Log in

No account? Create an account
color cycle (slow)

Kistaro Windrider, Reptillian Situation Assessor

Unfortunately, I Really Am That Nerdy

Previous Entry Share Next Entry
Meaningless Numbers
color cycle (slow)
There are three types of lies: lies, damn lies, and statistics. -- Samuel Clemens, A.K.A. Mark Twain

From http://story.news.yahoo.com/news?tmpl=story&cid=572&e=14&u=/nm/health_marijuana_dc

"And it found the age of youths using marijuana is falling. The teens aged 12 to 17 said on average they started trying marijuana at 13-1/2. The same survey found that adults aged 18 to 25 had first tried it at 16."

Does anybody else see what's wrong with this number?

Here's the thing: It's comparing a difference between two distinct populations, when that which decides between populations is numerically related to the difference.

Think about this. If you're polling those in the age range of 12 to 17, it will be impossible to get a value of 18 or higher. All numbers will fall in the range from zero to 18.

Polling adults? The range is now zero to 25. So, perhaps, the exact same data point count could be plotted over the curve of those who answered in the 12 to 17 poll- but then there are the additional possible values, pulling the number to the older. The number given is completely meaningless.

Great reporting, Yahoo.

  • 1
Psych courses + math courses + Yahoo stats = Adam wins.

What might be a more valuable statistic is to take the numbers from the range of 12-17 year olds, and then take it from 18-23 year olds (same time-span), and discard any number of 18 or greater. (That's essentially what's being done with the other group- points of 18 or greater haven't happened yet.) That shows what the same poll would have given five years ago; repeat the procedure for those of ages 24-29 and again discard responses of 18 or greater. That's the only way to get a meaningful number from this.

  • 1