quiz-a-thon update
this'll help me decide when to start collecting social security, assuming social security stays solvent long enough. a female acquaintance received a time of death estimate of 79 from the same quiz. i'd like to see some more results, for comparison purposes--not that i attach much significance to this result; for instance, it doesn't take into account a nutritional supplement i'm taking that's proven to prolong life in rats.
i sent email to the test makers suggesting 2 enhancements: a median predicted age for each gender, and a correlation computation for fatal test & nerd test scores.
for anyone not math or social science savvy, correlation is a number between -1.0 and 1.0, with -1.0 and 1.0 being perfect negative and positive correlation, respectively, and 0 being no correlation. 0 correlation does not mean that 2 things never happen at the same time, which is perfect negative correlation; rather, it means 2 things happen together 1/2 the time. to conclude with a familiar example, heavy smoking and lung cancer have high correlation, say 0.7 or 0.8, but it couldn't be 1.0, not only because it doesn't always cause cancer, but because a heavy smoker could die from other causes, not just disease, but also accidents, foul play, suicide, and so on, before developing cancer.
incidentally, i found an interesting discussion of the smoking/cancer correlation here:
http://www.journaloftheoretics.com/Editorials/Vol-1/e1-4.htm
the correlation might be lower than i thought; on the other hand, this article overstates things the other way, because a smoker who dies in a car accident at a young age doesn't prove that smoking doesn't cause cancer, yet the statistics he uses includes such people among the ranks of smokers that don't develop cancer.