I have been taking a look at the series and the problem is that the brackets are just not granular enough. If you were born in 1970 you would be 25 in 1995, the year the dotcom bubble started to inflate. That was a really good time for someone with networking experience to be starting out. If you were born in 1979 then you would have been finishing your degree right around the time of the dotcom bubble bursting. Not that that deterred everyone. What I am getting at here is that the structure inside the 1970s bracket may be rather significant. The survey results are consistent with two separate interpretations: 1) There is a nice bell curve and nothing to worry about. 2) There is a large cluster aged 35-45 and a couple of bumps at 21-25 (grad students) and 55-65 (aging Cerf era grad students) Another problem with the survey is that there is clearly a lot of self-selection going on. We know how many people are on the IAB and IESG so we can work out their response rate (>80%). Working out the number of WG chairs that attended is a but trickier but there are 130 WGs and this would suggest a max of 260 chairs attending but not all WGs meet and not all chairs attend yadda yadda. 80 chairs responded to the survey. If we posit a number of 160 WG chairs attending (probably high) then that would imply a 50% response rate. Those figures suggest that the people with a greater investment in the organization are more likely to respond. Now I remember why I gave up as an experimentalist: You ask for figures and then when you have them you find they don't answer your question. On Fri, Apr 27, 2012 at 11:27 AM, Yoav Nir <ynir@xxxxxxxxxxxxxx> wrote: > > On Apr 27, 2012, at 6:05 PM, Carsten Bormann wrote: > >> On Apr 27, 2012, at 16:41, Yoav Nir wrote: >> >>> Before 1950 2.9% >>> 1950 - 1960 16.6% >>> 1961 - 1970 33.7% >>> 1971 - 1980 32.8% >>> After 1980 14.0% >> >> Nice bell curve, יואב, but you can't pop that soap bubble of perception with the bluntness of raw data :-) > > Only 350 out of 1200 people answered the survey, so all caveats about bias apply. It's also possible that the 20-somethings tend to sit in the back more, and go to the microphone less. Maybe them young'uns are too busy clicking "Like" on pictures of LOL-cats :-) > >> Maybe just the areas where PHB likes to work in are growing old? :-) > > The old people in the security area do tend to look older than the old people in other areas. Maybe the bell curve for the security area is different. > >> Many of the people doing the real work in CoRE are in their 20s, or have left that age range just recently. And no, they aren't all academics. I think we have a healthy age mix, with some pretty good gray-haired input as well. >> >> I'm going to argue for an age column on the blue sheets so we get better data :-) >> >> Grüße, Carsten >> >> PS.: Please, don't take any of this seriously. Except for the CoRE age statistics. >> Dave Cridland's observations also definitely don't apply to CoRE, except that we do have the stunning range of experience that makes the IETF so valuable. >> >> PPS.: Is the overall median really 42? > > Wait a couple of years, and most participants won't get why "42" is funny. -- Website: http://hallambaker.com/