Tag Archives: statistics

Numbers don’t lie, politicians do

Posted on by

In this debate season we are hearing all sorts of numbers from politicans that tell us how wonderful they are and how evil their opponents are.  If you don’t care much for the truth, take the numbers at face value and stop reading at this point.

Since you care about the truth, let me tell you a story of numbers to show how accurate numbers can be used to deceive.  When I was a lieutenant in the Army in the 1980’s, I played goalkeeper for my post’s soccer team.  I was a very average goalkeeper.  Because Ft. Bliss was a NATO post, we got to play teams from other nations.  One game that we played was against the team from the United Arab Emirates.  If you were to ask me how our team did, I could have honestly answered by saying, “I stopped 50 shots on goal.”  That sounds impressive.  If you were to ask me a follow-up question like, “How many shots on goal did they take?,” to be honest I would have answered 63.  With one answer, I sounded ready for the U.S. World Cup team.  With the other answer, I sounded to be as porous as Spongebob Squarepants. One last question would give you some real meaningful information.  It is “How many goals did your team score?”  The answer is 1.  We lost 13 to 1.

Mr. Obama claims that during his administration 4.5 million new jobs were created.  That number is correct.  However, the combination of workers retiring at an older age and the addition of new workers to the job force mean that to break even, 90,000 new jobs must be created each month.  He has been president for 46 months.  46 times 90,000 equals 4.14 million.  4.5 minus 4.14 equals 360,000 net jobs added.  Claiming 4.5 million new jobs when people think that a net gain of 4.5 million jobs occurred is a lie.  When someone asks me how my team did in a soccer game, that person wants to if we won or lost.  If I tell that person that I stopped 50 shots on goal, I am lying to that person because I am making that person think we did well.

Here is one more numerically accurate lie.  The Bureau of Labor Statistics reported unemployment at 7.8% at the end of the summer, something Mr. Obama took credit for.  That number was somewhat accurate depending on how you define “employed.”  The lie is tied up in the fact that millions of college students who were employed full-time through the end of August left their jobs to return to school.  A reasonable person hearing the number 7.8% unemployment in mid-October thinks that is the current unemployment figure.  It isn’t and that’s why it is a lie.

Advertisement

A Second Open Letter to Sen. Tom Harkin

Posted on by

On September 24, 2012, I posted the following letter to Senator Tom Harkin’s Senate Web page.

________________________________________________________________________

Senator Harkin:

As a veteran and taxpayer, I appreciate your desire to have the money used to educate our veterans be used effectively and efficiently. However, the presentation that you authored has issues. I’m trained in statistics and because of that, I could make a laundry list of issues. You are a very busy man so instead of going item for item, please allow me to point out one blatantly wrong slide. You have a slide entitled “For-Profit Colleges Employ Many Recruiters But Few Placement Staff.” From a subjective perspective, to have a slide with a title that indicates an issue with For-Profit Colleges as a group but then indicts a single one is disingenuous and ineffective because your case is that many schools have this problem, yet you showcase only one. It also smacks of a lack of objectivity and personal bias. Even if your chart is accurate, it does not make the point that this is a systemic problem.

Objectively, the slide in incorrect. University of Phoenix, the largest school under the Apollo Group umbrella, has an extremely robust Career Services site that covers the gamut of services from resume writing to career research to interview preparation. The site also lists major employers that Phoenix partners with to help find jobs for its graduates.

Your data was gathered from 2007 to 2010. I started at Phoenix in 2010 and I don’t know what its career placement services were like then. It could be that in 2010, your statement was accurate. However, it is nearly 2013 and it definitely is not accurate.

If I had the 20-30 hours free to write it, I would write a report that shows slide for slide, just how bad this report is. I don’t have the time to write it and anyone as busy as you doesn’t have the time to read it. Please consider making this “update” to your report as a show of good will. I’ll be the first person to thank you for that by making a post to that effect on my blog, andyknaster.com.

Since your Senate Web site has a link to that report on its home page, it appears that this is a report you are proud of and feel the people need to read. Don’t you owe it to your constituents and all of America to be accurate and up-to-date?

This is my second letter to you. I requested a response to the first one and I have not received it. I posted to your Facebook page and got no response. Please respond to this message. This is a very non-partisan issue. In this highly contentious and partisan time, taking action on a non-partisan issue is something I think would resonate well with many Americans.

Respectfully yours,
Andrew Knaster, BA, MA, MCP
University of Phoenix Master of Information Systems student, class of 2013

Statisticians and actuaries

Posted on by

A good statistician knows how to not lie and yet fail to tell the truth. The same holds true for actuaries. The only difference is that a statistician can use data to convince you that what didn’t happen, did, while an actuary, armed with the same data, can convince you that what won’t happen, will.