Mitigating invalid data bias in the estimation of sexual orientation disparities in a survey of youth in US and Canada

Abstract

The current commentary explored the applicability of the methods described in “Mitigating invalid and mischievous survey responses: A registered report examining risk disparities between heterosexual and lesbian, gay, bisexual, or questioning youth” by Dr. Joseph Cimpian and colleagues to explore sexual orientation disparities in preexisting data from a nonprobability sample. Understanding Affirming Communities, Relationships, and Networks was a study of mostly White (77.4%) 9674 sexual and gender-minoritized youth aged 15–29 from the US and Canada. The influence of invalid data on the prevalence ratios of four health outcomes was assessed. The methods yielded similar effects to the original paper. The accuracy varied by outcome prevalence and was robust to misspecification of the model. Therefore, the applicability of this method to preexisting data seems feasible.