Learning Objective: We currently do not have direct measures of household gun ownership at geographic levels smaller than Census region. At the conclusion of the session, participants will be able to describe the limitations this imposes on research and will be able to describe the strengths and limitations of a proposed proxy.
Study Design: To identify a valid and reliable proxy of firearm prevalence at the state level, we calculated cross-sectional correlations among 8 proxies for gun prevalence and compared these proxies to temporally matched, survey-based estimates of household gun ownership rates at the state and regional level. Regressions with and without fixed effects for region and time were compared to assess how well proxies perform as cross-sectional vs. longitudinal estimates.
Results: Among the 8 candidate proxies one measure performed better than the rest in every validation and reliability test we conducted; namely, the percentage of suicides committed with a firearm (FS/S). The wide differences in firearm prevalence exhibited across states exhibit stability over recent decades. Limitations in our gold-standard survey sample size relative to the small changes in regional firearm stocks over recent decades make it difficult to assess the usefulness of our chosen proxy as a longitudinal estimate of survey-based gun ownership.
Conclusion: Among candidate proxies for firearm prevalence, the percentage of suicides committed with a firearm was most strongly and reliably associated with survey based measures of household firearm ownership and is a valid and reliable measure of firearm prevalence for cross-sectional studies.
Learning Objectives: Learning Objective: We currently do not have direct measures of household gun ownership at geographic levels smaller than Census region. At the conclusion of the session, participants will be able to describe the limitations this imposes on research and will be able to describe the strengths and limitations of a proposed proxy.
Keywords: Firearms, Injury Control
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.