43% reduction in child obesity?? A great post by Razib Khan

Razib Kahn writes an article in Slate that I would love to have written.

The Obesity Rate for Children Has Not Plummeted - 
Despite what the New York Times tells you.

Apparently "he is currently a Ph.D. student studying the evolutionary genomics of the domestic cat" however I think his real calling could be debunking poorly explained statistics. Even the New York Times, erstwhile home of Nate Silver, is not above some shoddy analysis.

Thanks Razib!

http://www.slate.com/articles/life/family/2014/02/obesity_rate_for_children_has_not_plummeted_despite_that_43_percent_headline.html

Missing a key data element ...

On April 1st from Rick Newman at US News and World Report published "Why the Middle-Aged Are Missing Out on New Jobs"

To summarize the article: based on the latest US job report, 1.9 million net new jobs were added over the past 15 months from Jan/10 to Mar/11. However the 35-44 and 45-54 age groups have lost jobs, while all others gained about 2.4 million jobs. These are facts. The rest of the article goes on to speculate on why this might be occurring, and what are the implications for the US economy.

I'm in my late 40s, so this news was unwelcome. Fortunately as a small business owner I created my own job along with a few more so I didn't take the news too personally...

Mr. Newman's speculation is quite reasonable. But ... can the "facts" be taken at face value?

These facts are ignoring the change in size of the age groups. I'm no demographer, but I am aware of the Baby Boom. As a proxy for change in age group size, here are the US Birth rates that would impact the age groups. Over the 15 month period, and ignoring immigration, deaths, etc., the 35-44 age group has shrunk by about 800K (645K*1.25) and the 45-54 age group by 430K (344K*1.25). Admittedly rough calculations ...

However, this implies that the 35-44 age group has shrunk faster than their job pool (-800K to -143K). It remains a fact that this group has fewer jobs than 15 month earlier. But, since the number of people in the group is notably lower, it's also a fact that a higher percentage of people in this group now have jobs. So I don't think it's reasonable to conclude that this group has any kind of hiring issue. For the 45-54 age group the numbers are close (-430K to -454K). Again, on this basis not likely a hiring issue.

Finally, I recognize that this post is not doing a lot better than the original publication ... I'm not accurately comparing the CHANGE IN SIZE of each age group with CHANGE IN NUMBER of jobs before I draw a conclusion. That's what is needed. My excuse is that I was up early today, so I took 30 extra minutes to write this post. I'm respectfully hoping for larger time investment and for better quality from the US News and World Report.

All comments welcomed. Marv

PS. For ease of reference, I've attached the relevant statistical breakdown from the article:

Age group	Job gains last 15 months	Unemployment rate
All adults 16 and over	1.9 million	8.8%
16 – 24	490,000	17.6%
25 – 34	709,000	9.1%
35 – 44	-143,000	7.2%
45 – 54	-454,000	7.1%
55 and over	1.3 million	3.1%

(Note: The broken-down job numbers don't completely add up to the total due to seasonal adjustments and other factors.)

PPS. Yes, I could probably find a better proxy than the US Birth rate numbers with some more research ... feel free to help me out ...

Today is the day I couldn't stand it any longer ...

I read the financial news this morning by spending 30 minutes browsing some of my usual favorites ... those include the Wall Street Journal, the New York Times, Yahoo Finance, and Seeking Alpha. For better or worse (mostly worse), I am one of those people who does a lot of mathematics in their heads when they read an article that tries to use statistics to make a point. I fundamentally believe the quote popularized by Mark Twain "There are three kinds of lies: lies, damned lies, and statistics!"

However, there is one "use" of statistics that I find even worse than misusing them to support a point. That's when someone uses a statistic and clearly has no idea of what it means or whether it is even relevant to the point they are discussing. I've named these "damned statistics", in honor of Mark Twain (or perhaps Disraeli or ... see Wikipedia entry for "Lies, Damned Lies, and Statistics").

For personal therapy, I'll be pointing out some egregious [mis]uses of statistics ... I am not sure where this blog will go. I do know there is a LOT of fodder out there to work with!

Anyway, as you can see, I am not really a Twitter guy, but I will try to keep this brief. Enjoy