Wednesday, December 20, 2017

Junk Science: "Surge" of Accidental Gun Deaths After Sandy Hook



A recent study from Wellesley College economists, published in Science magazine  is a good example of how to torture data until you get the result you want.

The mass murder at Sandy Hook took place on December 14, 2012. President Obama of the need for more gun control on 19 December.  He talked about specific restrictions on 16 January, 2013. The events are so close to the end of the year, that an increase in accidental deaths should show in the annual data.

WISQARS CDC data base

The article in Science make the following claim. From sciencemag.org:
About 60 additional unintended shooting deaths, roughly 20 of them in children, occurred in the 5 months after the shooting, conclude the study’s authors, economists Phillip Levine and Robin McKnight of Wellesley College in Massachusetts. For all of the 2012 calendar year, there were 545 accidental shooting deaths, or about 45 per month, according to the U.S. Centers for Disease Control and Prevention. So a 60-death bump in a 5-month period is a considerable one.
Look at the actual numbers as shown by the CDC. 2011 had 591 deaths from firearm accidents, a rate of .19 per 100,000.

2012 had 545 deaths from firearm accidents, significantly lower than 2011. The rate was .17 per 100,000. Sandy Hook happened at the very end of 2012.

According to Science, an increase of about 60 accidental firearms then occurred. Those 60 did not occur in 2012, because 2012 had 43 fewer accidental firearm fatalities than in 2011.

In 2013, the number of fatalities from firearm accidents was 505. That is 40 deaths lower than the number in 2012. Where are the supposed 60 fatality increase? They do not exist. It is a made up number based on extremely short term statistical trends tortured out of very limited data.

Between 1945, (the earliest figure available for per capita firearms, Kleck) to 2014 ( the latest figure, using BATFE numbers and Kleck's methodology), the number of per capita firearms has risen from .351 to 1.176. That is more than a three-fold increase.

During the same period, the per capita rate of fatal firearm accidents has declined from 1.84 per 100,000 to .15 per 100,000, over a 91% drop.

Using the earliest number for fatal firearm accidents, in 1933, the drop has been nearly 94%.

A clue to the torture of the data are the significant findings the authors refused to publish in their paper. Good science publishes findings both for and against a thesis. Junk science only publishes findings that are in favor of their thesis.

The authors tested the data against homicides and suicides. They found that their was no "systematic effect".  They chose not to publish that information in the abstract, or to make charts or tables about it. They mention it briefly in a paragraph near the end of the paper. From the study at sciencemag.org:
A third limitation is that our reported analysis has focused only on accidental firearm deaths, ignoring the potential impact on firearm-related suicides and homicides. Individuals who responded to concerns about gun control by purchasing or handling a firearm are unlikely to be motivated by an intention to kill themselves or others with those guns. We have also examined these mortality outcomes and found no systematic effect of the spike in gun sales. This finding is consistent with prior research that analyzed the short-term impact of increased firearm purchases due to local gun shows on homicides and suicides and found no evidence of impacts on those outcomes (19).
They misstate the results of the gun show study. That paper found homicides were decreased in a statistically significant way, with increased purchases at gun shows. The authors of the gun show study downplayed the effect, but they published it. From the gun show study:
"But our results provide little evidence of a gun show-induced increase in mortality in Texas. In fact, we find that in the two weeks following a gun show, the average number of gun homicides declines in the area surrounding the gun show. Aggregating across all gun shows in the state, we find that there are approximately 16 fewer gun homicides resulting from the 200 gun shows in the average year. In the sections below, we discuss several possible explanations for this counterintuitive finding. However, it is important to keep in mind that while these results are statistically significant, they are quite small – representing just one percent of all homicides in Texas in the average year."
The Wellesley study published in Science has many red flags that indicate junk science. It uses a very abbreviated data set, only looking at one event, while there are several events that show jumps in gun purchases.

The authors include all of December in their post Sand Hook period, when December was half way through when the Sandy Hook event took place. The authors focus on rates of very small numbers, but do not show us the actual numbers.

Finally, the authors ignore significant results of their own studies, choosing to publicize fatal gun accidents while ignoring homicides and suicides.

There are about 9,000 annual homicides with guns. A bump up of guns after Sandy Hook did not result in any increase.

There are about 18,000 suicides with guns annually. The bump up of guns after Sandy Hook did not result in any increase.

There are about 550 fatal gun accidents with guns annually, trending downward. The authors were able to torture the data to show an imaginary increase (fatal accidents were actually reduced) over the next year. That is what they publicized, and what was blared out over the world media.

The Science article was picked up by at least 83 news outlets.

©2017 by Dean Weingarten: Permission to share is granted when this notice and link are included.

Gun Watch















No comments: