Faulty Statistical Reasoning by Mlive on the Dangers of Guns and Motorcycles

[XLFD]

Two articles recently appeared on Mlive, two articles that discussed statistics dealing with guns, traffic deaths, and motorcycle helmets.  Both articles offered the data that has been compiled by what one must believe are reliable sources, and then took that data to editorialize against guns and riding a motorcycle without a helmet.  As Mark Twain once said:  "Facts are stubborn, but statistics are more pliable." and "There are three kinds of lies: lies, damned lies and statistics." 

The statistics in both articles seem to promote the authors own biases, and even though the two articles are not promoted as editorials, they wear the mantle of statistics to try to assert that there is only one way to interpret what they supposedly show.  It is a typical activity of today's media to take a set of statistics and twist it to their own use.  Now it's my turn; but I admit my bias firsthand that I believe law-abiding citizens that own guns and/or ride a motorcycle without a helmet do not pose a threat to society or themselves.

Guns

The 4-10-2013 article starts:

"As cars get safer, gun deaths eclipse traffic fatalities in Michigan

If you live in Michigan, you’re more likely to die from a bullet than from a motor vehicle accident.

In 2010, the death toll from motor vehicle traffic accidents in Michigan was 942, the National Highway Traffic Safety Administration reported. That same year, 1,056 people died from gunshot wounds, including 440 homicides and 599 suicides, according to the Michigan Department of Community Health."

Let's analyze:  First, the deaths are almost the same total, so the use of the word "eclipse" is not warranted, and used to sensationalize their point.  In the data, 57% of gun deaths are suicides, only 42% are homicides.  In both, a person is effectively killed by a bullet, but the traffic deaths do not include those people that commit suicide by starting their vehicle in an enclosed area, so why do they include gun suicide deaths for the gun stats?  If we consider only non-suicides, we find that 457 deaths by gun remain, less than half of Michigan's traffic deaths of 942.

The article then covers, in order:  other causes of death in Michigan, national pushes in gun control, rates across racial lines, and finished with studies and statistics on peripheral issues.  Throughout, they avoided the any conceptual analysis by the experts they put forth, the main one being an advocate for more 'gun control' methods, but framing it as 'gun safety' as she does unabashedly in this link in the fourth paragraph.  A 'gun control' by any other name, would smell as violative of individual rights, as Shakespeare might say.

Motorcycle Helmets

I would dare say a hard batch of statistics showing that riding without a helmet is a lot dangerous than riding with would change the minds of those who prefer to choose what (or what not) to put on their head when riding their chopper.  But I would be just as likely to believe that those who believe helmet-wearing is the only way to go were just as hard-headed in their beliefs.  These people can look at last year's statistics on motorcycle accidents and fatalities, and justify that riding without a helmet is the dumbest and most dangerous thing a motorcyclist can do.  But last year's statistics do not reflect that prejudice, so they twist the statistics for their agenda.  This was looked at last year in two Ludington Torch articles:  Helmets cause accidents? June 2012 Surprising Stats Dec. 2012

 

John Barnes starts his 4-10-2013 Mlive article:

Medical and insurance coalition urges lawmakers to repeal law allowing helmetless motorcyclists (with crash database)

Dan Petterson, who heads a pro-helmet motorcycle group, called the helmet-choice law “a tragic human experiment” that lawmakers should repeal.

"The helmet law has turned out to be a disaster that has cost human lives. They have the opportunity to take a mulligan, the opportunity to do a do-over," said Petterson, president of SMARTER, for Skilled Motorcyclist Association - Responsible Trained and Educated Riders.

"While we can’t turn the clock back to save lives, we can restore our helmet law now. There is no shame in correcting your course. The shame is in continuing on a direction that’s proven to be dead wrong."

The effort comes on the heels of a University of Michigan study that found the fatality rate for helmetless riders in crashes was nearly three times higher than for those who wore helmets. An MLive Media Group investigation published in December came to the same conclusion.

The U-M study, unveiled at the state’s Traffic Safety Summit two weeks ago, also concluded there would have been 26 fewer deaths and 49 fewer serious injuries had the helmet mandate been in place. Fifty-five helmetless riders died in 2012 after the law went into effect; 194 were seriously injured."

 

Let's analyze:   The pie graph below refers to last year's accidents which is linked to by this article.   You will notice that the fatalities are almost equally matched, with the surprising result that there was more helmeted fatalities.  But the oddity that you will never hear about from the helmet brigade is that helmeted riders got in three times as many accidents as those without. 

Statistical models reflect that the use of helmets would go down to around 57% or less of the riding population.  Why are they getting in 75% of the accidents?   Could it be that wearing a helmet does deaden your senses somewhat when you travel down the road?  Is it because the biker thinks he's more invulnerable?  Why isn't this 'anomaly' addressed by the people above.  One could extrapolate from the same data, that there would have been a considerable smaller amount of accidents if everyone did not wear helmets, and an actual reduction in serious injuries and deaths. 

                      From the interactive graph at:  MI Motorcycle Accidents 2012

 

The article continues:  "...Kris  Nicholoff, executive director of the Michigan Osteopathic Association, drew a parallel between the timing of the helmet-choice law and what he called the poor condition of Michigan roads.

“I cannot believe we repealed this law when we have an obstacle course every single day,” Nicholoff said.

Heather Drake, manager of state  government relations for AAA Michigan, said the company has received three claims so far for helmetless riders injured in crashes. Each claim will exceed $1 million, she said.

“We are now entering a new riding season and we expect claims to increase,” Drake said. “The data is clear. The experiment is over. Helmets save lives.”"

 

Let's analyze:  Kris, if Michigan roads are in bad shape, someone without a helmet is more likely to be able to navigate the roads as their vision range and reaction time increase without a bulky helmet limiting their perceptions.  Sure they may have a worse injury without a helmet, but if they are likely to avoid it about three times as often, it may be in their best interest to go without a helmet.

Heather, why do you not tell us of the hundreds of incapacitating injuries that occurred to helmeted riders and the comparative cost to society and note that they actually cost us more overall?  Why do these opponents of personal choice decide to ride recklessly on flawed reasoning and present the statistics in a way to favor their own personal goals?

[Willy]

Good points X on both subjects. I especially like the way, as you said, the articles author includes suicide gun deaths with homicide death because people would have killed themselves even if guns were not available, so the suicide statistic is irrelevant to his argument. So as you point out motor vehicle deaths are twice as high as gun homicides. One thing I did notice is that no mention was made of "accidental" gun deaths. Were those included? If so why were they not given a separate category?

[XLFD]

It just goes to show you that no matter what statistics say, you can almost always make a reasonable argument for your cause if you expand certain 'helpful' parameters and restrict other 'harmful' parameters. 

In a way, I did the same myself, but I try to work beyond my own prejudices.  If you put your complete trust in one year's statistics, the 'facts' are: 

1)  you are about three times as likely to get in an accident if you wear your helmet, but your three times as likely to die in an accident without one.  Using rounded figures, we would expect out of every 60 accidents, 45 to be to helmeted riders, with 1 fatality and 15 riders with no helmets with 1 fatality.  It only proves to me that helmets make riding less safe.

2)  you are more likely to die by a bullet than you are by a traffic accident in Michigan.  The only reason this is the case is because suicides and urban violence skews the figures; these are choices and lifestyle choices that should not be part of the debate on disarming law-abiding Americans. 

One could say you are a lot more likely to die by an abortion doctor than both of the above combined.  The 23,300 abortions induced each of the two years (2010, 2011) truly 'eclipses' the 2000 deaths each year from gunshot and traffic accidents.