For rare events / overestimation is likely even if the misclassification is not random . The Myth of Millions of Annual Self - Defense Gun Uses : A Case Study of Survey Overestimates of Rare Events

A criminologist has been using national self-report surveys to estimate the incidence of self-defense gun use in the United States (Kleck 199 I). His most recent estimate is that civilians use guns in self-defense against offenders more than 2.5 million times each year (Kleck and Gertz 1995). This figure has been widely cited, not only by the National Rifle Association but in the media and in Congress. A criminologist in Canada is touting high self-defense estimates for that country, derived from similar self-report surveys (Mauser 1995). All attempts at external validation of the 2.5 million figure show that it is an enormous overestimate (Hemenway, in press). For example, in 34% of the times a gun was used for self-defense, the offender was allegedlycommitting a burglary. In other words, guns were reportedly used by defenders for self-defense in approximately 845,000 burglaries. From sophisticated victimization surveys, however,we know that there were fewer than 6 million burglaries in the year of the survey and in only 22% of those cases was someone certainly at home (1.3 million burglaries). Since only 42% of U.S. households own firearms,