By: Angie Laird, PhD, Florida International University (https://doi.org/10.18116/18b3-1v85)

Large, open datasets have emerged as important neuroimaging resources that offer exciting opportunities for innovative discoveries. Before engaging in secondary data analyses, it is essential that researchers consider relevant ethical issues for responsible data use.

Training Considerations for Responsible Conduct of Research

Training in responsible conduct of research typically emphasizes important topics related to the protection of human subjects, animal welfare, and laboratory safety. However, responsible conduct of research also extends to data management, sharing and ownership, scientific rigor and reproducibility, and responsible authorship and publication.

Increased sharing of large, open datasets must be accompanied by heightened attention to ensuring the protection of participant identity, including individuals from more vulnerable populations, such as patients with clinical disorders and/or from historically underrepresented groups.

Responsible Data Analyses Require Advance Planning

Beyond concerns about participant privacy, responsible data analyses require advance planning, becoming familiar with the data acquisition protocols, and understanding the limitations of the acquired data.

Generally, when planning to collect study data, scientists can choose the research instrument that best addresses their research question. However, when working with existing data, the study design cannot be manipulated. Consequently, the scientific process is reversed, and the research question must be designed based on the appropriateness of the available instrument. Secondary data analysis projects can be conducted relatively quickly - the data are already collected and available for download. But responsible data use requires that researchers pause before engaging with the data and think carefully to ensure that research questions are suitable for the given dataset.

Preventing Stigmatizing Research

Finally, prior to any data analysis or interpretation, researchers must engage responsibly and fully consider the psychological, social, economic, and any other potentially harmful impacts their research could have on individuals, communities, and society. Specifically, this means that responsible use of variables related to race, ethnicity, gender, and sex must be thoughtfully considered prior to conducting analyses of neuroimaging data. Comparisons across participants who are grouped by race and/or ethnicity can potentially be interpreted as evidence of biomarkers that explain neurobiological mechanisms through which some communities experience lower rates of achievement and poorer life outcomes. To discourage continuation of this biological deficits framework, it is imperative that data analysts recognize that ethical conduct in research includes ensuring that analyses prevent further stigmatization, marginalization, and injustice toward individuals because of racial, ethnic, or gender status.

Additional discussion of these issues can be found in two recent publications, including a review article and a practical guide for analyzing data from the ABCD Study.

Guidelines for Preventing Stigmatizing Research have been adapted from the Responsible Conduct of Research developed by All of Us, the NIH’s Precision Medicine Initiative.

Quarterly Newsletter Article from June 14, 2022