Poster #124 - Risk of Violent Victimization among Transgender Individuals: Prevalence Estimates Using a Propensity Score Matching Approach

• In Event: Poster Session II (posters 80 – 155)

Abstract

Although it is well established that the incidence of violence against those who are transgender is excessive, little work has been done to isolate the unique effect of gender identity from cooccurring risk factors such as socioeconomic marginalization, marital status, and age. Separating out the effect of gender identity on victimization risk is difficult because (1) low base rates of violent victimization and (2) a proportionally small transgender sub-population both combine to preclude comparisons of covariate-adjusted estimates of victimization risk. Using data from the nationally representative National Crime Victimization Survey, we deal with these methodological constraints by taking a propensity score matching (PSM) approach to match transgender respondents with cisgender controls on key sociodemographic characteristics (e.g., poverty, age, marital status, metropolitan residence) thought to be predictive of both transgender identity and victimization risk in order to estimate the net effect of transgender identity on the risk of violent victimization. This research will demonstrate the utility of a PSM approach to account for the risks of rare events amongst small sub-populations and it has the potential to underscore the extent to which having a marginalized identity itself is associated with disproportionate levels of vulnerability to harm.

Authors

• Kathryn DuBois, Washington State University

• Amelie Pedneault, Washington State University