Statistical Analyses and Applications for Studying Intersectionality and Disability Inclusion

Kaitlyn Cook Chair
Smith College
Johanna Hardin Discussant
Pomona College
Shiya Cao Organizer
Smith College
Wednesday, Aug 7: 10:30 AM - 12:20 PM
Invited Paper Session 
Oregon Convention Center 
Room: CC-252 



Main Sponsor

Committee on Statistics and Disability

Co Sponsors

Caucus for Women in Statistics
Health Policy Statistics Section
Justice Equity Diversity and Inclusion Outreach Group
Social Statistics Section


Intersectionality and Disability in Multiple Sclerosis: A statistical mediation analysis approach using imaging biomarkers

Prior research has shown that Black American (BA) patients with multiple sclerosis (MS) experience a more severe disease progression than their White American (WA) counterparts [1-5], however causal pathways of this relationship have not been explored. In this talk, we model the role of race, gender, and socioeconomic status on MS related disability outcome using imaging biomarkers as the mechanism of mediation. We explore a variable selection approach to find the imaging mediating features on the relationship between race and disability. Our aim is to explore the causal pathways of such relationships. Our findings could ultimately have direct benefits for clinical care by explaining distinct race-related processes that impact disease progression.  


Susan Gauthier, WCMC
Nara Michaelson, Mass General Brigham


Sandra Hurtado Rua, Cleveland State University

Intersectionality of Race and Mental Health Among People who Inject Drugs from Respondent Driven Sampling

Respondent driven sampling (RDS) has been practiced in studies targeting people who inject drugs (PWID), who, otherwise, are difficult to recruit. Although lacking stability, typically existing for a short term, PWID do form social networks. RDS capitalizes on this nature of the PWID community and recruits its members.
We conducted the Project Positive Attitudes towards Health (PATH), an RDS study of Southeast Michigan PWID in 2022-2023. Reflecting the demographic composition of the study area, our sample comprised a large proportion of minoritized PWID. This offers us a unique opportunity to examine health outcomes by race among PWID. Further, because RDS hinges on existing social networks, its byproducts (e.g., whether a respondent successfully recruits a PWID peer) along with various social network measures implemented in the PATH (e.g., "how many PWID do you know?") allow us to examine not only the PWID social networks by race but also the role of social networks on mental health and in the relationship between mental health and race. We illustrate this intersectionality through network visualization in conjunction with modelling techniques. 


Sunghee Lee, University of Michigan


Stephanie Morales

Model interpretation after using random projections: An applied study on travel disability data

The National Household Travel Survey (NHTS) asks respondents whether they have a medical condition "that makes it difficult to travel outside of home", which is defined as travel disability in this research. The NHTS allows us to investigate the effects of disability on travel behavior, however, it may release some sensitive medical conditions and travel data. We use a differential privacy algorithm – random projection to get a random dataset that contains the summary statistics of the sample dataset so useful aggregate information can be released and used for the intended purposes, while the privacy of the individuals in the sample dataset is preserved. The main idea of this differential privacy algorithm is to use random projection to project a sample dataset (n by p) to a random dataset (k by p). We fit a linear regression model for the random dataset and compare the statistics of interest of the random dataset with those of the sample dataset. With this differential privacy algorithm, we can examine the accuracy of our random projection compared to the original sample and then make statements about statistics of interest of the true population. 


Keegan Kang, Bucknell University


Shiya Cao, Smith College