Saturday, August 29, 2009

How to liberals lie with statistics

The classic can't hold a candle to liberals' tricks. An occasional commenter at my friend Karol's blog linked to perhaps the most dishonest "study" I've ever seen. My reply:
Stop trying, Toowoozy. You keep failing.

1. You're engaging in the same old fallacy that because country X's population lives to only __ years, that country Y's greater longevity means it has superior health care. It's a fallacy because it assumes every cause of death is preventable and/or treatable. Notice something at the top of the chart? "All causes." So this "study" is just another way of presenting what we already know: American life expectancy is lower because of higher homicide rates.

Try a study that is restricted to natural causes, and adjust for Americans' higher-fat, higher-cholesterol diets. Oh, that can't be done? Well, coincidentally, neither can socialized medicine -- effectively, that is.

2. Presumably he's talking only about South Korea, but to call it just "Korea" is idiotic.

And now the biggie:

3. He's a liar. He so massaged the statistics that they're meaningless.
a. He's using males against the entire population's life expectancy. We already know that American men, on average, die a bit younger than the average life expectancy for both men and women. American women have a life expectancy of about two years more, as I recall, so it's not surprising to see numbers that show American men don't live as long as the average.

To put it in simple terms for you, let's say Jack died at 78, Jill died at 80. This study would look at Jack only, then say "American health care is inferior because Jack didn't live until the average of 79!"

b. He doesn't need to use "per 100,000 males" for this kind of average. He could have easily had it "For each average male." However, he needs this lie to inflate his numbers.

According to his own numbers, it comes to 0.06397 years per American male. That's 23.35 days. The "best" country on his list, Japan, therefore comes to 13.55 days per male.
So this study proves "nothing" except that when it comes to dying, there's more gender equality in other countries.

And you're welcome for the free lesson on how to look past statistics.
Then my P.S.:
Just saw that I forgot one last thing:

You now see (or SHOULD see) that the putz is lying about his data. But let's assume he wasn't. Let's give him the benefit of the doubt, shall we?

His "study" would be comparing the average of 100,000 individuals to the average of that particular population. The difference would be...zero. ZERO. Do you understand that?
Always be skeptical of data that purports to prove something. I don't even ask people to trust things that I cite. Never take anything at face value, or anyone at his word, and especially question hard some clever new "analysis."

Labels: , , , ,

0 Comments:

Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home