top of page

Summary of "Instructing Animosity: How DEI Pedagogy Produces the Hostile Attribution Bias"


In November 2024, the Network Contagion Research Institute (NCRI), (2) in collaboration with researchers from Rutgers University's Social Perception Lab (1), published a report examining the psychological effects of certain diversity, equity, and inclusion (DEI) training materials. Titled Instructing Animosity: How DEI Pedagogy Produces the Hostile Attribution Bias, the study investigates whether common "anti-oppressive" and "anti-racist" approaches in DEI pedagogy achieve their intended goals of reducing bias and fostering empathy—or whether they inadvertently heighten perceptions of hostility and prejudice.


The research focuses on materials drawn from prominent DEI sources, testing their impact across three domains: race (drawing from scholars like Ibram X. Kendi and Robin DiAngelo), Islamophobia, and caste discrimination. Using controlled experiments, participants exposed to these DEI texts were compared against control groups reading neutral content. The findings suggest that such materials can induce a "hostile attribution bias," leading participants to perceive discrimination, microaggressions, and unfairness in ambiguous scenarios where no evidence of bias exists, while also increasing support for punitive responses.


The following is a detailed summary of the report's methodology, key results, and implications, prepared by CFBU research assistant, Natalie Roman. This document aims to provide an accessible overview for readers interested in evidence-based evaluations of DEI interventions, highlighting the need for rigorous testing of training programs to ensure they promote understanding rather than unintended division.

--

A Summary of

Instructing Animosity: How DEI Pedagogy Produces the Hostile Attribution Bias. NCRI, Nov. 25, 2024.


Purpose and Background

The purpose of the research was to examine the effectiveness of diversity, equity, and inclusion (DEI) training. The stated purpose of DEI training programs is to foster inclusivity by educating participants about prejudice and bias. (3) However, since their real-world effectiveness has been questioned, with some studies suggesting these interventions may backfire, the study was designed to address whether certain types of DEI training unintentionally foster intergroup hostility and increase perceptions of bias. 


Methodology

The primary questions asked in the research were whether the ideas and rhetoric used by DEI training foster inclusiveness or exacerbate conflict, and whether they increase empathy/understanding or increase hostility. The research examined the effect of DEI training related to race, religion (Islam), and the caste system. (4) Participants in the study were randomly assigned to read either a DEI-based article or a neutral article. They were then presented with a neutral scenario. Participants then completed a psychological survey that measured participants perceptions of bias, hostile attribution bias, authoritarianism, and punitive responses. 


Participants for the study on race were Rutgers University students and a national sample recruited via Amazon Prime Panels. Participants for the Islamophobia and caste study were recruited through Amazon Prime Panels.


Race

The study examined the effect of race-based DEI training that utilizes the “anti-oppressive” and “anti-racist” framing on removing bias and creating empathy. To this end, half of the participants read an article derived from writings of Ibram X. Kendi and Robin DiAngelo, two prominent DEI scholars. The other half read a neutral article describing corn production in the U.S. that contained no social, moral, or emotional framing. Both groups answered a survey after they were presented with the following scenario: “Eric Williams applied to an elite East Coast university in Fall 2023. During the application process, he was interviewed by an admissions officer, Michal Robinson. Ultimately, Eric’s application was rejected.” 

There was a statistically significant difference between the two groups.


Compared to the participants who read the essay on corn, those who read the “anti-oppression”/“anti-racist” article were more likely to perceive that the applicant was a person of color and that the admissions officer performed microaggressions against the applicant, was biased, and was unfair in rejecting the applicant. Further, those exposed to the DEI article were also more likely to support that the admissions officer should suffer consequences. Possible consequences included a public and personal apology, an investigation by the DEI office, being required to take a DEI course, and students’ protest of the admissions officer.


The study states that there was no measurable difference in the two groups with regards to their warmth or coldness towards people of color after reading the article. However, the table that is meant to show that is missing from the article.


Islamophobia

The study also examined whether anti-Islamophobia training is effective at counteracting anti-Muslim bias or whether it distorts perceptions of bias and fairness. For this purpose, the study had half of participants read an article containing excerpts of DEI Islamophobia training from the Institute for Social Policy and Understanding. The other half read the neutral article describing corn production in the U.S. Both groups answered a survey after they were presented with the following scenario: “After bombing a local government building, Ahmed Akhtar/George Green was put on trial and convicted of terrorism.” The names Ahmed Akthar and George Green were randomly flipped. 


For those who read the corn article, there was no difference in the perception of the fairness of Ahmed’s and George’s trial. However, for those who read the anti-Islamophobia training, Ahmed’s trial seemed significantly less fair than George’s trial, even though the scenario was identical for both.


Caste

Caste-based discrimination is not a core of DEI training in the U.S., but it has been gaining more attention in the last few years. Most Americans are also unfamiliar with caste discrimination, which the authors felt could be used as a “blank slate” to evaluate how DEI training affects perceptions. (5)


For this portion of the study, two different articles on the caste system were used. The neutral article was compiled from writings of a broad range of academics and was free of accusatory or sensationalized language. The second article consisted of excerpts from anti-oppression writings used by Equity Labs, a DEI firm that provides training for caste discrimination in North America. Both groups then answered a survey after they were presented with the following scenario: “Raj Kumar applied to an elite East Coast university in Fall 2022. During the application process, he was interviewed by an admissions officer, Anand Prakash. Ultimately, Raj’s application was rejected.”


Those who read the excerpts from Equality Labs were significantly more likely to perceive microaggressions, caste bias, and harm. They were also significantly more likely to support negative consequences, including not permitting Brahmins to wear sacred clothes on campus, firing and suspension of the admissions officer, and forced DEI training and investigation. Further, those who read the Equality Labs excerpt were significantly more likely to perceive Hindus as racists and were significantly more likely to agree with “Hitler-like” statements (Brahmins are parasites, Brahmins are a virus, Brahmins are the devil personified). 


Given the lack of knowledge of both groups of participants about the caste system, it is reasonable to conclude that the difference in the responses was based directly on the content of the articles with which they were presented.


Implications

The data presented in the study indicates that at least some anti-oppressive DEI training materials fail at the goal of combating bias and creating understanding and empathy. Instead, they seem to provoke “baseless suspicion and encourage punitive attitudes.” (6) Even where there is no evidence of bias or unfairness, some DEI training appears to “create a hostile attribution bias.” (7) This can also lead to undermining trust in institutions (8) and can “heighten racial suspicion, prejudicial attitudes, authoritarian policing, and support for punitive behaviors in the absence of evidence for transgression.” (9) Therefore, DEI training ought to be tested for effectiveness at reducing bias and increasing empathy and to examine potential harms.

Notes:

  1. The goal of the SPL is to study “how people think about, understand, judge, evaluate, and perceive” others. Social Perception Lab, Rutgers.

  2. NCRI was founded by psychologist Joel Finkelstein. It states that its mission is to “track, expose, and combat misinformation, deception, manipulation, and hate across social media channels.” Network Contagion Research Institute (NCRI): Reports | Library of Congress. 

  3. “What Is DEI Training? Building Inclusion,” Robert f. Smith, 31 Oct 2023; “8 Types of Diversity, Equity and Inclusion Training to Implement Within Your Organization,” InStride, 1 Apr 2024.

  4. The caste system, used primarily in India, divides Hindus into four main hierarchical categories (Brahmins, Kshatriyas, Vaishyas, and Shudras) and one category that is outside the caste system who are considered untouchables (Dalits). “What is the Caste System?” BBC, 19 June 2019.

  5. Instructing Animosity, p. 8-9

  6.  Instructing Animosity, p. 7 

  7.  Instructing Animosity, p. 8

  8.  The authors also point out that when extreme “egalitarian rhetoric" is framed as a moral imperative, it can lead to “coercive control, intolerance, and punitive attitudes,” which are part of authoritarian movements. This is further enhanced in a person who already has authoritarian attitudes. Instructing Animosity, p. 13

  9.  Instructing Animosity, p. 14

 
 
bottom of page