Net54 age distribution

[BicycleSpokes]

Thank you all so much for the great voter turnout!

After an "official" 24 hour sampling period, here are some results for the Net54 age distribution poll:

Total population sampled: 289
mean age: 45.4 years
median age: 44.5 years
standard deviation: 11.6 years
skewness: 0.132

For the mean and other statistics, I just assumed a mean value for each age group i.e. 14.5, 24.5, 34.5, etc.. While not perfect, this is easier than compiling the nearly 300 values by hand! (and also gets to within a few years of the mean values Leon mentioned previously in the thread, as a gut check.)

More than the mean, I am personally actually most interested in the skewness, which is a measure of how asymmetric the distribution is about the mean. The slightly positive skewness means that the distribution is actually "leftward leaning" compared to a perfectly symmetric normal distribution i.e. leaning towards the younger population (e.g. the 30s outweigh the 50s; also the "most likely" median age is less than the mean). I was actually expecting the opposite, so this is a pleasant surprise. This in particular would seem to point to a rather healthy, possibly growing, active Net54 population. In fact, given that many potential collectors won't have much disposable income until around say 30 (at least I know I didn't), this may create an artificial lower age barrier to entering the hobby, hence the leftward lean might be even more pronounced than indicated by the raw statistics (I am only speculating here). On the other hand, as many of you noted, the online sampling might also skew young compared to the pre-war collecting population as a whole. In any event, a graph of the 24-hour sampling results is attached as a JPEG file, in case anybody is interested.

If it doesn't annoy you guys too much, I would like to repeat this "census survey" at approximately annual periods, and continually add the new yearly results onto the chart below. I don't expect much difference year to year necessarily, but over say 5 or 10 years, I am curious if any noticeable and consistent drifts can be observed. (Sorry, I am an engineer / scientist, and just like doing these sorts of things! Also, anybody who has actual expertise in statistics, polling, etc. is very welcome to weight in!)

Thank you all again for participating (the poll remains open).

Cheers,
David