Seed Grant Awards

Each spring, the Owens Institute for Behavioral Research (OIBR) seeks applications for faculty seed grant projects in the social and behavioral sciences. All faculty in the Social and Behavioral Sciences are encouraged to apply, but OIBR gives preference to proposals from faculty who are OIBR distinguished scholars, affiliates, or grant development program participants.

More information about the program.

Year	Awardee	Title	Amount	Abstract
2027	Shana Adise	Assistant Professor, Nutritional Sciences	$15,000	Optimizing a Neuroinflammation-Sensitive MRI Battery for Pediatric Brain Health Research Childhood obesity affects approximately 20% of youth in the United States and is associated with differences in brain structure, function, and cognitive outcomes, including deficits in executive function (EF), the set of neurocognitive processes governing inhibitory control, reward processing, and food-based decision-making. Concerningly, obesity in adulthood is linked to accelerated cognitive decline and dementia risk, suggesting that neurobiological consequences of excess weight gain may be set in motion well before adulthood. Neuroinflammation arises through diet- and adiposity-driven systemic inflammation, and cellular and psychosocial stress signaling, both of which are disproportionately experienced by children with obesity. When chronically activated, neuroinflammatory pathways erode myelin, disrupt brain microstructure, and impair the prefrontal and limbic circuits that support self-regulation and appetite control. Emerging preclinical evidence suggests that neuroinflammatory perturbations during this window may not be fully recoverable, underscoring the urgency of characterizing these processes in humans before they compound into lasting cognitive deficits. Pediatric human research has been constrained by the absence of noninvasive tools capable of measuring neuroinflammation-sensitive brain health in living children. This project will address this gap by optimizing a 45-minute neuroinflammation-sensitive MRI battery in 20 children ages 9–10 and examining associations between MRI-derived brain health markers and EF to generate preliminary effect size estimates for future investigations.
2027	Sarah Haight	Assistant Professor, Epidemiology & Biostatistics	$14,402	Provider Perspectives and Practices in Perinatal Mental Health (P4MH): A multi-method pilot study in rural Georgia Perinatal mood and anxiety disorders (PMADs) affect approximately 1 in 7 individuals, yet remain severely underdiagnosed and undertreated, with stark racial and ethnic disparities in diagnosis and care. In Georgia, the situation is especially concerning – the state has one of the highest maternal mortality rates in the U.S. and mental health conditions are a leading cause of pregnancy-related mortality. Of Georgia’s 159 counties, 120 are rural and over 40% do not have an OB/GYN, a number projected to be even higher when specifically considering obstetrics. These gaps leave the identification and treatment of PMADs to providers like family medicine physicians, pediatricians, nurse practitioners, certified nurse midwives, and physicians assistants. These providers are often the first and only point of contact for perinatal mental health concerns, yet little is known about how the perspectives and practices of providers may contribute to the identification and treatment of perinatal mental health conditions. Importantly, inequities in perinatal mental health care may stem from differential screening, diagnosing, and treating behaviors by patient factors like race and ethnicity, but existing research relies on self-reported survey data that cannot capture real-time clinical decision-making or implicit bias. This pilot study will develop and test a novel two-pronged approach combining provider surveys with virtual reality (VR) simulation to examine provider perspectives and practices in perinatal mental health care. By comparing self-reported practices (survey) with observed behaviors (VR) across diverse patient presentations, this research will identify clinical decision points where implicit bias may contribute to perinatal mental health disparities—providing the foundation for targeted, evidence-based provider training interventions to improve equity in perinatal mental health care.
2027	Roberta Salmi	Associate Professor, Anthropology	$9,600	Kinship and Social Relationships as Behavioral Mechanisms of Gene Flow in a Primate Hybrid Zone This project advances a theoretical framework in which reproductive isolation is not governed solely by genetic incompatibilities or ecological divergence but is dynamically shaped by social decision-making within group-living species. Drawing from behavioral ecology, kin selection theory, and models of socially mediated gene flow, we examine how dominance relationships, affiliative bonds, dispersal patterns, and mate choice behaviors regulate reproductive boundaries between two hybridizing langurs: the Endangered northern purple-faced langur (Semnopithecus vetulus philbricki) and the Vulnerable tufted gray langur (Semnopithecus priam thersites) in Sri Lanka. Although hybridization is increasingly recognized as common in primates, the behavioral mechanisms structuring gene flow remain poorly understood because few systems integrate validated parentage reconstruction with direct behavioral observation in habituated populations. This project addresses that gap by establishing reliable noninvasive genetic pedigrees and linking them to observed mating interactions, social affiliation, proximity patterns, and aggressive encounters. By integrating genetic pedigrees with observed social relationships, this project will create the first kinship-informed map of reproductive structure in this hybrid population. Rather than treating genetics and behavior as separate domains, this approach explicitly links them: pedigrees reveal reproductive outcomes, while behavioral data illuminate the social processes that shape them. Establishing this foundation is essential before scaling to whole-genome analyses of hybridization extent, directionality, and history.
2027	Suhang Song	Assistant Professor, Health Policy & Management	$15,000	Artificial intelligence-powered image-based dietary assessment for Alzheimer’s disease and related dementias epidemiological research Alzheimer’s disease and related dementias (ADRD) affect 7.2 million Americans aged 65 and older, a number projected to nearly double by 2050, creating substantial clinical, social, and economic burden. Diet is a key modifiable risk factor for cognitive decline and ADRD; however, progress in understanding diet-cognition relationships has been constrained by limitations in dietary assessment. Traditional self-report methods, such as 24-hour dietary recalls (24HRs), rely heavily on episodic memory and are prone to measurement error. These challenges are amplified among older adults with mild cognitive impairment (MCI) or ADRD, contributing to attenuated associations, inconsistent findings, and limited ability to examine dietary influences on ADRD progression. Recent advances in artificial intelligence (AI) enable automated, image-based dietary assessment. However, most existing systems were developed in younger, cognitively intact populations, lack interpretability, and are not optimized for real world image capture conditions common in aging populations. No validated, aging-tailored, interpretable image-based dietary assessment tool has been rigorously evaluated against a true ground-truth reference and directly compared with traditional methods (i.e., 24HR) in older adults across the cognitive spectrum. This project will develop and pilot-test an AI-powered, image-based dietary assessment tool designed for older adults, including individuals with MCI and ADRD. The proposed tool aims to reduce reliance on memory-based reporting, operate under real-world image capture conditions, generate interpretable reasoning outputs, and capture detailed dietary characteristics such as preparation methods.
2027	Julie Stanton	Associate Professor, Cellular Biology	$3,150	AI and Learning in the Life Sciences: Characterizing metacognition as students solve problems with AI Generative AI (GenAI) is a powerful yet fallible resource that is shaping the way undergraduates think in the life sciences. As college students increasingly turn to GenAI to support their learning, there is an urgent need to help them develop the metacognition required to assess AI-generated outputs. Metacognition is defined as awareness and control of thinking, and it can improve student learning, achievement, and problem solving in science (e.g., Halmo et al. 2022; Ohanti & Hisasaka, 2018; Wang et al., 1990). Students with strong metacognitive skills can monitor their understanding, which involves determining what they do and don’t know. They can also evaluate their solutions to problems, which includes assessing the accuracy and relevance of their solutions. When students use GenAI as a partner for life science learning, their metacognitive skills are essential for ensuring they think critically about GenAI outputs and evaluate whether the model’s response is correct and applicable. The major objective of this project is to characterize the metacognition life science undergraduates use while they solve problems with help from GenAI. To reach this goal the project will address two specific aims: (1) Compare the ways life science undergraduates monitor their understanding when they solve problems with and without help from GenAI, and (2) Compare the ways life science undergraduates evaluate the accuracy and relevancy solutions when they solve problems with and without help from GenAI. The proposed studies will determine the metacognition instructors can reasonably expect life science students to use with GenAI without direct instruction. The results will also define specific areas where students require explicit guidance from instructors to use metacognition with GenAI. This foundational knowledge is essential for guiding the development of interventions that target the areas where students need the most support.
2026	Kristen Bub	Professor, Educational Psychology	$9,685	Harnessing the power of wearable technology to monitor real-time stress in early childhood educators There is a nationwide teacher shortage – with teachers leaving the profession at a considerable rate (8% turnover and 16% mobility; Taie & Lewis, 2023) and fewer individuals choosing to join the education workforce (Reinke et al, 2025). One of the leading causes of low recruitment and retention is the level of stress associated with the teaching profession, especially among educators who work with historically and institutionally marginalized populations (Vincente et al., 2024). Most research on educator stress has relied primarily on teacher self-reports of perceived stress or other distal markers such as retention rather than on more direct physiological markers or a combination of both teacher report and physiology to understand educator stress. The present study seeks to (1) develop models of early childhood educators’ normative physiological stress patterns (2) evaluate the similarities and differences between self-reported perceived stress and experienced stress with experienced stress measured through physiological markers and (3) to understand the feasibility of ecological momentary assessment (EMA) approaches to understanding educator stress by examining the added value of wearable technology. This study will inform our understanding of the connection of perceived stress to biometric stress levels and contribute to the future development of just-in-time interventions for teachers.
2026	Gaurav Sinha	Assistant Professor, Social Work	$10,000	Emerging Adults’ Awareness, Use, and Engagement with AI-Driven Chatbots and Large Language Models for Mental Health Emerging adults (ages 18-25) face several barriers to accessing mental health services, including stigma, discrimination, lack of awareness, limited availability of trained professionals, and financial constraints (Arnett et al., 2014). In the absence of formal mental healthcare and due to the widespread availability of technologies like cellphones, computers, and the internet, many emerging adults turn to AI-driven chatbots and Large Language Models (LLMs) such as ChatGPT or Co-pilot for advice or emotional support (Rubenstein, 2024). While these platforms offer a sense of anonymity, they also present risks, such as misinformation or negative reinforcement, which could potentially impact mental health outcomes (Torous et al., 2021). Although AI-driven chatbots and LLMs are becoming more accessible, emerging adults’ perceptions and engagement with these tools, particularly in the context of mental health, remain underexplored. The proposed project aims to explore the factors influencing emerging adults’ engagement with AI-driven chatbots and LLMs for mental health issues, and identify barriers to their adoption. Specifically, the project will: (a) Examine the associations between awareness, usage patterns, and perceptions of AI-driven chatbots and LLMs for mental health among emerging adults, and (b) Identify the barriers and conditions influencing the adoption of these tools among this demographic.
2026	Liyuan Wang	Assistant Professor, Health Promotion & Behavior	$9,900	Navigating Digital Spaces: Understanding Social Media Use and Mental Health Among Gender Expansive Adolescents in the Deep South Gender expansive adolescents (GEA), whose gender identities or expressions extend beyond traditional cisgender definitions, face disproportionately high risks for depression and suicidal ideation. They are three to thirteen times more likely to be diagnosed with mental health conditions than their cisgender peers, a disparity is magnified by discriminatory environments, limited access to gender-affirming care, and pervasive stigma. Although social media spaces can foster discovery, peer-led support, and identity affirmation for GEA, they can also amplify transphobia, discrimination, and cyberbullying. This study aims to investigate how social media use (SMU) influences the mental health of GEA in the Deep South, particularly those residing in rural areas. Findings from this study will inform just-in-time adaptive interventions (JITAI) to mitigate harmful online interactions and reinforce positive coping strategies.
2025	Hilda Kurtz	Professor and Department Head; Geography	$949	Modelling Tensions and Tradeoffs in Community Responses to Large Scale Solar Development In the contentious and polarized environment of the United States in the 21st century, finding common ground and achieving compromise between opposing viewpoints has become increasingly difficult. Given the far-reaching impacts of climate change, this proposal responds to a pressing need to identify actionable areas of agreement in the domain of energy transition. A massive shift to renewable energy is essential for addressing climate change, improving air quality, conserving resources, enhancing energy security, fostering economic development, and increasing resilience to disasters. Ambitious federal targets for energy transition include accelerated development of large scale photovoltaic solar installations (LSS), raising solar from meeting 3% of demand in 2021 (80 GW) to meeting 44%–45% of demand (up to 1570 GW) by 2050. These targets call for 10.3 million acres of land nationwide to be repurposed for large scale solar installations by 2050 [1]. These ambitious targets for solar power production are imperiled, however, by social polarization regarding the costs and benefits of large-scale solar installations (LSS). Clear and early communication with community stakeholders is a vital element of successful siting for renewable energy [2-3]. Yet without a more sophisticated understanding of intersecting concerns about prospective solar developments, and whether and how compromise is possible, LSS projects are slowed and stymied, putting an urgently needed energy transition in the U.S. at risk. This project seeks to fill significant knowledge gaps regarding how environmental, economic and social values and concerns coalesce to form distinct and nuanced perspectives on LSS development. The purpose of this project is to develop proof of concept for a novel interdisciplinary approach to modeling differing viewpoints on LSS installations. Specifically, we proposed a staged combination of Q-method, generative AI, and choice experimentation with which to address the questions: How do different values and concerns intersect to form identifiable segments of opinion about local development of LSS? What combinations of factors are most important to the different segments of community opinion? What features of proposed solar developments generate the most support or the most opposition? The insights generated from this novel approach can inform policy and decision-making processes related to large-scale solar siting projects. With better understanding of community preferences and concerns, policymakers and project developers can tailor project proposals and mitigation strategies to better align with community values, potentially reducing opposition and increasing the likelihood of project acceptance.
2024	Glenna Read	Associate Professor, Advertising & Public Relations	$9,600	Effect of point-of-view on interpretation of body-worn camera footage: A psychophysiological investigation of cognitive processing and evaluation of culpability Body-worn cameras (BWC), small cameras worn on the body that record and provide footage of police encounters from a first-person point of view (POV), are used by an increasing number of police agencies around the United States (Chapman, 2018). Footage from these devices can be used to promote improved officer-citizen relations, deter breaches of procedural justice, and increase transparency within the justice system (e.g., White House, 2014; BJA, 2015). Some reports indicate that their use contributes to a positive relationship between law enforcement and citizens (BJA, 2015). However, despite the advantages of BWC to both police and citizens, recent research indicates that people interpret BWC footage heterogeneously in ways that can facilitate biased outcomes (Bailey, Read et al., 2021; Salerno & Sanchez, 2020; Wilson et al., 2017). For example, one study found that viewers of BWC footage attributed a violent interaction to situational factors when told the footage came from a policewoman. When told that the same BWC footage came from a policeman, viewers attributed the interaction to the officer's aggression (Salerno & Sanchez, 2020). This research, and others, demonstrates that societal stereotypes and other cognitive processes bias interpretation of BWC video evidence. This is problematic because the collection and release of BWC videos shape public discourse around policing and are often used to assess citizen/officer culpability in both formal (e.g., trial) and informal (e.g., public opinion) contexts. For this reason, we propose to build on our previous work by (1) assessing how the first-person POV inherent to BWC footage interacts with social factors such as citizen race to affect perceptions of citizen/officer culpability, (2) testing if instructions that promote awareness of these potential biasing effects can reduce bias in interpretation of BWC footage and, (3) examining implicit, automatic processes underlying reception of both the videos and the instructions using psychophysiological measures and eye-tracking.
2023	Daniel Jung	Assistant Professor, Health Policy & Management	$7,500	The Impact of COVID-19 Pandemic on the Gaps in Care Quality and Health Outcomes between Medicare Home Health Care Patients with and without Alzheimer's Disease and Related Dementia While it is known that the COVID-19 pandemic had a disparate impact on vulnerable populations and communities in terms of access to high-quality health care and health outcomes (Werner & Bressman, 2021), we have less information on its impact in the context of Medicare home health care. The pandemic has intensified challenges such as workforce shortages as well as insufficient access to essential resources in home health care (Denise Tyler et al., 2021; Sama et al., 2020), which may affect care quality and health outcomes differently based on pre-existing patient and HHA-level characteristics. Especially, ADRD patients are known to have worse functional and cognitive abilities which require complex needs and heavily rely on caregivers (J. Burgdorf et al., 2019; Konetzka et al., 2020). Considering these characteristics, ADRD home health care patients may have been disproportionately affected by the pandemic. Identifying the unique impact of the pandemic on home health care and ADRD patients will inform expectations and policies aimed at improving Medicare home health care and mitigating the gap in care quality and health outcomes, as well as the need for further monitoring of the impact of COVID-19 pandemic. Moreover, identifying factors that affect quality of care for ADRD patients during the pandemic will inform policymakers not only on how to support home health agencies (HHAs) and disadvantaged subgroups who were disproportionately affected by the pandemic, but also on how to take action to reduce gaps in Medicare home health care beyond the effect of the pandemic.
2023	Cassia Roth	Associate Professor, History	$5,150	Determining Reproductive Value: Slave Prices, Gender, and Reproduction in Nineteenth-Century Brazil This project tests the relationship between the prices of enslaved women of reproductive age (15–35) and changing legal parameters in the largest slaveholding society of the nineteenth century: Brazil. It uses a differences-in-differences (DiD) methodology and a large sample size (n ≈ 20,000) from original archival research in three different Brazilian regions (São Paulo, Rio de Janeiro, Salvador da Bahia) to understand the role that reproduction played in the abolition of slavery. It thus has relevance for current debates on the lasting legacy of the intersection of racial and gender inequality in former slave societies such as Brazil and the United States.
2022	Drew Abney	Assistant Professor, Psychology	$9,955	The Role of Multi-Level Parent-Child Synchrony on Child Self-Regulation This pilot study brings together researchers with expertise in varied areas of scholarship to examine relations between parent-child synchrony and child self-regulation in a community sample of 72 families from diverse sociodemographic backgrounds. Self-regulation – the ability to manage behavior and emotions in response to environmental demands – underlies healthy development across the lifespan and is foundational for adaptative development. There is strong support for the notion that the developmental process of self-regulation has biological, social, and environmental bases, that the process unfolds largely in the context of the caregiver-child relationship, and that the dyadic unit itself serves key functions that can either facilitate or impede healthy child self-regulation. It is therefore the goal of the current study to investigate the dyadic unit through the concept of synchrony: the matching of behavioral, physiological, or neural states1. Prior work has largely examined parent child synchrony at the behavioral and physiological levels, but emerging theoretical insights2, informed by metaanalytic results3, suggests a deeper link between meaningful behavioral, physiological, and neural co-variation among children and their parents. This study will examine parent-child synchrony as a predictor of child self-regulation using behavioral, physiological, and hyper-scanning functional near-infrared spectroscopy (fNIRS) neuroimaging methods in an adapted conceptual framework that makes connections between the brain, body, and environment. Data from this innovative pilot study will serve as proof of concept and inform the development of a larger external grant proposal submitted using either the NIH R01 mechanism (PA-16-160) or the NSF: Developmental Sciences mechanism (PD 08-1698). This work will ultimately inform models of child self-regulation and potentially identify novel targets for improving prevention and intervention programs for children and their families.
2022	Brett Clementz	Professor, Psychology	$9,990	Sensory Target Engagement for Psychosis B-SNIP Biotype-1 (STEP-1) This project will enhance the B-SNIP consortium, by using biomarkers as Biotype-specific targets for clinical application. DSM-type diagnoses do not provide such treatment targeting information. The intervention will be a sensory training regimen that could be easily delivered in any psychologist’s office. This approach will transition psychosis treatment from a subjective, to an at least partially objective, enterprise. We will use already recruited psychosis cases and identify BT1 cases with a target biomarker feature (deficient early sensory processing due to low neural response to salient stimuli, as seen in EEG event-related potentials; ERPs). The psychosis comparison group (BT2) have deficient early sensory processing due to high EEG intrinsic brain activity, yet normal response to salient stimuli. BT3, who are normal on cognition and sensory processing, will be excluded. We will target the defining feature of neurophysiological dysfunction in BT1 using an established sensory intervention (Tone Matching Training; TMT) to normalize their neural deviation. Previous sensory training studies have been marginally successful but intervened on heterogeneous sychosis rather than a psychosis subgroup with a specific physiological dysfunction for which the treatment should be specifically efficacious. We have submitted this study as a large multi-site trial (R01), but reviewers wanted to see an indication that the intervention will specifically target BT1 like we predict (this is the only part of the project for which we lack prodigious preliminary data). Ergo, we are requesting these OIBR funds to support that specific request. We will stratify participants to one of two groups until completion of 22 cases (n=11/Biotype). TMT will be administered in 20 min sessions x 4/day, 2-3 days/week over 3 weeks, totaling 32 sessions. Baseline, midpoint, and post-training assessments will target specific neurophysiological outcomes. Target engagement (TE) hypotheses: (TE1) TMT will augment ERP amplitudes and improve sensory signal fidelity in BT1 but not BT2; and (TE2) training will improve sensory discrimination ability in BT1 but not BT2.
2022	Caree Cotwright	Associate Professor, Family and Consumer Sciences	$8,534	Evaluating an interactive eLearning training to improve healthy beverage policy implementation among Early Care and Education Teachers in Georgia It is essential to reduce the prevalence of obesity among young children because obese children have a higher risk of developing chronic diseases later in life. Decreasing intake of sugar-sweetened beverages (SSBs) and increasing consumption of healthy beverages (e.g., water and milk) among young children can help address the problem of childhood obesity. SSBs are drinks that contain sucrose, fruit concentrate, high fructose corn syrup, or other caloric sweeteners (e.g., soda, fruit drinks, juice drinks). National Health and Nutrition Examination Study data show 62% of children aged 2-5 years consume SSBs, contributing 127-130 kcal to daily caloric intake. As a large majority of young children spend a significant time in the early care and education (ECE) setting, it has been recognized as an important environment for childhood obesity prevention. The majority of ECE teachers in Georgia have never had a beverage policy training. There is a lack of evaluation research regarding the use of eLearning to promote obesity prevention policies to ECE teachers. This pilot study will use an intervention-control design to test the effectiveness of an eLearning beverage policy training. Study outcomes will provide vital information on the use of an eLearning beverage policy training to improve policy implementation among ECE teachers. The proposed study may contribute to reducing obesity risk and promoting healthier diets among preschool children in Georgia.
2022	Justine Tinkler	Associate Professor, Sociology	$4,500	Intergroup Relations and Prosocial Behavior in the Wake of the COVID-19 Pandemic Recent research on intergroup behavior suggests that individuals respond to a new threat from an outgroup by prioritizing the most privileged identity to which they feel they belong (Abascal 2015). The current study builds on and extends Abascal’s work by investigating whether the framing of the pandemic as a threat from Asia affects relations not only between Asian Americans and other groups but also has spillover effects on relations between white and Black Americans. To test our thesis, we designed an Internet-based experiment that exposes people to a news article about an ongoing pandemic threat from Asia and then measures 1) respondents’ prosocial behaviors toward a same-race versus a different-race recipient in a one-shot dictator game and 2) the strength of their racial and American identities. We have collected data from 1000 white Americans via Amazon Mechanical Turk. Preliminary results show that our treatment prime successfully activated threat perceptions and that white participants react to the prime by donating significantly less money to Asian American recipients. We propose to use the seed grant to collect data from an additional 1200 Black and Asian Americans. This research is interdisciplinary by bringing together theories from sociology, psychology, and race and ethnic relations to specify how an outside threat affects racial and national identities that have spillover effects on interracial behavior.