Academic Legal Writing - Eugene Volokh [184]
A survey of Chicago robberies in 1975 revealed that, of those victims taking no resistance measures, the probability of death was 7.67 per 1000 robbery incidents, while the death rate among those taking self-protection measures was 64.29 per 1000 robbery incidents. The victim was 8 times more likely to be killed when using a self-protective measure than not!
Here, source A does correctly summarize the original study (source C). But let's compare the student article with source A:
(vi½) As in error (vi), source A talks about one place, Chicago, one time, 1975, and one crime, robbery; but the student article talks generally about “a person who uses a handgun in self-defense.” The reader should be alerted to this limitation, since the specific data may not apply equally to self-defense more broadly (for instance, to self-defense against burglary, assault, rape, or attempted murder), to the country generally, or to the year that the article was written.
(vii) But the big error is the leap from “self-protection measures” to “us[ing] a handgun in self-defense.” Neither source A nor the original study, source C, explicitly equates the relative risk of self-protection measures generally with the relative risk of self-protection using a handgun. The student article falsely claims something about a specific subset when the data only relates to a broader set.
Look again at the table from source C:
The study found that when a handgun was used for self-protection, 0 out of 6 robberies led to death—not 18 out of 280 (64.29 per 1000), the ratio on which source A relies, and which covers weaponless selfdefense, self-defense with weapons, verbal response, and flight. The study doesn't tell us how effective handguns really are for self-defense, since six cases are far too few to justify any inference. But the study also does not show that “a person who uses a handgun in self-defense is eight times more likely to be killed than one who quietly acquiesces.”
So the author of the student article made a bad mistake. But the author of source A also erred because his citation of source C is likely to mislead readers. Three sentences shortly before the “8 times more likely” sentence and the one sentence immediately after had to do with selfdefense using a handgun:
A Case Western Reserve University study showed that a handgun brought into the home for the purposes of self-protection is six times more likely to kill a relative or acquaintance than to repel a burglar .... The handgun is also of questionable utility in protecting against robbery, mugging or assault .... The element of surprise the robber has over his victim makes handguns ineffective against robbery .... A survey of Chicago robberies in 1975 revealed that, of those victims taking no resistance measures, the probability of death was 7.67 per 1000 robbery incidents, while the death rate among those taking self-protection measures was 64.29 per 1000 robbery incidents. The victim was 8 times more likely to be killed when using a self-protective measure than not!
Although handguns possess little or no utility as self-protection devices, some may have a socially acceptable value when properly marketed under restricted guidelines [such as to the police].
The sentences that cite source C are easily misread as focused on defensive handgun uses, rather than on what they literally discuss, which is self-protection generally—and the author of the student article seems to have misread these sentences exactly this way. Had source A explicitly said that it was extrapolating from general self-defense data to handgun data, the student article's author might have recognized the limitations of the data, and at least made them clear to his readers.
APPENDIX III: SAMPLE COVER LETTERS
A. For Sending an Article to Law Reviews
[Your name]
[Your address]
[Your phone number]
[Your e-mail address]
Articles Department
[Law review name and address]
[Date]
Dear Madam or Sir:
In the attached