4 ms·

And? How does that dispute what's put forth in the article? "The percentage has spiked by almost 10 points since 2007 — in 2014"

by zajd 10y ago

kbart 10y ago
If 1/3 of population is >50 years old, a random subset of this population is expected to have 1/3 of >50 years old as well. After reading this title, one could get an impression that this part is un-proportionally big, which is not.
- zajd 10y ago
  It wasn't as big 7 years ago. I don't get what's so hard about this to understand? The statistic grew from ~25% to 33%. This is noteworthy.
  - kbart 10y ago
    "The statistic grew from ~25% to 33%. This is noteworthy." That's the whole point! Compare to, for example: "Homeless people in America that are over 50 increased by 30%". See the difference? "A third" is just what one would expect from pure statistics.
    - peeters 10y ago
      I'm really confused as to what your point is. Just because over-51s now represent their demographic in homeless population as well, doesn't negate the fact that there was a MASSIVE upward trend in over-51s in the homeless population. If the trend continues, wait a few years and they'll be severely over-represented. Imagine people 50 and under all of a sudden accounted for 70% of Alzheimers diagnoses. Would you say that's not notable, because 70% of the population is 50 or under?
- brudgers 10y ago
  In the US, a meaningful fraction of the population over fifty is receiving direct cash distributions in the form of Social Security [and perhaps other forms of 'retirement checks']. Such direct cash support is much less common among the rest of the population. One way of looking at it is that the rate of homelessness among people over 50 is what it is despite the rates at which individuals within it recieve of direct cash support.
- zwieback 10y ago
  Except it's not a 'random subset', it's the subset of homeless people. If you take the subset of poor people or the subset of sick people I would expect a different distribution than that of the whole population.

And? How does that dispute what's put forth in the article? "The percentage has spiked by almost 10 points since 2007 — in 2014"

by zajd 10y ago

And? How does that dispute what's put forth in the article? "The percentage has spiked by almost 10 points since 2007 — in 2014"

kbart 10y ago

If 1/3 of population is >50 years old, a random subset of this population is expected to have 1/3 of >50 years old as well. After reading this title, one could get an impression that this part is un-proportionally big, which is not.

zajd 10y ago
It wasn't as big 7 years ago. I don't get what's so hard about this to understand? The statistic grew from ~25% to 33%. This is noteworthy.
- kbart 10y ago
  "The statistic grew from ~25% to 33%. This is noteworthy." That's the whole point! Compare to, for example: "Homeless people in America that are over 50 increased by 30%". See the difference? "A third" is just what one would expect from pure statistics.
  - peeters 10y ago
    I'm really confused as to what your point is. Just because over-51s now represent their demographic in homeless population as well, doesn't negate the fact that there was a MASSIVE upward trend in over-51s in the homeless population. If the trend continues, wait a few years and they'll be severely over-represented. Imagine people 50 and under all of a sudden accounted for 70% of Alzheimers diagnoses. Would you say that's not notable, because 70% of the population is 50 or under?
brudgers 10y ago
In the US, a meaningful fraction of the population over fifty is receiving direct cash distributions in the form of Social Security [and perhaps other forms of 'retirement checks']. Such direct cash support is much less common among the rest of the population. One way of looking at it is that the rate of homelessness among people over 50 is what it is despite the rates at which individuals within it recieve of direct cash support.
zwieback 10y ago
Except it's not a 'random subset', it's the subset of homeless people. If you take the subset of poor people or the subset of sick people I would expect a different distribution than that of the whole population.