Originally Posted by
RubeRad
No, that's what statistics is. You don't have to see every BMW driver to be able to make valid observations about BMW drivers. 1% would be quite a large sample size actually, given the size of the complete population.
Having seen thousands of BMW drivers on the road and observing a pattern of more aggressive driving than other car owners establishes, with a certain amount of statistical significance, that the center of the bell curve of BMW drivers is more aggressive than the center of the bell curve of the rest of drivers.
1% would be a statistically valid sample size if you could remove all other variables
But this 'study' doesn't.
First off the observer has an admitted bias, which means poorly driving BMW are more likely to be noticed than well driving BMW's, and poorly driving Acuras are less likely to be noticed than poorly driving BMW's. That inherent bias from the observer skews the stats in favor of BMW drivers being worse than other drivers. So the statistics are immediately invalid, based on this observation.
Then there is the location. The original observation was in Munich, where BMW's are made. Now without doing any research on this I'm willing to bet Munich has a much higher than average percentage of BMW's on the road than any other city in the world, due to hometown pride in the local product and all.
So is BMW drivers in Munich are bad, who's to say it's because they are driving BMW's and not just that Munich drivers are awful in general? We don't know from this sample set. It could just be a correlation. You put BMW's in a city with terrible drivers....that doesn't make the car brand the problem. They would be just as bad in Ford Fiestas. This is a variable that is not accounted for in the 'statistical analysis'
Finally, there is no hypothesis given as to what makes BMW drivers worse than any other make of car. What ties all this data together? BMW's aren't necessarily more expensive than other cars. Some higher end BMW's are. But mid range models are lower cost than many other makes such as Porsche and Land Rover. And BMW's are well made cars. They last a long time. So used BMW's are readily available in the sub $10,000 price arena, where used Ford Fiestas tend not to be as readily available in the used market since they don't last as long. So owning a BMW doesn't immediately make one wealthy, nor doing being wealthy immediately make one a BMW owner, so that removes that as a predictive variable in this equation. So what is it? What about BMW's makes someone a poor driver?
And what about the other makes and models of cars? Have you independently collected statistical data on those to ensure that your BMW results aren't just observer bias?
What about other common threads on this very forum where people claim pick up truck drivers are the worst and most aggressive on the roads? How do you account for their 'statistical observations' not replicating your own? What makes your results valid and their results wrong?
Don't try to make a couple of people's non scientifically collected biased observations into a valid statistical conclusion. It is an exercise in futility.