Furthermore: IF THERE ARE REAL NEGATIVE EFFECTS (as all the studies you cite suggest) THAT CAUSE SOME PEOPLE TO GET INTO ACCIDENTS BECAUSE OF THEIR CELL PHONE USE, THERE MUST BE SOME OFFSETTING EFFECT THAT LEADS OTHER CELL PHONE DRIVERS NOT TO GET INTO ACCIDENTS THEY OTHERWISE WOULD, THUS RESULTING IN THE SAME 6% RATE.
I think it is very useful to ask what are those offsetting effects.
So now you admit that your belief is a hypothesis. That is a good start. Now if you test your hypothesis and prove it to be true, let me know.
You should’t just take two statistics and put them side by side and claim that together they mean something. That isn’t how hypotheses are tested and proven.
These two “6%” stats put together don’t say ANYTHING about whether or not the accidents involving cell phones were caused by the use of cell phones. So it appears you are drawing that conclusion out of thin air. These two “6%” stats don’t prove you right and they don’t prove you wrong.
If my explanation still doesn’t make any sense to you, I invite you to discuss the matter with a statistician or any type of scientist. Maybe they can explain it better than I can. Perhaps you could ask someone who has studied mathematical logic at the college level.
What you are saying is a little like saying “When I opened the door it was noisy. It isn’t noisy anymore. Therefore, the door must be closed.” The door might really be closed, but logically, the conclusion drawn might be false. There might be other reasons the noise has disappeared.
Again, if you are witholding some kind of proof other than the conclusion that you THINK has been proven by two unrelated statistics, please let me know. I fully understand your position, but it is based on a fallacy.
I love the one (from the NHTSA) that says:
A Prevention Magazine survey in 1995 reported that 85% of cell phone owners use their phones while driving at least some of the time. If such statistics hold true, there are now about 81 million drivers talking on their phones while driving, and about 10 million of these use the phone during every trip.
…and fewer people had cell phones back then.
But that isn’t as good at this one:
I am here today because I am a co-author of the Department of Transportation’s 1997 report entitled “An Investigation of the Safety Implications of Wireless Communications in Vehicles”, and because of my involvement in crash investigation research. I review hundreds of crash investigation reports every year. I have seen the faces of the dead, I know the stories of the injured, and I believe that the use of cell phones by drivers creates an unnecessary risk to the driving public.
…or this one…
What is striking about all the recorded cell phone-related crashes is that they fall into the two categories of striking something in front of them, or leaving their lane of traffic. They do not reflect the pattern of all crashes which also include intersection collisions, rear ends and rollovers. Something is different about cell phone crashes, and I suggest to you that it is driver inattention.
The more research I conduct, the more support I find for my side of this arguement. I don’t find a single piece of evidence that supports your so called “conclusion.” Why do you think that is the case? Have I been looking in the wrong place? If so, where should I be looking for evidence that supports your so called “conclusion?” Is there anyone in the world that shares your belief that people are better drivers when they are on the phone?
Zombie, I just showed your original post to one of my colleagues who has a background in math and statistics. He says you have made an invalid transitive assumption.
Here is how he explained it to me. Let’s say there is a football game between Florida State and the University of Texas and Florida State wins. Then the University of Texas beats the University of Georgia. According to your logic, Florida State should beat the University of Georgia. However, when they play the game, the University of Georgia might beat Florida State, proving that the transitive assumption between Florida State and the University of Georgia was invalid.
Is that the best you can do to talk intelligently about this very serious matter? Your condescending comments don’t contribute much. In fact, I feel a bit embarrassed for you that you need to resort to that.
Look at what you just wrote to refute my argument:
These two “6%” stats put together don’t say ANYTHING about whether or not the accidents involving cell phones were caused by the use of cell phones.
So? What’s your point here? Are you saying that the numbers reported by Harvard might even be too high? I’m sure many accidents were wrongly blamed on cell phone use…OK, but that makes it look even safer to use cell phones while driving. Previously, you argued that the cell usage rate was UNDER reported by the NHTSB methodology which again would make it look even more surprising that so few accidents involve cell phones.
Transitive smangitive! Nonsense! What I am saying is that the rate of accidents involving cell phone drivers should be higher than the random chance of a driver using a cell phone! It’s that simple. The same way that the rate of accidents involving drunk drivers is higher than the proportion of drivers who happen to be drunk. If there was no increase, than I would say that drunk driving is safe too!
When I am proven correct, this will be forever known here after as “The Zombie Theorem”:
“Any dangerous activity must result in a higher accident rate than the rate at which the activity commonly occurs.”
~The Zombie
Testing? Your lack of interest in pure theoretical automotive analysis astounds me.
Zombie…I have an advanced degree in math. I believe your 6% statistics are independent. The first 6% gives you a sample from which the second 6% is drawn, but is not a causal factor for the other. Without some additional information or “linking” statistic, they would be mutually exclusive. No conclusion can be drawn in my opinion. I think that’s what Whitey was saying. I wouldn’t put the two statistics together and draw any conclusion.
Like the statement: if A implies B implies C than A implies C; We’re missing B.
I do understand your willingness to assume a linkage as it does seem logical given the wording in the two statements. Statistically though, it’s too much of an assumption without additional info.
That the two “6%” were from two different studies invalidates their common inclusion…now if the one or the other had a study that included both, you could look for that linkage.
Just my two cents.
Of course they are independent studies and it’s not an exact science, so of course there would be differences in methodology, wording, etc. Are you telling me that after you see a cell phone usage survey with an N=43,000(!), you are not comfortable using that number to create reasonable expectations about other things that have to do with cell phone use? And it doesn’t surprise you when your expectations differ from studies that attempt to measure exactly what you are estimating? Of course it shouldn’t be exactly identical, but you don’t find it odd that there is no measured increase in the accident rate? By one or two percent? It sure looks to me like there is no causation. And just to be clear, both of you claim that there is no way to conceptualize any of this?
Even if the methodology and wording were indentical and beyond reproach, your conclusion is a logical fallacy. There is neither evidence of causation nor evidence of a lack of causation.
To be perfectly clear, to conceptualize the conclusion you have drawn would require me to assume facts that you have not proven.
Ok, so assuming the 6% cell usage rate is very accurate, what would you expect the measured cell phone accident rate to be if there was absolutely no causation between cell phone use and accidents?
The silence is deafening…I’ll answer that for you: 6%.
And if there was some sort of a safety danger inherent in cell phone use, that number should be higher by some amount. I don’t know by how much, but by something at least. Well, this quack university did a study and they determined that number to be…hold your breath…6%. Now, I’m not saying there is EXACTLY no relationship in all or any one particular accident, but I’m saying that the inherent danger overall must not be very much if this fly-by-night, bought-and-paid-for university cannot even measure it.
Just to put this into perspective with the drunk driving analogy, according to the NHTSB ( http://www.alcoholalert.com/drunk-driving-statistics-2004.html ), in 2004 7% of all accidents were “alcohol related”. Now, imagine somebody came out with a study that said: “We stopped 43,000 random drivers and we found that 7% of them happened to be drunk”. Now, we would all be shocked if that were the case, because we all know that the rate of alcohol related accidents should be way higher than the random rate of drunk driving. Also, we know that 7% sure seems too high based on what we see in our everyday lives (I’m guessing).
My point is that this is similar to the cell phone analogy except that we are not surprised that 6% of drivers use cell phones (in fact, it was earlier argued on this blog that this was being under-reported).
“Are you telling me that after you see a cell phone usage survey with an N=43,000(!), you are not comfortable using that number to create reasonable expectations about other things that have to do with cell phone use?”
That’s exactly right…not and relate it to another unrelated study. Each study should be argued on it’s own merits. It would take a third study to validate any conclusion that would link the two. It’s a pretty slippery slope for me to officially state a relationship between the two. If for friendly argument purposes you want to make a personal inference with “I believe that to be true” or “in my opinion”…than go for it. Then you have a hypothesis that you now have an opportunity to prove.
That’s the approach I would use with my students in teaching statistics and calculus. That’s all I feel comfortable saying. But then, you have more freedom to say what you want about this matter, and I as an instructor have the obligation to say… “I don’t agree.”
OK, yes, these are my opinions. I have not personally surveyed 300 million people and I have not confirmed their reports by following every single one of them around. I am basing my opinions on two abstract concepts: (1)The idea that 6% of all drivers at any one time are using cell phones and (2)The idea that 6% of all accidents involve drivers who are on cell phones. These concepts might or might not be true, but it is my opinion that I respect the credibility of the two studies I cited for what that is worth and I assume their margins of error are negligible, in fact I am ignoring the margins of error, and I am extrapolating these rates to apply to the entire driving population.
There, are you happy? Now let’s talk intelligently about the merits of the concepts and what can be logically deduced if we accept them.
Can we at least agree that as a theoretical abstraction, having identical rates of usage and related accidents would actually DEFINE a lack of causation?
I think that if your phone rings, you should be able to answer while you are pulling off the road to talk. After your conversation, you can resume driving.
Or just ban it altogether, that’s fine too.
“There, are you happy? Now let’s talk intelligently about the merits of the concepts and what can be logically deduced if we accept them.”
Now my equally unverifiable argument would be that because the two 6% do agree, there appears to be a one to one correspondence between the number of cell phone users and that number as a percent that appear in accidents. I therefore conclude that as the number of cell phone users increases (which ii is sure to do w/o intervention), so will the number of cell phone related accidents.
This makes it imperative that we pass a law restricting cell phone use in automobiles. Cutting the cell phone use by 6% by enacting a law would save many thousands of lives…6% of all accidents.
This makes no more sense to me and is likewise unacceptable.
I guess you’re now assuming that you have been talking intelligently and the rest of us have not…another motion I can’t accept.
By the way…I favor “hands free” as the only acceptable usage of cell phone in a car, and mandated as a safety feature in all new cars. The hardware is nearly already in cars, it’s just added software.