skip to main content


Title: Examining the Effects of Race on Human-AI Cooperation
Recent literature has shown that racism and implicit racial biases can affect one’s actions in major ways, from the time it takes police to decide whether they shoot an armed suspect, to a decision on whether to trust a stranger. Given that race is a social/power construct, artifacts can also be racialized, and these racialized agents have also been found to be treated differently based on their perceived race. We explored whether people’s decision to cooperate with an AI agent during a task (a modified version of the Stag hunt task)is affected by the knowledge that the AI agent was trained on a population of a particular race (Black, White, or a non-racialized control condition). Thesedata show that White participants performed the best when the agent was racialized as White and not racialized at all, while Black participants achieved the highest score when the agent was racialized as Black. Qualitative data indicated that White participants were less likely to report that they believed that the AI agent was attempting to cooperate during the task and were more likely to report that they doubted the intelligence of the AI agent. This work suggests that racialization of AI agents, even if superficial and not explicitly related to the behavior of that agent, may result in different cooperation behavior with that agent, showing potentially insidious and pervasive effects of racism on the way people interact with AI agents.  more » « less
Award ID(s):
1849869 2218226
NSF-PAR ID:
10321024
Author(s) / Creator(s):
; ;
Editor(s):
Thomson, R.; Hussain, M. N.; Dancy, C. L.; Pyke, A.
Date Published:
Journal Name:
14th International Conference on Social Computing, Behavioral-Cultural Modeling, & Prediction and Behavior Representation in Modeling and Simulation
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. In this study, we drew on the m(ai)cro framework, which centers racism as a macrosystem, to examine how college-going emerging adults made meaning about society and themselves during the 2020 U.S. presidential election and 2021 inauguration. This period was marked by racial justice protests, a global pandemic, anti-Asian violence, and the storming of the U.S. Capitol by predominantly white Trump supporters. Using the constructs of critical consciousness and racial identity meaning making, we analyzed participants’ reports of recent race related conversations. Our sample included 47 students ( Mage = 19.71, SD = 1.72; 81% female, 17% male, 2% other; 45% Asian/Asian American, 30% white, 13% Latinx/Hispanic, 4% Black/African American, 4% Multiracial, 2% Middle Eastern/Arab) at a private, predominantly white university in the U.S. Midwest. Hybrid inductive-deductive analysis showed that a majority reported conversations with peers, focused primarily on racial inequity and justice. For many participants of color, conversations about topics including protests and anti-Asian violence were woven into their racial identities. In contrast, although many white participants discussed events such as the Capitol insurrection, none made links to their racial identities. Our findings highlight connections between critical consciousness and racial identity, and the importance of context and participant positionality in developmental research.

     
    more » « less
  2. Background: Even though Historically Black College and Universities (HBCUs) make up only 3% of higher education's institutions, they play a pivotal role in producing Black scientists by virtue of the fact that many received either their undergraduate or doctorate degree from a HBCU. HBCUs are credited with providing a more supportive and nurturing environment that thrives on communal mindsets and practices, emphasizing the importance of relationships, offering opportunities for Black students to "see themselves" as part of the academic and social milieu whereas Historically White Institutions (HWIS) are characterized as being hostile and discriminatory. Mentoring is said to be pivotal in the attainment of the PhD. Mentorships have an inherent gatekeeping mechanism, better positioning those who receive effective mentorships while disadvantaging those who do not. It has potential to harm and marginalize when not engaged with deliberate care and a culturally liberative mindset. Mentoring, when not under the thumb of colonizing mindsets, can contribute to more equitable experiences and outcomes for students who hail from AGEP population groups. Literature has indicated that Black students are less likely to have a mentor or be engaged in effective mentorships. The HBCU narrative of supportive environment is consistently told but has scant empirical validation for Black students pursuing STEM doctoral degrees. In fact, the lure of having faculty and peers who look like you is something of an enigma given that even at HBCUs there are limited numbers of Black faculty in STEM. How are same race, same gender mentorships attained when, not unlike their HWIS counterparts, HBCU STEM faculties have a large number of White and Asian men? If the environment is indeed different at HBCUs, is it different for Black STEM doctoral students? Is STEM doctoral mentoring at HBCUs emblematic of anti-Blackness or is it yet another tool used to oppress marginalized students? Theoretical Framework: Anti-black racism and critical capital theory serve as critical theoretical frameworks and were selected because they highlight the ways violence is enacted through taken for granted colonized practices such as mentoring. Fanon understood that thoughts and mindsets are the progenitors of violence and dehumanization is the process through which violence is enacted. Anti-black racism and critical capital theory can be useful in unearthing the structural inequalities that uphold the current system in place for STEM doctoral learning. Research Design: An embedded multiple qualitative case study research project sought to understand the nature and quality of STEM doctoral mentorships at an HBCU. The analysis on the HBCU subcase asked, how are STEM doctoral mentorships understood by Black STEM doctoral students at HBCUs? Black STEM HBCU students were interviewed and completed a mentoring competency assessment survey. In addition STEM doctoral students from three universities also completed the survey. The qualitative data was analyzed using narrative analysis and the survey data was analyzed using descriptive and inferential statistics. This project is part of a larger NSF AGEP sponsored research study. Research findings: The findings from this study expose that Black STEM doctoral students at HBCUs have not reached the proverbial Promise Land. In spite of being in a space that is more diverse, they manage to simultaneously be invisible and hypervisible. An unmerited sense of assumed cultural belonging was highlighted with students reporting a lack of selfethnic reflectors in their programs. In many ways the systemic and institutional structures on HBCUs with respect to STEM doctoral programming mirrored the colonial structures more often associated with HWIS. Their culture and cultural-based experiences as domestic students as well as their academic strengths were often not recognized by mentors while that of international students were. Three themes were supported by the data: Conspicuous Absence, Race Still Matters, and Invisibilized Hypervisibility. Implications: Better understanding how STEM doctoral mentoring is facilitated at HBCUs holds the promise of informing a mentoring practice that supports cultural liberation instead of cultural degradation and suppression. It becomes one avenue as the “The Call'' suggests to "confront our own complicity in the colonial enterprise" by holding STEM doctoral mentors and the institutions they represent accountable for socially just mentoring practices. Greater intentionality as well as mandated training informed by the study's results are recommended. HBCU faculty doctoral mentors are challenged to be scholar activists who engage mentoring from an advocacy and accomplice framework. The development of STEM scholar activists is the aspiration of more culturally liberative STEM doctoral mentorships. Black students need mentors who are willing and equipped to be advocates and accomplices in their success. 
    more » « less
  3. ackground: Historically Black College and Universities (HBCUs) have for decades played a pivotal role in producing Black scientists. Research found that HBCUs, despite being under funded and resourced, were responsible for over 10% of Black scientists with doctorates. Even though most earn their doctorates at Historically White Institutions (HWIS), understanding the experience of Black STEM doctoral students at HBCUs is of paramount importance to impacting opportunity for success for underrepresented population groups. HBCUs are recognized for approaches to learning and learning environments that are more relational, encouraging peer to peer and student to faculty relationships, particularly in the form of same-race and same sex mentorships, resulting in less negative racialized gendered experiences and less competitive atmospheres. In spite of what appears to be accepted truths, such as HBCUs offering more culturally affirming experiences, some researchers suggests that little empirical research exists on the quality of support structures available for graduate students at HBCUS in STEM academic fields, particularly mentoring. Increased understanding would provide essential framing necessary for developing more effective mentors at HBCUs, especially given that there are limited numbers of Black faculty in STEM, even at HBCUs. Theoretical Framework: Anti-racism and critical capital theory are employed as theoretical frameworks. Both are well suited for questioning taken-for-granted assumptions about the lived experiences of racialized others and for deconstructing systemic issues influencing common faculty practices. These frameworks highlight the contextual experiences of STEM doctoral learning. Research Design: The researchers were interested in understanding how STEM doctoral faculty at HBCUs perceive their role as mentors. An NSF AGEP sponsored social science research project explored the dispositions, skills, and knowledge of eight STEM faculty at a HBCU. Attitudes towards culturally liberative mentoring were explored through a qualitative case study. The participating faculty were involved in an institutional change program and were interviewed for an average of 60 minutes. Constant comparative data analysis method was used. Additionally, STEM faculty from participating departments completed two mentoring competency and attitude inventories. This case was drawn from a larger multiple embedded case study. Research Findings: The research findings indicate that STEM doctoral faculty mentors at HBCUs express attitudes about mentoring that are not all that different from their PWIS counterparts. They have a tendency to hold deficit views of domestic Black students and have minimal awareness of how culture inhibits or facilitates a positive learning experience for Black students. Further the culture of science tended to blind them from the culture of people. Research Implications: In order to enhance the learning experiences of Black STEM doctoral students at HBCUs, the Black student experience at HBCUs must be deromanticized. Understanding the impact of anti-Black racism even within an environment historically and predominantly Black is imperative. Recognizing the ways in which anti-Black attitudes are insidiously present in faculty attitudes and practices and in environments perceived as friendly and supportive for Black students highlights opportunities for STEM faculty development that can move toward a more culturally liberative framework. 
    more » « less
  4. This study examined how well people can recognize and relate to animated pedagogical agents of varying ethnicities/races and genders. For both Study 1 (realistic-style agents) and Study 2 (cartoon-style agents), participants viewed brief video clips of virtual agents of varying racial/ethnic categories and gender types and then identified their race/ethnicity and gender and rated how human-like and likable the agent appeared. Participants were highly accurate in identifying Black and White agents but were less accurate for Asian, Indian, and Hispanic agents. Participants were accurate in recognizing gender differences. Participants rated all types of agents as moderately human-like, except for White agents. Likability ratings were lowest for White and male agents. The same pattern of results was obtained across two independent studies with different participants and different onscreen agents, which indicates that the results are not solely due to one specific set of agents. Consistent with the Media Equation Hypothesis and the Alliance Hypothesis, this work shows that people are sensitive to the race/ethnicity and gender of onscreen agents and relate to them differently. These findings have implications for how to design animated pedagogical agents for improved multimedia learning environments in the future and serve as a crucial first step in highlighting the possibility and feasibility of incorporating diverse onscreen virtual agents into educational computer software.

     
    more » « less
  5. Abstract Background and aims

    Despite widespread negative perceptions, the prevalence of alcohol‐impaired driving (AID) in the United States remains unacceptably high. This study used a novel decision task to evaluate whether individuals considered both ride service cost and alcohol consumption level when deciding whether or not to drive, and whether the resulting strategy was associated with engagement in AID.

    Design

    A two‐sample study, where sample 1 developed a novel AID decision task to classify participants by decision strategy. Sample 2 was used to cross‐validate the task and examine whether decision strategy classifications were predictive of prior reported AID behavior.

    Setting

    A laboratory setting at the University of Missouri, USA.

    Participants

    Sample 1 included 38 student participants from introductory psychology classes at the University of Missouri. Sample 2 included 67 young adult participants recruited from the local community.

    Measurements

    We developed a decision task that presented hypothetical drinking scenarios that varied in quantity of alcohol consumption (one to six drinks) and the cost of a ride service ($5–25). We applied a Bayesian computational model to classify choices as consistent with either: integrating both ride cost and consumption level (compensatory) or considering only consumption level (non‐compensatory) when making hypothetical AID decisions. In sample 2, we assessed established AID risk factors (sex, recent alcohol consumption, perceived safe limit) and recent (past 3 months) engagement in AID.

    Findings

    In sample 1, the majority of participants were classified as using decision strategies consistent with either a compensatory or non‐compensatory process. Results from sample 2 replicated the overall classification rate and demonstrated that participants who used a compensatory strategy were more likely to report recent AID, even after accounting for study covariates.

    Conclusions

    In a hypothetical alcohol‐impaired driving (AID) decision task, individuals who considered both consumption level and ride service cost were more likely to report recent AID than those who made decisions based entirely on consumption level.

     
    more » « less