Statistics

posted January 13th, 2019 at 10:10 PM

I keep reading these posts and comments that quote some statistics that purportedly support the poster’s viewpoint. However, as anyone who has ever studied statistics would know, virtually every “study” can be manipulated to produce the results one is looking for. There’s also that issue of what specific stats you’re referring to. The mean? The median? Which is it? What was the margin of error? So, to the people who consistently quote these stats, if you feel so compelled to do so, kindly include your study references so that your numbers can be verified. If you can’t do that, then all you’re doing is stating your own opinion, just like the rest of us.

3 Comments

Yes

Jan 13, 2019 at 11:30pm

Nailed it!

Sam Clemens

Jan 14, 2019 at 12:16am

There are lies, damn lies and Statistics.

Meaningless

Jan 14, 2019 at 12:35am

Did you know that 77.3 percent of all studies referenced on the internet are fabricated? No?
Hm. It may or may not be true, but I just made it up.
Let's put aside the fact that most of the population is too innumerate to understand the distinctions you're trying to make, and the field of statistics more generally. The terms you use are hardly understood outside mathematics, mean and median being just one example (admittedly, I chose a critically useful and relevant one).
I'll cut to the quick.
People believe what they want to believe. I could quote studies and statistics at you, but this flaw in human thinking has been widely known and empirically established for millennia. Belief trumps (ahem) facts. Almost always. When confronted with contradicting facts, belief merely hardens into certainty.

So to sum up, it would be a waste of resources of the writer and publisher, and counterproductive for most readers, to provide all the statistical context. That generalization is valid within the normal margin of error :-)
People who want to believe particular statistics (and the conclusions they suggest) will agree. And the skeptics will deem it fanciful or suspect, no better than sophistry. Facts be damned.