Friday, September 28, 2018

Evidence of Selection Bias involving Firearms at the Center for Disease Control

The scientists at the Centers for Disease Control have shown their prejudice against firearms ownership once again. They have done so with selection bias and word manipulation.

Scientists are supposed to evaluate factual information to arrive at testable theories about objective reality.

A problem for all scientists is selection bias.  Scientists, as with nearly all humans, are more likely to include data that bolsters their pre-conceptions, and exclude data that refutes their theories.

The CDC has committed both errors in their National Violent Death Reporting System (NVDRS). It takes self discipline and self examination to fight against selection bias.

The CDC is making a value judgment in favor of one political side of the debate about an armed population. They do so by the selection of what to study, and how to define violence.

Consider the choice of words and the study issue first.

Violence is morally neutral, like gravity. Violent acts can be good, bad, or neutral. A thunderstorm is violent. It is morally neutral. Defense of self or the nation can be violent. It is morally good. Aggression against individuals or groups can be violent, and tends to be morally bad.

Some people consider suicide to be a bad, violent act, although the primary victim and perpetrator are the same. Others consider suicide to be gentle, non-violent, and good, when sanctioned by the state in the form of "assisted" suicide.

The CDC selects certain violent actions, which may be good, bad, or neutral, and lumps them all together in one category, violent deaths, with the implication that violence is always bad.  Lumping some violent acts together lumps the good, the bad, the ugly all together in a mishmash. Other violent acts are excluded based on value judgments. The numbers are used to promote political values of one side.

This particular instance of selection bias involves firearms related deaths in the National Violence Death Reporting System  (NVDRS). The CDC includes unintentional firearm deaths in the violent death count. They exclude unintentional deaths in vehicles, drowning, with poisons and other causes.

If a fatal gun accident is violent, so is a plane crash, a fatal car accident, a fatal poisoning accident, and a fatal drowning accident.

The CDC acknowledges this bias. From
NVDRS defines a death due to violence as "a death resulting from the intentional use of physical force or power against oneself, another person, or against a group or community." NVDRS collects information about homicides, suicides, deaths by legal intervention-excluding executions-and deaths of undetermined intent. In addition, information about unintentional firearm injury deaths (i.e., the individual did not intend to discharge the firearm) is collected, although these deaths are not considered violent deaths by the above definition.
Why are unintentional firearms deaths included? Because the CDC wanted them included.  Why are unintentional plane crashes, accidental motor vehicle deaths, drownings, and poisoning excluded? Because the CDC wanted to exclude them.

While not perfect, the rates of death in the 27 NVDRS states is close to the rates of death in the United States as a whole, as recorded in the CDC WISQARS data base. The "violent death rate" in the 27 NVDRS states is calculated at 19.67 per 100,000 population, age adjusted.

The rate of death including all age adjusted firearm related deaths in the NVDRS is 9.90 per 100,000 population. The rate in the entire United States in the CDC WISQARS data base for the same year, 2015, age adjusted, is 11.03, about 11% higher.

The rate of unintended deaths per 100,000 by poisoning (14.92),transportation (11.95 ), falling (10.40), drowning (1.10), and fire/burns (.76) in the WISQARS data for the entire United States add up to 39.13 deaths per 100,00 population, age adjusted.

39.13 dwarfs the entire "violent death" list put together by the CDC NVDRS of 19.67.

The unintentional death rate by firearm for 2015 is .15 per 100,000 population, age adjusted, in the United States. That is about .4% of the total unintentional death rate. Those numbers are virtually lost in the noise of other unintentional deaths.

Using the NVDRS subjective and biased definition of violent deaths, firearm related deaths are 50.3% of the total. 

However, when all injury-related deaths are tallied, firearms-related deaths, which include homicide, suicide, and unintentional injury, are in fact only 17.3% of the total. (In WISQARS for 2015, the total comes to 63.76 per 100K population. All firearms-related deaths are 11.03 per 100K).

Gun Violence is a propaganda term invented by those who desire a disarmed population. It is used to transfer volition and accountability from humans to inanimate objects,  guns. I suspect the CDC/NVDRS definition of violent deaths was chosen to prop up this effort. Intentional acts are not a disease.

I do not expect the CDC to reform their partisan definition of violent deaths.  The CDC excludes most unintentional deaths as violent deaths while including unintentional firearms deaths. They have done this deliberately.

It is selection bias, and it shows the wisdom of forbidding the CDC from using tax dollars to produce firearm related political propaganda.

©2018 by Dean Weingarten: Permission to share is granted when this notice and link are included.

Gun Watch


ScienceABC123 said...

Again, you can't be a political activist and a scientist they're mutually exclusive.

ExpatNJ said...

Paul Weber said: "you can't be a political activist and a scientist; they're mutually exclusive".

Sorry to bust your paradigm, Paul. But, I am a Scientist in the Pharma/HAPPI (Household & Personal Product) Industries, AND I am a staunchly pro-2A political activist. Unfortunately, I became active only after being a victim of violent crime in 1984. Thanks to reaching a proper and educated epiphany way back then, I was able to successfully repel a armed, home invasion in 2008, and survive uninjured.

Please note the admonition of one of the greatest writers of American literature:
"All generalizations are false, including this one" - Mark Twain.

However, I have found - and I do agree with you, Paul - that the majority of my colleagues in the Scientific community are 'liberal', if not strongly Progressive, and anti-2A, if they have an opinion on the issue. Maybe I am just a rare exception ...