Do not trust any statistics you did not fake yourself

[Corny]

This time it's a masterpiece

http://mediamatters.org/blog/201006280019

 

[laikaNYC]

Which part is fake exactly?

 

[Corny]

the interpolation

 

[rareboy]

^^ Yeah. Any idiot can see it is all about the curve, not the pure numbers.

 

[CowboyBob]

Does Fox really skew data and report lies?

My favorite was when they edited out the applause at President Obama’s commencement speech at West Point. Now they call it a "technical error".

http://mediamatters.org/blog/201005240010

http://www.huffingtonpost.com/2010/05/25/fox-news-video-omits-appl_n_588447.html

 

[MercuryJones]

Which part is fake exactly?

The title of the chart, for one thing.

It claims to show "Job loss by quarter," but doesn't.

 

[laikaNYC]

The title of the chart, for one thing.

It claims to show "Job loss by quarter," but doesn't.

Are the numbers fake?

 

[MercuryJones]

^

Moving the goalpost.

 

[laikaNYC]

That's what I thought

 

[Zu-Mendel]

If that was a poor attempt at sarcasm then you really don't know how to lie with statistics, do you? Seriously, take a Stat course at your local community college; the first 2 chapters of my Stat 101 book detailed how you can twist any number to say anything you damned well please.

 

[opinterph]

... you can twist any number to say anything you damned well please.

That might work if nobody checks your methodology. The problem with most statistical misrepresentations is that whatever entity reports the finding doesn’t make available the process they used to achieve the results. If nobody can check my calculations, then its not the number that is saying “anything” – it’s me.

 

[JayHawk]

It is obvious the flat line after the ecnomy collapsed completely is what FOX fails to represent... In toher words the real statistic I see by comparing the two is that Obama's job spurring massive spending has done nothing.

Similar to the Japanese correction...they tried to fix it with cash and it started a 10 year long recession.... We are headed for worse by scale I am afraid.

 

[MercuryJones]

In toher words the real statistic I see by comparing the two is that Obama's job spurring massive spending has done nothing.

And how do you see that?

The trend line is that the number of unemployed are increasing sharply from about June 2008 to about June 2009, then hits a rocky plateau.

And doesn't that period cover only a small amount of the spending?

 

[Corny]

If that was a poor attempt at sarcasm then you really don't know how to lie with statistics, do you?

hu? I know there are a gazillion of ways to lie with statistics. What is your point? Are you telling me this graph isn't misleading because there are better ways to mislead or what?

 

[JayHawk]

http://mediamatters.org/blog/201006280019

And how do you see that?

The trend line is that the number of unemployed are increasing sharply from about June 2008 to about June 2009, then hits a rocky plateau.

And doesn't that period cover only a small amount of the spending?

SO the period of spending isnt reflected in the virtually flat line from June 2009 to June 2010?

Not creating jobs after spending Billions of dollars to do so would seem as if it was not working.

 

[opinterph]

SO the period of spending isnt reflected in the virtually flat line from June 2009 to June 2010?

Not creating jobs after spending Billions of dollars to do so would seem as if it was not working.

The trend line demonstrates a remarkable change in the period(s) associated with late 2009. That invites the question, “Why?”

 

[MercuryJones]

SO the period of spending isnt reflected in the virtually flat line from June 2009 to June 2010?

Not creating jobs after spending Billions of dollars to do so would seem as if it was not working.

The period of June 2009 to June 2010, when joblessness hits a rocky plateau after the upward trend coincides with the introduction of the Recovery Act. But it's not the entire distribution - in fact the combination of cuts and spending is the least of three years in 2009. Those elements in 2010 are more than twice what's in 2009, and a bit more in 2011 than 2009.

So this data can't conclude that the Recovery Act didn't deliver, because it accounts only for the start of the intervention, not the whole.

 

[Zu-Mendel]

hu? I know there are a gazillion of ways to lie with statistics. What is your point? Are you telling me this graph isn't misleading because there are better ways to mislead or what?

I was responding to laikaNYC's post directly above mine.

 

[JayHawk]

So the info doesn't point to it not delivering...t hat is to come at a later date. Is that going to be before or after the ten year recession cause by trying to prop up the losses in a free market correction?

Will Obama get credit in ten years or will it be whatever hack is in office?

Finally, with the ongoing oil spill catastrophe you cant possibly believe that the even marginal improvements Obama's PR was trying to push out will be maintained. Or do you?

 

[JayHawk]

The trend line demonstrates a remarkable change in the period(s) associated with late 2009. That invites the question, “Why?”

Certainly you can say that the spending coincided with a flattening of the trendline. However was that due to a bottom or due to the spending? Dont think it will ever be possible to assess that.

We can't assess it because we are about to roll into a much wider and deeper recession.