Education Activities for Responsible Analyses of Complex, Large-Scale Data (R25- Clinical Trial Not Allowed)

Funding Opportunity Description The NIH Research Education Program (R25) supports research educational activities that complement other formal training programs in the mission areas of the NIH Institutes and Centers. The overarching goal of this R25 program is to support educational activities that complement and/or enhance the training of a workforce to meet the nation’s biomedical, behavioral and clinical research needs. To accomplish the stated over-arching goal, this notice of funding opportunity (NOFO) will support creative educational activities with a primary focus on methodological rigor in the analysis of large complex datasets involving brain, behavior, genomic, and socioenvironmental data. This could involve: Courses for Skills Development: For example, advanced statistics and research design courses in a specific discipline or research area, rigor and robustness in research practice, or ethical conduct of biomedical research. Curriculum or Methods Development: For example, to improve biomedical, behavioral or clinical science education, or develop novel instructional approaches or computer-based educational tools that support methodological rigor in analysis of large complex datasets. Background: With the widescale availability of large, open datasets (e.g., Human Connectome Project, the Adolescent Brain Cognitive DevelopmentSM Study, the Healthy Brain and Cognitive Development Study , and the All of Us Research Program), human neuroscience research now frequently includes population-level approaches to understanding brain structure and function and its association with behavior and substance use/mental health outcomes. The wide adoption of such approaches has led to novel insights into fundamental neurocognitive function and accelerated efforts towards understanding risk for and protection from substance use disorder (SUD) and other psychiatric disorders . However, because of the large sample sizes and heterogeneity of these datasets, there is a high likelihood that analyses will produce statistically significant results with small effect sizes that may or may not be clinically or biologically meaningful. Responsible use of such data therefore requires knowledge of analytical and statistical considerations specific to large datasets, such as population inference, sampling variability, covariate inclusion, and non-random data missingness. Additionally, with the increasing availability of open datasets, many investigators are attempting analyses of large-scale, complex data for the first time and with varied levels of training and knowledge about how to use these large datasets. Recently, interest has grown for understanding the social determinants of health or “the societal, environmental, and economic conditions that impact and affect health outcomes” (World Health Organization Commission final report, 2008). With such unprecedented motivation and opportunity to use large scale datasets to study socioenvironmental influences comes a responsibility to incorporate the broader social context and the suitability of the dataset when designing research questions and conducting the analyses. At the same time, it is important to engage communities with lived experiences pertinent to the research questions and the data analyses, and to solicit their input to inform both.Taken together, there is an urgent need for research education to help the scientific community understand how to consider study design and statistical analyses in the context of structural socioenvironmental factors. Objectives: This NOFO encourages applications that seek to advance methodological rigor in biomedical and behavioral research by supporting training on the responsible analyses of complex, large-scale datasets involving brain, behavioral, genomic, and socioenvironmental data. Topics of interest include, but are not limited to: Analytical approaches for large-scale, longitudinal data Enhanced rigor and robustness in research practice (e.g., pre-registration of experimental protocols, plans, and analyses) Estimation of meaningful associations, including population inferences, effect sizes, control of covariates, and interpretation of associations Ethical conduct of biomedical and behavioral research, including consideration of social constructs such as race/ethnicity and gender, and the potential for stigmatization Community-partnered approaches to inform data analyses and interpretation, including secondary analyses of existing data Consideration of socioenvironmental contexts known to introduce inequities – at the individual, community, and/or structural level – such as family income and education, employment, housing, neighborhood-level characteristics, and exposure to violence Factors to consider when examining the influence of socioenvironmental factors (e.g., non-random data missingness, sampling methodologies). In addition, the National Institute on Drug Abuse (NIDA) emphasizes responsible analyses of data related to neurodevelopment and neurocognition, as it relates to the substance use trajectory. Research education programs may complement ongoing research training and education occurring at the applicant institution, but the proposed educational experiences must be distinct from those training and education programs currently receiving Federal support. R25 programs may augment institutional research training programs (e.g., T32, T90) but cannot be used to replace or circumvent Ruth L. Kirschstein National Research Service Award (NRSA) programs.