Originally Posted by
350htrr
Originally Posted by
MMACH 5
For 14% of individuals, those preparations will prove absolutely necessary. Hopefully they will have health insurance.
True, but nobody personally knows who that actual risk will be including... Same with helmets, for the people who's head bounces off the pavement wearing one when it happens is better than not wearing one in general I predict.
Thus the helmet thread, some people are willing to take more risk than others, some use the population's risk % factor as to weighing their risk, others use their own risk calculations done by their own senses/idea of personal risk. Not that the risk is big but it's there...
We can use statistical inference related to conditional probabilities to get a more precise probability. In the heart disease example, Baye's Theorem. Not intending to get technical, but in addition to the 14% base probability you'd need to know two more things about the individual with respect to a particular dependent probability. Such as exercise which we know improves our chances of not suffering heart failure. We'd need to know the base probability, the percentage of people who exercise, and the statistical probability that people who exercised died of heart disease. From this and using the formula that expresses Baye's theorem we obtain a new probability that the person who exercises regularly will die of heart failure.
Similarly for other factors such as diet, stress levels etc, knowing the related statistical data for those factors we can derive a probability for the individual - more precisely, for progressively smaller subsets of the population which include the individual. To be clear, this represents the probability that the given individual will experience heart failure.
One common fallacy in interpreting statistics is when you
ignore the base rate. This error occurs when someone is presented a base rate probability, and a specific case, and instinctively focuses on the specific case while ignoring the base rate. We see it a lot in helmet literature, and sad to say, in this list. If we say for example, 97% of the people who died from heart failure did not take Vitamin C supplements, and from just that fact advise to take the supplements regularly, it's an example of the base rate fallacy (and one or two more).
With regards to fatalities related to helmet use or non-use, the base rate probability is extremely small which makes avoiding the base rate fallacy particularly important.
From accident data there are more elements, other statistics, from which probabilities may be inferred. Where you ride, time of day, even age group, and you can get more specific as well with variously credible data sets. You can formally calculate a refined "population's risk %" from these statistics, or, if you are one to "use their own risk calculations done by their own senses/idea of personal risk" I posit that both are the same general concept. You are using statistical data from observed events (your own or aggregated or both) and inferring a probability representing risk.