How Hemenway and Solnick Distorted the Effectiveness of Defensive Gun Use
15 Pages Posted: 29 Aug 2020 Last revised: 31 Aug 2020
Date Written: June 28, 2020
Hemenway and Solnick (2015) assessed the effectiveness of crime victims using defensive gun use (DGU) to reduce their risk of injury. They obtained a positive (injury-elevating), though nonsignificant, estimate of this effect. This result is misleading because they distorted their estimates of the effect as a result of three research flaws:
(1) They needlessly reduced their sample size, relative to available data, by 75%, thereby artificially reducing the chances of obtaining statistically significant estimates of the association between victim SP actions and injury.
(2) They misclassified incidents in which injury was not inflicted after the victim’s self-protection action as incidents in which injury was inflicted after the victim’s SP action, artificially supporting the claim that the SP action provoked the offender to attack.
(3) Most importantly, they systematically omitted from their injury models virtually all of the known confounders of the DGU/injury relationship - all but one of which tended to bias that relationship upward. Thus, the failure to control these confounders artificially made it seem that DGU had more of a positive association with injury than it actually has.
Suggested Citation: Suggested Citation