Predicting implicit attitudes with natural language data.

Large-scale language datasets and advances in natural language processing offer opportunities for studying people's cognitions and behaviors. We show how representations derived from language can be combined with laboratory-based word norms to predict implicit attitudes for diverse concepts. Our approach achieves substantially higher correlations than existing methods. We also show that our approach is more predictive of implicit attitudes than are explicit attitudes, and that it captures variance in implicit attitudes that is largely unexplained by explicit attitudes. Overall, our results shed light on how implicit attitudes can be measured by combining standard psychological data with large-scale language data. In doing so, we pave the way for highly accurate computational modeling of what people think and feel about the world around them.

All human social groups are human, but some are more human than others: A comprehensive investigation of the implicit association of "Human" to US racial/ethnic groups.

All human groups are equally human, but are they automatically represented as such? Harnessing data from 61,377 participants across 13 experiments (six primary and seven supplemental), a sharp dissociation between implicit and explicit measures emerged. Despite explicitly affirming the equal humanity of all racial/ethnic groups, White participants consistently associated Human (relative to Animal) more with White than Black, Hispanic, and Asian groups on Implicit Association Tests (IATs; experiments 1-4). This effect emerged across diverse representations of Animal that varied in valence (pets, farm animals, wild animals, and vermin; experiments 1-2). Non-White participants showed no such Human=Own Group bias (e.g., Black participants on a White-Black/Human-Animal IAT). However, when the test included two outgroups (e.g., Asian participants on a White-Black/Human-Animal IAT), non-White participants displayed Human=White associations. The overall effect was largely invariant across demographic variations in age, religion, and education but did vary by political ideology and gender, with self-identified conservatives and men displaying stronger Human=White associations (experiment 3). Using a variance decomposition method, experiment 4 showed that the Human=White effect cannot be attributed to valence alone; the semantic meaning of Human and Animal accounted for a unique proportion of variance. Similarly, the effect persisted even when Human was contrasted with positive attributes (e.g., God, Gods, and Dessert; experiment 5a). Experiments 5a-b clarified the primacy of Human=White rather than Animal=Black associations. Together, these experiments document a factually erroneous but robust Human=Own Group implicit stereotype among US White participants (and globally), with suggestive evidence of its presence in other socially dominant groups.

Implicit Sexual Cognitions in Women with Ambiphilic Sexual Attractions: A Comparison to Androphilic and Gynephilic Women.

Previous research using indirect cognitive measures (sometimes referred to as implicit measures) of sexual attraction have shown that women who are attracted to men (androphilic women) show category non-specific responses, whereas those who are attracted to women (gynephilic) show a category-specific bias to women. The purpose of the present study was to examine whether women who explicitly report approximately equal attraction to men and women (ambiphilic) would show similar non-category specific attraction at this implicit level or whether their responses would be more similar to those of gynephilic women. An implicit association task and a priming task were given to 169 women alongside measures of their self-labelled sexual orientation and an explicit measure of their sexual attraction to men and women. The results replicated previous findings of little bias towards either gender in androphilic women and of a strong bias towards females in gynephilic women. The ambiphilic women also showed a strong bias towards females. The findings clearly show that early automatic associations to sex are biased towards females in ambiphilic women and are not consistent with their explicit statements of preference.

Perceptions of older and younger adults who wear hearing aids.

OBJECTIVE: To investigate older and younger adults' perceptions of older and younger adults who wear hearing aids. DESIGN: Participants completed two Implicit Association Tests: One with images of older adults (OA-IAT) and one with images of younger adults (YA-IAT), either wearing or not wearing hearing aids. Participants also rated age, attractiveness, and intelligence of younger and older adults pictured with or without a hearing aid. STUDY SAMPLE: Thirty older adults (M age = 70 years, SD = 4.38) and 30 younger adults (M age = 23 years, SD = 3.01) who reported not having hearing aids or a diagnosed hearing impairment. RESULTS: For both IATs, older and younger participants responded faster and more accurately when images of individuals wearing hearing aids were paired with negative words in comparison to positive words. Photo ratings did not vary in relation to the presence or absence of hearing aids for either age group. CONCLUSION: Although the photo rating tasks indicate neutral explicit attitudes towards individuals who wear hearing aids, our interpretation of the IAT results indicates that younger and older adults may hold negative implicit attitudes towards both older and younger hearing aid users.

Using facial expressions instead of response keys in the implicit association test.

Previous research found that when people are instructed to smile toward liked objects and show negative facial expressions toward disliked objects, their facial response is faster and more intense than when they are required to smile toward disliked objects and express negative facial response toward liked objects. The present research tested a technologically innovative indirect evaluation measure that was based on that finding. Participants completed an implicit association test (IAT)-a common indirect measure of evaluation, responding with their emotional facial expressions, rather than by pressing response keys. In two web studies, using emotional facial expression detection through a webcam, we found that the Facial Response IAT (FR-IAT) is a reliable and valid measure of evaluations, comparable to the keyboard IAT. Because facial responses provide more information than key responses, pursuing future improvements of the FR-IAT's methodology, software, and data analysis is a promising direction for enhancing the quality of indirect evaluation measurement. The same methodology and technology may also enhance other indirect measures of evaluation and cognitive tests related to emotion and judgment.

The Course of Anxiety-Specific Cognitive Bias Following Daycare/Inpatient Treatment in Youths with Social Phobia and School Absenteeism.

Social phobia (SP) is a common mental disorder in youth often accompanied by absence from school, which may require daycare or inpatient intervention (DC/IN). Objective: The present explorative study investigates changes in anxiety-specific implicit assumptions and interpretation bias following DC/IN. Methods: The study included 16 youths with SP (M age = 15.8 [SD = 1.24], females: 62.5 %) participating in DC/IN. We assessed the main outcomes using the Implicit Association Test and Affective Misattribution Procedure. Results: A large effect was shown for reducing implicit assumptions of feeling anxious (p = .142; η2p = .171) and for reducing the implicit interpretation bias (p = .137; η2p = .162). No change was indicated by effect size in implicit assumptions of feeling socially rejected (p = .649; η2p = .016). Social phobia symptoms initially correlated with changes in implicit assumptions of feeling anxious (r = .45). Conclusion: Effect sizes indicate that implicit anxiety-specific assumptions and interpretation bias descriptively improved following DC/IN. Thus, DC/IN may lead to meaningful improvements of anxiety-specific cognition in some individuals with high SP symptoms, emphasizing the relevance of cognitive behavioral approaches in the treatment of SP. Several limitations are discussed.

Implicit racial bias among dental hygienists practicing in the United States.

PURPOSE: Provider bias has been shown to be a contributing factor to racial and ethnic disparities observed in health care settings. The purpose of this study was to examine implicit racial bias among dental hygienists. METHODS: A convenience sample of licensed and practicing dental hygienists within the United States was recruited through email and national dental hygiene social media groups via snowball sampling. A two-part survey design was used for data collection. Participants completed a 10-item demographic survey through an online platform and were then routed to the Race Implicit Association Test (IAT). Descriptive statistics and linear regression analyses were used to compare demographic data and implicit racial preference scores (d-scores). RESULTS: Data from 404 licensed dental hygienists were included in this study. Over two-thirds (67.8%) of participants showed a preference for European Americans over African Americans. A significant difference was found between implicit racial preference scores and participant age (Estimate: 0.01, 95% CI: 0.00, 0.01), years worked comparing <5 years to 21 or more years (Estimate: 0.19, 95% CI: -0.30, -0.09) and race comparing non-White to White (Estimate: -0.17, 95% CI: -0.27, -0.07). No difference was found with task order, previous Race IAT experience, or previous self-reported implicit bias training. CONCLUSIONS: Findings suggest that dental hygienists may harbor implicit racial preferences for European Americans over African Americans. Non-White participants had more positive implicit preferences toward African Americans compared to White participants. Further research is needed to determine the extent to which implicit racial biases contribute to disparities in oral health.

Right amygdala and caudate activation patterns predict implicit attitudes toward people with autism spectrum disorders and physical disabilities, respectively.

Attitudes toward people with disabilities tend to be negative, regardless of the visibility of the disorder traits. Individuals with autism spectrum disorder (ASD) experience stigmatization that is rooted in negative attitudes or prejudice toward them because of their social awkwardness. The neural underpinnings of attitudes toward people with disabilities remain unclear. In this study, we focused on implicit attitudes toward people with ASD and physical disabilities, which are more visible than ASD, and investigated whether implicit attitudes were predicted by using neural activity with multivoxel pattern analysis (MVPA) in the prejudice network. Thirty-six, right-handed, Japanese, university students without disabilities participated. Two implicit association tests (IAT) toward people with ASD and physical disabilities revealed negative implicit attitudes. In the MRI scanner, participants performed a one-back task by using the same picture sets of IATs to examine their neural responses toward people with ASD and physical disabilities. According to the MVPA results, activation patterns of the right amygdala and right caudate significantly predicted implicit attitudes toward people with ASD and physical disabilities, respectively. These results suggest that implicit attitudes toward ASD and physical disabilities can be predicted by using neural signals from different regions within the prejudice network.

Regional Measures of Sexual-Orientation Bias Predict Where Same-Gender Couples Live.

Regional explicit and implicit bias are associated with real-world discrimination and marginalization. We extended this research area by focusing on sexual minorities and where same-gender couples live. Using data on 2,939 U.S. counties from Project Implicit and other publicly available sources, we found that measures with known associations with systemic anti-lesbian, gay, and bisexual (anti-LGB) bias are similarly associated with regional implicit and explicit anti-LGB bias. Furthermore, we found that fewer same-gender couples reside in counties with more explicit and implicit anti-LGB bias, above and beyond other factors that likely influence same-gender-couple residency. These findings further suggest that explicit and implicit measures of regional bias are capturing similar, if not the same, construct of a region's culture of bias toward particular groups. Couched specifically within the ongoing systemic political antagonization of the lesbian, gay, bisexual, transgender, queer, plus (LGBTQ+) community, these findings also highlight the importance of considering contextual (in addition to individual) factors that reinforce systemic inequality.

Oppressed Groups Engender Implicit Positivity: Seven Demonstrations Using Novel and Familiar Targets.

Across seven preregistered studies in online adult volunteer samples (N = 5,323), we measured implicit evaluations of social groups following exposure to historical narratives about their oppression. Although the valence of such information is highly negative and its interpretation was left up to participants, implicit evaluations of oppressed groups shifted toward positivity, including in designs involving fictitious, well-known, and even self-relevant targets. The sole deviation from this pattern was observed in an experiment using a vignette about slavery in the United States, in response to which neither White nor Black Americans exhibited any change in implicit race attitudes. In line with propositional perspectives, these findings suggest that implicit evaluations (including, notably, implicit evaluations of well-known and self-relevant social groups) tend to change toward positivity in response to extremely negative information involving past oppression. However, macro-level phenomena, such as public awareness of histories of oppression, can modulate such updating processes.

A comparison of self-reported risk and protective factors and the death implicit association test in the prediction of future suicide attempts in adolescent emergency department patients.

BACKGROUND: Concerns have been raised about the utility of self-report assessments in predicting future suicide attempts. Clinicians in pediatric emergency departments (EDs) often are required to assess suicidal risk. The Death Implicit Association Test (IAT) is an alternative to self-report assessment of suicidal risk that may have utility in ED settings. METHODS: A total of 1679 adolescents recruited from 13 pediatric emergency rooms in the Pediatric Emergency Care Applied Research Network were assessed using a self-report survey of risk and protective factors for a suicide attempt, and the IAT, and then followed up 3 months later to determine if an attempt had occurred. The accuracy of prediction was compared between self-reports and the IAT using the area under the curve (AUC) with respect to receiver operator characteristics. RESULTS: A few self-report variables, namely, current and past suicide ideation, past suicidal behavior, total negative life events, and school or social connectedness, predicted an attempt at 3 months with an AUC of 0.87 [95% confidence interval (CI), 0.84-0.90] in the entire sample, and AUC = 0.91, (95% CI 0.85-0.95) for those who presented without reported suicidal ideation. The IAT did not add significantly to the predictive power of selected self-report variables. The IAT alone was modestly predictive of 3-month attempts in the overall sample ((AUC = 0.59, 95% CI 0.52-0.65) and was a better predictor in patients who were non-suicidal at baseline (AUC = 0.67, 95% CI 0.55-0.79). CONCLUSIONS: In pediatric EDs, a small set of self-reported items predicted suicide attempts within 3 months more accurately than did the IAT.

Is the performance at the implicit association test sensitive to feedback presentation? A Rasch-based analysis.

The Implicit Association Test (IAT) is commonly used for the indirect assessment of psychological constructs. While the features of the IAT that might influence the performance of the respondents have been extensively investigated, the effect of informing the respondents about the correctness of their responses (i.e., feedback presentation) has been poorly addressed so far. The study addresses this issue by presenting an across-domain (implicit prejudice and food preference) Rasch-based analysis of IAT data obtained with and without feedback presentation. Results showed that speed was influenced by the interaction between feedback presentation and associative condition, whereas accuracy was influenced by the associative condition. This result varied across-domain. Results suggested that IATs administered with feedback presentation provide more accurate information on the construct of interest.

Implicit assumptions and interpretation bias in youth with severe, chronic social phobia.

Interpretation bias and dysfunctional social assumptions are proposed to play a pivotal role in the development and maintenance of social phobia (SP), especially in youth. In this study, we aimed to investigate disorder-specific implicit assumptions of rejection and implicit interpretation bias in youth with severe, chronic SP and healthy controls (CG). Twenty-seven youth with SP in inpatient/day-care treatment (M age = 15.6 years, 74% female) and 24 healthy controls (M age = 15.7 years, 54% female) were included. The Implicit Association Test (IAT) and the Affect Misattribution Procedure (AMP) were completed to assess implicit assumptions and interpretation bias related to the processing of social and affective stimuli. No group differences were observed for the IAT controlling for depressive symptoms in the analyses. However, group differences were found regarding interpretation bias (p = .017, η2p = .137). Correlations between implicit scores and explicit questionnaire results were medium to large in the SP group (r =|.28| to |.54|, pall ≤ .05), but lower in the control group (r =|.04| to |.46|, pall ≤ .05). Our results confirm the finding of an interpretation bias in youth SP, especially regarding the implicit processing of faces, whereas implicit dysfunctional social assumptions of being rejected do not seem to be specific for SP. Future research should investigate the causal relationship of assumptions/interpretation bias and SP.

Discovering the Roots: A Qualitative Analysis of Medical Students Exploring Their Unconscious Obesity Bias.

Phenomenon: Bias against individuals with obesity in medical settings has negative implications for patients, including stigmatization, poor health outcomes, and reduced healthcare utilization. This study explored reflections of medical students when confronted with their own implicit obesity bias.Approach: A group of 188 pre-clinical second-year medical students from George Washington University School of Medicine and Health Sciences completed the Weight Implicit Association Test (IAT) in 2020 and were instructed to write a reflective response based on their results. Participants reflected upon their preferences ("fat" vs. "thin") and described the factors that influenced their perceptions of obesity. Inductive coding techniques were used to generate themes from medical students' responses using Dedoose Version 8.3.35 (SocioCultural Research Consultants LLC, Los Angeles, California).Findings: Regarding IAT results, 7% of medical students preferred "fat over thin," 14% had no preference, and 78% preferred "thin over fat." Reflection themes highlighted medical students' difficulty accepting IAT results, perspectives on the origins of obesity in individuals, personal and family challenges with obesity and body image, medical training's perceived influence on bias, reservations about discussing obesity with patients, and desires to change current and future practices.Insights: Many medical students expressed a desire to provide optimal care for patients of all weight classes despite demonstrating a strong unconscious bias against individuals with obesity on the IAT. Medical school should provide targeted opportunities to acknowledge and mitigate obesity bias by expanding on medical students' pre-established and often harmful understandings of obesity and highlighting the complexities of this disease. Such training would better equip medical students to facilitate successful interactions with patients as future physicians.

We like it 'cause you take it: vicarious effects of approach/avoidance behaviours on observers.

We present five studies investigating the effects of approach and avoidance behaviours when individuals do not enact them but, instead, learn that others have performed them. In Experiment 1, when participants read that a fictitious character (model) had approached a previously unknown product, they ascribed to this model a liking for the object. In contrast, they ascribed to the model a disliking for the avoided product. In Experiment 2, this result emerged, with a smaller effect size, even when it was clear that the behaviours followed specific instructions from a third party. The model had been a mere executor instead of behaving autonomously. Finally, in Experiments 3, 4, and 5, we showed, with direct and indirect measures of attitudes, that reading that the model had approached vs avoided products was sufficient to create preferences in the participant for the approached one, regardless of whether it was explained that the model was a mere executor. This research highlights the largely unexplored effects of vicarious approach/avoidance behaviours. Theoretical and practical implications and possible developments of this line of research are discussed.

Using Behavioral Measures to Assess Suicide Risk in the Psychiatric Emergency Department for Youth.

Suicide screening is critical in pediatric emergency departments (EDs). Behavioral measures of suicide risk may complement self-report measures. The current study examines suicide-specific behavioral measures and tests their potential short-term within-person effects among respondents, ability to discriminate future suicide attempt from suicidal ideation, and translation into interpretable categorical composite scores. The sample included 167 youth (10-17 years), presenting for suicide-related reasons to a pediatric psychiatric ED. During their ED visit, participants completed the Death/Life Implicit Association Test (IAT) and the Suicide Stroop Task. Recurrent suicidal thoughts and attempts were assessed within 6 months of the ED visit via medical records and email surveys. Youth displayed a decrease in the levels of distress and self-injurious desires (negative mood, desire to hurt themselves, and desire to die) after completing the behavioral tasks. The Death/Life IAT prospectively differentiated with 68% accuracy between youth who attempted suicide after their ED visit and those who had suicidal ideation but no attempt, p = 0.04, OR = 5.65, although this effect became marginally significant after controlling for self-report and demographic covariates. Neither the Suicide Stroop Task, nor the categorical composite scores predicted suicide attempts, ps = 0.08-0.87, ORs = 0.96-3.95. Behavioral measures of suicide risk administered in the ED do not appear to increase distress or self-injurious desires. They may be able to distinguish those who go on to attempt suicide (vs. consider suicide) within six months after discharge.

IAT faking indices revisited: Aspects of replicability and differential validity.

Research demonstrates that IATs are fakeable. Several indices [either slowing down or speeding up, and increasing errors or reducing errors in congruent and incongruent blocks; Combined Task Slowing (CTS); Ratio 150-10000] have been developed to detect faking. Findings on these are inconclusive, but previous studies have used small samples, suggesting they were statistically underpowered. Further, the stability of the results, the unique predictivity of the indices, the advantage of combining indices, and the dependency on how faking success is computed have yet to be examined. Therefore, we reanalyzed a large data set (N = 750) of fakers and non-fakers who completed an extraversion IAT. Results showed that faking strategies depend on the direction of faking. It was possible to detect faking of low scores due to slowing down on the congruent block, and somewhat less with CTS-both strategies led to faking success. In contrast, the strategy of increasing errors on the congruent block was observed but was not successful in altering the IAT effect in the desired direction. Fakers of high scores could be detected due to slowing down on the incongruent block, increasing errors on the incongruent block, and with CTS-all three strategies led to faking success. The results proved stable in subsamples and generally across different computations of faking success. Using regression analyses and machine learning, increasing errors had the strongest impact on the classification. Apparently, fakers use various goal-dependent strategies and not all are successful. To detect faking, we recommend combining indices depending on the context (and examining convergence).

The project implicit international dataset: Measuring implicit and explicit social group attitudes and stereotypes across 34 countries (2009-2019).

For decades, researchers across the social sciences have sought to document and explain the worldwide variation in social group attitudes (evaluative representations, e.g., young-good/old-bad) and stereotypes (attribute representations, e.g., male-science/female-arts). Indeed, uncovering such country-level variation can provide key insights into questions ranging from how attitudes and stereotypes are clustered across places to why places vary in attitudes and stereotypes (including ecological and social correlates). Here, we introduce the Project Implicit:International (PI:International) dataset that has the potential to propel such research by offering the first cross-country dataset of both implicit (indirectly measured) and explicit (directly measured) attitudes and stereotypes across multiple topics and years. PI:International comprises 2.3 million tests for seven topics (race, sexual orientation, age, body weight, nationality, and skin-tone attitudes, as well as men/women-science/arts stereotypes) using both indirect (Implicit Association Test; IAT) and direct (self-report) measures collected continuously from 2009 to 2019 from 34 countries in each country's native language(s). We show that the IAT data from PI:International have adequate internal consistency (split-half reliability), convergent validity (implicit-explicit correlations), and known groups validity. Given such reliability and validity, we summarize basic descriptive statistics on the overall strength and variability of implicit and explicit attitudes and stereotypes around the world. The PI:International dataset, including both summary data and trial-level data from the IAT, is provided openly to facilitate wide access and novel discoveries on the global nature of implicit and explicit attitudes and stereotypes.

Neurocomputational mechanisms underlying cross-modal associations and their influence on perceptual decisions.

When exposed to complementary features of information across sensory modalities, our brains formulate cross-modal associations between features of stimuli presented separately to multiple modalities. For example, auditory pitch-visual size associations map high-pitch tones with small-size visual objects, and low-pitch tones with large-size visual objects. Preferential, or congruent, cross-modal associations have been shown to affect behavioural performance, i.e. choice accuracy and reaction time (RT) across multisensory decision-making paradigms. However, the neural mechanisms underpinning such influences in perceptual decision formation remain unclear. Here, we sought to identify when perceptual improvements from associative congruency emerge in the brain during decision formation. In particular, we asked whether such improvements represent 'early' sensory processing benefits, or 'late' post-sensory changes in decision dynamics. Using a modified version of the Implicit Association Test (IAT), coupled with electroencephalography (EEG), we measured the neural activity underlying the effect of auditory stimulus-driven pitch-size associations on perceptual decision formation. Behavioural results showed that participants responded significantly faster during trials when auditory pitch was congruent, rather than incongruent, with its associative visual size counterpart. We used multivariate Linear Discriminant Analysis (LDA) to characterise the spatiotemporal dynamics of EEG activity underpinning IAT performance. We found an 'Early' component (∼100-110 ms post-stimulus onset) coinciding with the time of maximal discrimination of the auditory stimuli), and a 'Late' component (∼330-340 ms post-stimulus onset) underlying IAT performance. To characterise the functional role of these components in decision formation, we incorporated a neurally-informed Hierarchical Drift Diffusion Model (HDDM), revealing that the Late component decreases response caution, requiring less sensory evidence to be accumulated, whereas the Early component increased the duration of sensory-encoding processes for incongruent trials. Overall, our results provide a mechanistic insight into the contribution of 'early' sensory processing, as well as 'late' post-sensory neural representations of associative congruency to perceptual decision formation.

Patterns of Implicit and Explicit Attitudes: IV. Change and Stability From 2007 to 2020.

Using more than 7.1 million implicit and explicit attitude tests drawn from U.S. participants to the Project Implicit website, we examined long-term trends across 14 years (2007-2020). Despite tumultuous sociopolitical events, trends from 2017 to 2020 persisted largely as forecasted from past data (2007-2016). Since 2007, all explicit attitudes decreased in bias between 22% (age attitudes) and 98% (race attitudes). Implicit sexuality, race, and skin-tone attitudes also continued to decrease in bias, by 65%, 26%, and 25%, respectively. Implicit age, disability, and body-weight attitudes, however, continued to show little to no long-term change. Patterns of change and stability were generally consistent across demographic groups (e.g., men and women), indicating widespread, macrolevel change. Ultimately, the data magnify evidence that (some) implicit attitudes reveal persistent, long-term change toward neutrality. The data also newly reveal the potential for short-term influence from sociopolitical events that temporarily disrupt progress toward neutrality, although attitudes eventually return to long-term homeostasis in trends.