota et al., 2023) and reliance on an affect heuristic (Finucane et al., 2000; Keller et al., 2006) in non-U.S. populations.”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites Chandrashekar et al., 2021 source
    • cites Niszczota et al., 2023

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Findingown data
    § Abstract

    In all experiments, the effect of other target selection on the bias blind spot was mediated by affective evaluations.

    “In all experiments, the effect of other target selection on the BBS was mediated by affective evaluations.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Participants' affective evaluations of target others are causally prior to their attributions of bias, rather than being influenced by those attributions.
  • Findingown data
    § Results and Discussion

    Participants did not display a bias blind spot when evaluating the biases of a liked well-known other (M = −0.02, SD = 0.83), t(209) = −0.27, p = .784, d = −0.02, 95% CI [−0.15, 0.12].

    “Conversely, participants did not display a bias blind spot when evaluating the biases of a liked well-known other (M = −0.02, SD = 0.83), t(209) = −0.27, p = .784, d = −0.02, [−0.15, 0.12], as they extended favorable attributions of objectivity to these individuals.”
    Quote in the paperSubstantial claim
  • Findingown data
    § Results and Discussion

    Considerably less bias was attributed to liked well-known others (M = 3.50, SD = 0.93) compared to the average survey respondent (M = 4.72, SD = 0.77), t(422) = 14.71, p < .001, d = 1.43, 95% CI [1.21, 1.64].

    “considerably less bias was attributed to liked well-known others (M = 3.50, SD = 0.93) compared to disliked well-known others (M = 4.69, SD = 0.89), t(417) = 13.25, p < .001, d = 1.29, [1.08, 1.51], and the average survey respondent (M = 4.72, SD = 0.77), t(422) = 14.71, p < .001, d = 1.43, [1.21, 1.64].”
    Quote in the paperSubstantial claim
  • Classified as probably not a substantial claim. 28
  • May combine two claims. 28

  • Classified as probably not a substantial claim.

    Study design and conductcitation
    § Method

    The Experiment 1 sample size of 660 was chosen based on prior work and sensitivity power analyses showing at least 80% power to detect a minimum effect size of = .113 for mixed ANOVAs and d = 0.28 for follow-up independent samples t-tests.

    “This sample size was chosen on the basis of prior work on the bias blind spot (Mandel et al., 2022) with sensitivity power analyses using G*Power (Faul et al., 2007) revealing that this sample provided at least 80% power to detect a minimum effect size of = .113 for the conducted mixed ANOVAs and d = 0.28 for follow-up independent samples t-tests.”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites Mandel et al., 2022 source
    • cites Faul et al., 2007 source

  • May combine two claims.

    Premisecitation
    § Introduction

    Democrats and Republicans have negative feelings toward one another and positive sentiments toward co-partisans.

    “In Experiments 2 and 3, we leveraged Democrats’ and Republicans’ negative feelings toward one another and positive sentiments toward co-partisans (Iyengar et al., 2019) to assess the role of affect in attributions of bias within a polarized group context.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Iyengar et al., 2019 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data
    § Method

    The political identity strength scale showed excellent internal consistency (α = .93) and was aggregated to form a composite measure.

    “Scale items showed excellent internal consistency (α = .93) and thus were aggregated to form a composite measure of political identity strength.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Leach et al., 2008 source

      The available indexed passages contain bibliographic information for Leach et al. (2008), but none reports an α of .93 or aggregation into a political identity strength composite. Searches for political identity, internal consistency, and the stated alpha found no passage addressing those claims. (LLM, conf=0.72)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data
    § Abstract

    In Experiment 1, participants viewed themselves as considerably less biased than the average survey respondent and a personally-known disliked other, but not less biased than a familiar individual whom they liked.

    “In Experiment 1, participants viewed themselves as considerably less biased than the average survey respondent and a personally-known disliked other, but not less biased than a familiar individual whom they liked.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Abstract

    In Experiments 2 and 3, participants judged themselves as somewhat less biased than co-partisans but much less biased than their political opponents.

    “While participants judged themselves as somewhat less biased than co-partisans, they viewed themselves as much less biased than their political opponents.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results and Discussion

    The negative association between affective evaluations and bias blind spot was observed within each experimental group: average survey respondent r(212) = −.23, p < .001; liked well-known other r(208) = −.18, p = .010; disliked well-known other r(207) = −.31, p < .001.

    “Notably, this negative association was observed not only across conditions but also within each experimental group: Average survey respondent, r(212) = −.23, p < .001; Liked well-known other, r(208) = −.18, p = .010; Disliked well-known other, r(207) = −.31, p < .001.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results and Discussion

    Participants evaluating the average survey respondent (M = 1.00, SD = 0.97) or a disliked well-known other (M = 1.25, SD = 1.04) displayed large bias blind spots (p < .001 and d > 1.02 in both cases).

    “Participants evaluating the biases of the average survey respondent (M = 1.00, SD = 0.97) or a disliked well-known other (M = 1.25, SD = 1.04) displayed large bias blind spots (p < .001 and d > 1.02 in both cases).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results and Discussion

    In Experiment 2, participants viewed themselves as less biased than the average member of their political in-group, but displayed a larger bias blind spot when attributing bias to their political opponents.

    “while participants viewed themselves as less biased than the average member of their political in-group (p < .001, d = 0.41), they displayed a larger bias blind spot when attributing bias to their political opponents (p < .001, d = 0.75).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Method

    Aggregated responses across the seven selected items revealed no significant partisan differences in bias categorization (p = .605, d = 0.10) or affect judgments (p = .091, d = 0.34).

    “Aggregated responses across these seven items revealed no significant partisan differences in bias categorization (p = .605, d = 0.10) or affect judgments (p = .091, d = 0.34; Figure 2B).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results and Discussion

    In Experiment 3, Democrats displayed larger bias blind spots than Republicans when their self-attributions of bias were compared to their attributions of bias toward the average out-group, but not in-group member.

    “As a result, Democrats displayed larger bias blind spots than Republicans when their self-attributions of bias were compared to their attributions of bias toward the average out-group (p < .001, d = 0.38), but not in-group member (p = .651, d = 0.04).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results and Discussion

    Participants perceiving more negative feelings between Democrats and Republicans attributed more bias to the average member of their political out-group and exhibited larger bias blind spots.

    “participants perceiving more negative feelings between Democrats and Republicans attributed more bias to the average member of their political out-group and exhibited larger bias blind spots.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Participants' self-reported affective evaluations accurately reflect their actual affective reactions to the targets.

  • May combine two claims.

    Findingown data
    § Results and Discussion

    While a majority of participants endorsed a behavioral observation strategy when assessing the biases of a target other in Experiment 1, this preference was not observed in Experiment 2.

    “Likewise, while a majority of participants endorsed a behavioral observation strategy when assessing the biases of a target other in Experiment 1, this preference was not observed in Experiment 2.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results and Discussion

    Reliance on introspective evidence was weakly correlated with lower attributions of bias in Experiment 2, with no such relation being observed in Experiment 1.

    “Additionally, reliance on introspective evidence was weakly correlated with lower attributions of bias in Experiment 2, with no such relation being observed in Experiment 1.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Introduction

    Psychologists have documented numerous biases that distort human judgment in social and cognitive domains.

    “Psychologists have documented numerous biases that distort human judgment, both within the social (Nisbett & Ross, 1980; Nisbett & Wilson, 1977) and cognitive domain (Tversky & Kahneman, 1974).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Nisbett & Ross, 1980
    • cites Nisbett & Wilson, 1977 source
    • cites Tversky & Kahneman, 1974 source

  • May combine two claims.

    Premisecitation
    § Introduction

    Viewing oneself as objective while accusing others of bias can promote negative interpersonal interactions, facilitating distrust and impeding conflict resolution.

    “Furthermore, viewing oneself as objective while accusing others of bias can promote negative interpersonal interactions, facilitating feelings of distrust and impeding conflict resolution (Kennedy & Pronin, 2008; Pronin, 2008).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Kennedy & Pronin, 2008 source
    • cites Pronin, 2008 source

  • Classified as probably not a substantial claim.

    Premisenone given
    § Introduction

    Much of the literature surrounding the bias blind spot has investigated the tendency for people to rate themselves as less biased than a hypothetical average other.

    “Much of the literature surrounding the bias blind spot has investigated the tendency for people to rate themselves as less biased than a hypothetical “average other” (e.g., the average American).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisereasoning
    § Introduction

    The self-other asymmetry in strategies for assessing bias has been theorized to produce self-other asymmetries in attributions of bias that reflect a bias blind spot.

    “Due to the presumed nonconscious nature of many biases, this self-other asymmetry in strategies for assessing bias has been theorized to produce self-other asymmetries in attributions of bias that reflect a bias blind spot.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § Method

    People are less likely to acknowledge exhibiting biases that they perceive as socially undesirable.

    “People are less likely to acknowledge exhibiting biases that they perceive as socially undesirable (Pronin et al., 2002).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Pronin et al., 2002 source

  • Classified as probably not a substantial claim.

    Premisecitation
    § Introduction

    Much evidence demonstrates the significant role of affect in human judgment.

    “Much evidence demonstrates the significant role of affect in human judgment (Lerner et al., 2015; Loewenstein et al., 2001; Slovic et al., 2007).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Lerner et al., 2015 source
    • cites Loewenstein et al., 2001 source
    • cites Slovic et al., 2007 source

  • May combine two claims.

    Premisecitation
    § Introduction

    Believing that one objectively perceives reality leads people to assume others' beliefs will correspond with their own objective assessments; when this assumption is false, people tend not to question their own objectivity but attribute lack of information or bias to those with opposing views.

    “Believing that we objectively perceive reality leads us to assume that the beliefs of others will correspond with our own “objective” assessments. When this assumption is proven false, people tend not to question the objectivity of their judgments, instead attributing a lack of information or even bias to individuals with opposing views.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Griffin & Ross, 1991 source
    • cites Ross & Ward, 1996

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data
    § Method

    Data from 27 participants were excluded based on pre-registered exclusion criteria, leaving data from 633 participants.

    “We excluded data from 27 participants based on a set of pre-registered exclusion criteria (see “Data Preparation” below), leaving data from 633 participants (49% Female; Mage = 38.76, SDage = 11.37; 81% British, 7% Canadian, 3% American, 9% Other; 78% College educated).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data
    § Method

    The 14-item bias blind spot measure demonstrated good internal consistency (α = .82), supporting item aggregation.

    “Consistent with prior work (Scopelliti et al., 2015), this 14-item measure demonstrated good internal consistency (α = .82), supporting item aggregation.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Scopelliti et al., 2015 source

  • Classified as probably not a substantial claim.

    Premisecitation
    § Introduction

    Bias blind spots have been attributed to two complementary mechanisms: belief in objective perception of reality (naïve realism) and self-other asymmetries in the value given to introspective evidence.

    “Bias blind spots have been attributed to two complementary mechanisms (Pronin, 2007, 2008; Pronin et al., 2004): (a) people’s belief that they objectively perceive reality (i.e., naïve realism; Griffin & Ross, 1991; Ross & Ward, 1996) and (b) self-other asymmetries in the value1 given to introspective evidence (Pronin & Kugler, 2007).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Pronin, 2007 source
    • cites Pronin, 2008 source
    • cites Pronin et al., 2004 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given
    § Method

    All experiments were pre-registered through Open Science Framework.

    “All experiments were pre-registered through Open Science Framework.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § Introduction

    The cognitive processes that engender psychological biases are thought to be nonconscious and inaccessible to introspection.

    “The cognitive processes that engender psychological biases are thought to be nonconscious and, as such, inaccessible to introspection (Wilson & Brekke, 1994; Wilson et al., 2002).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Wilson & Brekke, 1994 source
    • cites Wilson et al., 2002 source

  • May combine two claims.

    Premisecitation
    § Introduction

    Affective reactions are generated quickly and automatically and guide how people attend to, encode, and interpret various events.

    “Affective reactions are generated quickly and automatically (Bargh et al., 1992; Zajonc, 1980), and guide how people attend to, encode, and interpret various events (Baumeister & Newman, 1994).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Bargh et al., 1992

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Zajonc, 1980 source
    • cites Baumeister & Newman, 1994 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data
    § Method

    Data from 34 participants were excluded based on pre-registered exclusion criteria, leaving data from 626 participants to be analyzed.

    “We excluded data from 34 participants based on pre-registered exclusion criteria, leaving data from 626 participants (51% Female; Mage = 39.77, SDage = 11.12; 50% Democrat; 68% College educated) to be analyzed.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data
    § Method

    The resulting 7-item, non-partisan bias blind spot measure showed acceptable internal consistency (α = .75), supporting item aggregation.

    “Furthermore, the resulting 7-item, non-partisan bias blind spot measure showed acceptable internal consistency (α = .75), supporting item aggregation.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § Results and Discussion

    In Experiment 2, affect contributes to intergroup differences in bias attribution and also shapes judgments within group boundaries.

    “Taken together, these findings suggest that while affect contributes to intergroup differences in bias attribution (consistent with in-group favoritism), it also shapes judgments within group boundaries.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § Results and Discussion

    In Experiment 2, political identity strength was linked to viewing in-group members as less biased, rather than strongly identified partisans attributing more bias to political opponents.

    “In sum, rather than strongly identified partisans attributing more bias to political opponents and exhibiting larger bias blind spots, political identity strength was linked to viewing in-group members as less biased—mirroring how individuals viewed themselves.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § Results and Discussion

    Research on affective polarization shows that negative feelings toward political out-groups shape individuals' political attitudes and behaviors.

    “Research on affective polarization shows that negative feelings toward political out-groups shape individuals’ political attitudes and behaviors (Abramowitz & Webster, 2016; Finkel et al., 2020; Kingzette et al., 2021).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Abramowitz & Webster, 2016 source
    • cites Finkel et al., 2020 source
    • cites Kingzette et al., 2021 source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationreasoning
    § Results and Discussion

    Increased perceptions of bias across political lines may diminish individuals' trust in bipartisan institutions and lead people to disengage from democratic processes.

    “Similarly, increased perceptions of bias across political lines may diminish individuals’ trust in bipartisan institutions and lead people to disengage from democratic processes.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning
    § Results and Discussion

    Rising partisan animosity may influence how people consume political information and participate in politics.

    “Rising partisan animosity, insofar as it increases the extent to which partisans view their political opponents as biased, may also influence how people consume political information and participate in politics.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning
    § Results and Discussion

    The methodology was high in ecological validity, taking advantage of participants' naturally occurring affective impressions.

    “This methodology was high in ecological validity, taking advantage of participants’ naturally occurring affective impressions.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning
    § Results and Discussion

    Rising partisan animosity may amplify attributions of bias across political divides, which can escalate conflict and lead to even greater partisan animus in the future.

    “Given peoples’ proclivity to attribute bias more readily to individuals they dislike, rising partisan animosity may amplify attributions of bias across political divides which, as we have discussed, can escalate conflict and lead to even greater partisan animus in the future.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning
    § Conclusion

    Initial feelings of dislike can facilitate the perception that a disliked individual or group is biased, encouraging conflict-escalating actions and hindering compromise.

    “They also have practical implications, suggesting that initial feelings of dislike can facilitate the perception that a disliked individual or group is biased, encouraging conflict-escalating actions and hindering compromise.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The observed relation between affective evaluations and attributions of bias is not due to extraneous variables (e.g., disagreement) on which the target others also differed.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationreasoning
    § Results and Discussion

    Perceiving political out-group members as biased can lead partisans to self-select into political echo chambers and disengage from political discussions that feature individuals with opposing views.

    “For example, perceiving political out-group members as biased can lead partisans to self-select into political echo chambers and disengage from political discussions that feature individuals with opposing views.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data
    § Results and Discussion

    Partisans who harbor strong animosity toward their political opponents may not only distrust their judgments in political contexts but also question their objectivity more broadly.

    “Thus, partisans who harbor strong animosity toward their political opponents may not only distrust their judgments in political contexts (e.g., their assessment of political events) but also question their objectivity more broadly.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given
    § Results and Discussion

    The results do not depend on other characteristics of the participants, materials, or context.

    “We have no reason to believe that the results depend on other characteristics of the participants, materials, or context.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Results and Discussion

    Much work has documented the rise of affective polarization and partisan animosity in the United States and other countries.

    “Much work has documented the rise of affective polarization and partisan animosity in the United States (Finkel et al., 2020) and other countries (Boxell et al., 2022).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Finkel et al., 2020 source
    • cites Boxell et al., 2022

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim.

    Premisecitation
    § Results and Discussion

    The cognitive processes that produce many psychological biases are inaccessible to introspection.

    “Given that the cognitive processes that produce many psychological biases are inaccessible to introspection (Wilson & Brekke, 1994; Wilson et al., 2002)”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Wilson & Brekke, 1994 source
    • cites Wilson et al., 2002 source

  • May combine two claims.

    Premisecitation
    § Results and Discussion

    Affective impressions are generated quickly and automatically and play a central role in human judgment.

    “Affective impressions are generated quickly and automatically (Bargh et al., 1992; Zajonc, 1980) and play a central role in human judgment (Lerner et al., 2015; Loewenstein et al., 2001; Slovic et al., 2007).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Bargh et al., 1992

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Zajonc, 1980 source
    • cites Lerner et al., 2015 source

  • Probably not a substantial claim: may not pass S1 (asserted).

    Premisecitation
    § Results and Discussion

    A prominent account of the bias blind spot states that this meta-bias results, at least in part, from people's tendency to value their own introspective evidence, while dismissing the introspective evidence of others.

    “A prominent account of the bias blind spot states that this meta-bias results, at least in part, from people’s tendency to value their own introspective evidence, while dismissing the introspective evidence of others (Pronin, 2007, 2008).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Pronin, 2007, 2008

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim.

    Premisecitation
    § Results and Discussion

    Distrusting the introspections of others, people instead focus on others' behavior when assessing their level of bias.

    “Distrusting the introspections of others, people instead focus on others’ behavior when assessing their level of bias (Pronin et al., 2004).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Pronin et al., 2004 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given
    § Conclusion

    APA ethical standards were followed in the conduct of all studies, and informed consent was obtained from all participants after an explanation of the nature and possible consequences of the study.

    “APA ethical standards were followed in the conduct of all studies, and we obtained informed consent from all participants following an explanation of the nature and possible consequences of the study they participated in.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning
    § Results and Discussion

    Interventions that target individuals' "hot" affective reactions may help attenuate the large bias blind spots that characterize social contexts marked by animosity.

    “Accordingly, interventions that target individuals’ “hot” affective reactions—as opposed to their “cold” cognitive strategies—may help attenuate the large bias blind spots that characterize social contexts marked by animosity.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationreasoning
    § Results and Discussion

    Such negative feedback loops may be most pertinent in group contexts characterized by dislike and disagreement, such as political contexts featuring competing groups with distinct values.

    “Such negative feedback loops may be most pertinent in group contexts characterized by dislike and disagreement, such as political contexts featuring competing groups with distinct values.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation
    § Results and Discussion

    This attribution of bias can then encourage individuals to engage in conflict-escalating actions that, over time, lead to stronger negative affect and consequently, greater attributions of bias.

    “This attribution of bias can then encourage individuals to engage in conflict-escalating actions (as has been shown in prior work; Kennedy & Pronin, 2008; Pronin et al., 2006) that, over time, lead to stronger negative affect and consequently, greater attributions of bias.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Kennedy & Pronin, 2008 source
    • cites Pronin et al., 2006 source

  • Probably not a substantial claim: may not pass S1 (asserted).

    Premisereasoning
    § Results and Discussion

    This self-other asymmetry in strategies for assessing bias has been theorized to produce the self-other asymmetry in attributions of bias that characterize the bias blind spot.

    “this self-other asymmetry in strategies for assessing bias has been theorized to produce the self-other asymmetry in attributions of bias that characterize the bias blind spot.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation
    § Conclusion

    Political partisans tend to overestimate the amount of dislike rival partisans feel toward them, with negative meta-perceptions being linked with ideological extremity and anti-democratic attitudes.

    “Specifically, this research finds that political partisans tend to overestimate the amount of dislike rival partisans feel toward them, with negative meta-perceptions being linked with ideological extremity and anti-democratic attitudes (Lees & Cikara, 2020; Moore-Berg et al., 2020).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Lees & Cikara, 2020

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Moore-Berg et al., 2020 source
50 of 50

  • The paper relies on this without support.

    The 14 psychological biases described in Experiment 1 are perceived by participants as bias-like and socially undesirable.

    “Participants completed a bias blind spot task adapted from Scopelliti et al. (2015). For this task, they were presented with descriptions of 14 biases (see Table 2) and asked to evaluate the extent to which they (Self block) and a target other (Other block) exhibited each bias.”

  • The paper relies on this without support.

    Participants' affective evaluations of target others are causally prior to their attributions of bias, rather than being influenced by those attributions.

    “In all experiments, the effect of other target selection on the BBS was mediated by affective evaluations.”

  • The paper relies on this without support.

    Participants' self-reported bias assessment strategies accurately reflect the strategies they actually used to assess bias.

    “Using a methodology adapted from prior work (Pronin & Kugler, 2007),4 we assessed the extent to which participants endorsed using different strategies when judging the biases of themselves and others.”

  • The paper relies on this without support.

    The average survey respondent is a valid and meaningful comparison target for the 'average other' in bias blind spot judgments.

    “In Experiment 1, participants judged the extent to which they and either a liked well-known other, a disliked well-known other, or the average survey respondent exhibited various psychological biases.”

  • The paper relies on this without support.

    The mediation analyses assume that the relationship between affective evaluations and bias blind spots is not confounded by unmeasured variables and that affective evaluations temporally precede bias attributions.

    “These analyses revealed that the effect of Other Reference Group on bias blind spot scores was partially mediated by participants’ affective evaluations of target others (b = 0.62, 95% CI [0.40, 0.80]; see Supplemental Materials Part C), suggesting that heightened attributions of bias toward political out-group members were driven, in part, by participants’ affective feelings toward each party.”

  • The paper relies on this without support.

    Participants' self-reported affective evaluations accurately reflect their actual affective reactions to the targets.

    “This methodology was high in ecological validity, taking advantage of participants’ naturally occurring affective impressions.”

  • The paper relies on this without support.

    The bias non-specific items and the specific bias blind spot items measure the same underlying construct of bias attribution.

    “Responses to items asking participants how biased they view themselves, the average Democrat, and the average Republican (bias non-specific), mirrored responses provided within the bias blind spot task (which focused on seven specific psychological biases).”

  • The paper relies on this without support.

    Political in-group and out-group categorization based on party affiliation captures meaningful social groups for participants.

    “Among participants assigned to evaluate the biases of the average out-group member, political identity strength was not correlated with bias blind spot scores”

  • The paper relies on this without support.

    The observed relation between affective evaluations and attributions of bias is not due to extraneous variables (e.g., disagreement) on which the target others also differed.

    “However, target others may have differed not only with regards to participants’ affective evaluations, but in additional ways that influenced bias attributions.”
9 of 9

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Consistentt(632) = 17.14reported p < .001 · recomputed p < .001

    “t(632) = 17.14, p < .001”

  • ConsistentF(2, 630) = 104.63reported p < .001 · recomputed p < .001

    “F(2, 630) = 104.63, p < .001”

  • Consistentt(417) = 13.25reported p < .001 · recomputed p < .001

    “t(417) = 13.25, p < .001”

  • Consistentt(422) = 14.71reported p < .001 · recomputed p < .001

    “t(422) = 14.71, p < .001”

  • Consistentt(421) = 2.58reported p = .01 · recomputed p = .010

    “t(421) = 2.58, p = .010”

  • Consistentt(209) = -0.27reported p = .784 · recomputed p = .787

    “t(209) = −0.27, p = .784”

  • Consistentr(631) = -0.52reported p < .001 · recomputed p < .001

    “r(631) = −.52, p < .001”

  • Consistentr(212) = -0.23reported p < .001 · recomputed p < .001

    “r(212) = −.23, p < .001”

  • Consistentr(208) = -0.18reported p = .01 · recomputed p = .009

    “r(208) = −.18, p = .010”

  • Consistentr(207) = -0.31reported p < .001 · recomputed p < .001

    “r(207) = −.31, p < .001”

  • Consistentr(631) = -0.5reported p < .001 · recomputed p < .001

    “r(631) = −.50, p < .001”

  • Consistentr(417) = -0.46reported p < .001 · recomputed p < .001

    “r(417) = −.46, p < .001”

  • Consistentr(417) = -0.52reported p < .001 · recomputed p < .001

    “r(417) = −.52, p < .001”

  • Consistentr(631) = -0.1reported p = .013 · recomputed p = .012

    “r(631) = −.10, p = .013”

  • Consistentz = -7.52reported p < .001 · recomputed p < .001

    “z = −7.52, p < .001”

  • Consistentt(580) = 3.05reported p = .002 · recomputed p = .002

    “t(580) = 3.05, p = .002”

  • Consistentt(625) = 19.06reported p < .001 · recomputed p < .001

    “t(625) = 19.06, p < .001”

  • ConsistentF(1, 622) = 152.26reported p < .001 · recomputed p < .001

    “F(1, 622) = 152.26, p < .001”

  • Consistentt(624) = 1.57reported p = .116 · recomputed p = .117

    “t(624) = 1.57, p = .116”

  • Consistentt(624) = 13.91reported p < .001 · recomputed p < .001

    “t(624) = 13.91, p < .001”

  • Consistentt(624) = 12.31reported p < .001 · recomputed p < .001

    “t(624) = 12.31, p < .001”

  • ConsistentF(1, 622) = 2.65reported p = .104 · recomputed p = .104

    “F(1, 622) = 2.65, p = .104”

  • Consistentr(624) = -0.55reported p < .001 · recomputed p < .001

    “r(624) = −.55, p < .001”

  • Consistentr(624) = -0.47reported p < .001 · recomputed p < .001

    “r(624) = −.47, p < .001”

  • Consistentr(304) = -0.3reported p < .001 · recomputed p < .001

    “r(304) = −.30, p < .001”

  • Consistentr(318) = -0.23reported p < .001 · recomputed p < .001

    “r(318) = −.23, p < .001”

  • Consistentr(624) = -0.52reported p < .001 · recomputed p < .001

    “r(624) = −.52, p < .001”

  • Consistentr(624) = -0.11reported p = .006 · recomputed p = .006

    “r(624) = −.11, p = .006”

  • Consistentz = -8.21reported p < .001 · recomputed p < .001

    “z = −8.21, p < .001”

  • Consistentr(318) = 0.09reported p = .111 · recomputed p = .108

    “r(318) = .09, p = .111”

  • Consistentr(304) = -0.15reported p = .009 · recomputed p = .009

    “r(304) = −.15, p = .009”

  • Consistentt(303) = -2.29reported p = .023 · recomputed p = .023

    “t(303) = −2.29, p = .023”

  • Consistentt(303) = -2.28reported p = .024 · recomputed p = .023

    “t(303) = −2.28, p = .024”

  • Consistentt(594) = 6.18reported p < .001 · recomputed p < .001

    “t(594) = 6.18, p < .001”

  • Consistentr(600) < 0.01reported p = .935 · recomputed p = .807

    “r(600) < .01, p = .935”

  • Consistentr(601) = -0.16reported p < .001 · recomputed p < .001

    “r(601) = −.16, p < .001”

  • Consistentt(635) = 22.97reported p < .001 · recomputed p < .001

    “t(635) = 22.97, p < .001”

  • ConsistentF(1, 634) = 415.88reported p < .001 · recomputed p < .001

    “F(1, 634) = 415.88, p < .001”

  • Consistentt(635) = 19.91reported p < .001 · recomputed p < .001

    “t(635) = 19.91, p < .001”

  • ConsistentF(1, 634) = 9.1reported p = .003 · recomputed p = .003

    “F(1, 634) = 9.10, p = .003”

  • ConsistentF(1, 634) = 30.19reported p < .001 · recomputed p < .001

    “F(1, 634) = 30.19, p < .001”

  • Consistentt(634) = 6.41reported p < .001 · recomputed p < .001

    “t(634) = 6.41, p < .001”

  • Consistentt(634) = 3.93reported p < .001 · recomputed p < .001

    “t(634) = 3.93, p < .001”

  • Consistentr(634) = -0.35reported p < .001 · recomputed p < .001

    “r(634) = −.35, p < .001”

  • Consistentr(634) = -0.41reported p < .001 · recomputed p < .001

    “r(634) = −.41, p < .001”

  • Consistentr(634) = -0.29reported p < .001 · recomputed p < .001

    “r(634) = −.29, p < .001”

  • Consistentr(634) = -0.32reported p < .001 · recomputed p < .001

    “r(634) = −.32, p < .001”

  • Consistentr(634) = -0.33reported p < .001 · recomputed p < .001

    “r(634) = −.33, p < .001”

  • Consistentr(634) = -0.4reported p < .001 · recomputed p < .001

    “r(634) = −.40, p < .001”

  • Consistentr(634) = -0.22reported p < .001 · recomputed p < .001

    “r(634) = −.22, p < .001”

  • Consistentr(634) = -0.31reported p < .001 · recomputed p < .001

    “r(634) = −.31, p < .001”

  • Consistentr(634) = -0.07reported p = .069 · recomputed p = .078

    “r(634) = −.07, p = .069”

  • Consistentr(634) = -0.25reported p < .001 · recomputed p < .001

    “r(634) = −.25, p < .001”

  • Consistentr(634) = -0.24reported p < .001 · recomputed p < .001

    “r(634) = −.24, p < .001”

  • Consistentt(633) = -2.08reported p = .038 · recomputed p = .038

    “t(633) = −2.08, p = .038”

  • Consistentt(633) = -4.74reported p < .001 · recomputed p < .001

    “t(633) = −4.74, p < .001”

  • Consistentr(634) = 0.26reported p < .001 · recomputed p < .001

    “r(634) = .26, p < .001”

  • Consistentr(634) = 0.18reported p < .001 · recomputed p < .001

    “r(634) = .18, p < .001”

  • Consistentt(633) = 1.06reported p = .289 · recomputed p = .290

    “t(633) = 1.06, p = .289”

  • Consistentt(633) = -7.02reported p < .001 · recomputed p < .001

    “t(633) = −7.02, p < .001”

  • Consistentt(632) = 1.61reported p = .109 · recomputed p = .108

    “t(632) = 1.61, p = .109”

  • Consistentr(634) = 0.24reported p < .001 · recomputed p < .001

    “r(634) = .24, p < .001”

62 of 62

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Self-check results appear here once an independent check has confirmed them.

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Walker, A. C., Collins, R. N., Walker, H. E. K., Fugelsang, J. A., & Mandel, D. R.
Affiliations
  • Brown University (US)
  • Defence Research and Development Canada (CA)
  • HumanSystems Inc., Guelph, ON, Canada
Corresponding author
Walker, Alexander C.
ORCID
Venue
Personality and Social Psychology Bulletin
Publisher
SAGE Publications
Published
2026-05-23
Pages
1461672261439247-1461672261439247
ISSN
  • 0146-1672 (ISSN-L)
  • 1552-7433
DOI
10.1177/01461672261439247
OpenAlex
W4393319687
Semantic Scholar
5f9b2895724d6cc1827eba6fd87cb1ce30dfc866
Cited by
2
References
27
Citations by year
2025: 2
FWCI
0.00
Type
journal article
Language
en
Open access
Licence
Locations
  • Personality and Social Psychology Bulletin (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • OSF Preprints (OSF Preprints) (submittedVersion)
  • PsyArXiv (OSF Preprints) (submittedVersion)
Funders
Canadian Safety and Security Program (CSSP-2018-TI-2394)
Topics
  • Social Sciences › Sociology and Political Science › Social and Intergroup Psychology
  • Social Sciences › Sociology and Political Science › Names, Identity, and Discrimination Research
  • Social Sciences › Decision Sciences › General Decision Sciences › Decision-Making and Behavioral Economics
Keywords
bias blind spot, affective evaluation, political polarization, partisan bias, target selection, liking and disliking

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-23.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-23.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review