Like many others who have been dragged onto facebook by a variety of people, I find myself bombarded by a variety of political posts from my friends still on the lefty side of things. Lo and behold, this post from “The Centrist Word” was popular:
When you think you have seen everything, that is when you find yourself shaking your head in disbelief. Since the general election campaign has kicked off, I have been following politics really closely to see what dirty tricks emerge. Well, I have found one – and in the age of social media, its a big one.
Yesterday I noticed a chart on a right-wing site that had an article that had been posted to Digg. In the post there was a chart. The chart was supposed to show deficit information about the last 6 presidents. I though to myself ‘wow that doesn’t look good for Obama’ but then I noticed something rather odd. The numbers did not match the chart. When I plugged the same numbers into excel, the chart looked completely different.
This piqued my interest. Was this just a one-off, or were there more to be found. It took me all of 15 minutes to find another one. However, this one used a different technique. This chart claimed to use data from the CBO. When I looked the information up on the CBO web site, the data for Obama was off by 25% for 2009 and 45% for 2010.
Mmmhmmm. Heads nod knowingly.
Wait a sec.
The only link in the entire post is to an Obama campaign site — the Truth Team. Funny how a right-wing site has no issue with linking to all sorts of sites, but one labeling itself as “centrist” can’t link to right-wing sites or demonstrate what they’re criticizing.
I see they have a handy list of numerical “untruths” I will get around to unpacking in later posts.
Anyway, as the only link was to the Truth Team, I decided to go there and see which graphs people had in mind. Hmmm, I see some graphs they did themselves, and I think I will make some new graphs, too, using the same data sources.
But come on guys, this is more annoying than vaguebooking. At least gossip blind items are dishing on something juicy. “Somebody’s graphs are sexed up” aren’t enough to go on without you pointing me to the graphs in question.
Here’s my offer: I’ll be a mini-truth team. I loathe misleading graphs, even when they supposedly support my preferred policy positions.
Suspicious of a graph? Post the link. I’ll check it out.
To be sure, all I’ll be checking is whether the graph is being distortionary of official numbers-sources. There are questions of how reliable certain numbers are.
I will NOT be checking out projections, by the way. I’m talking about checking out numbers of things that have already occurred.
So send along those suspicious graphs, whether left or right. If not, I’ll be checking out the items listed in the post I linked, as well as a few items I found on the Truth Team site.
The Centrist Word on May 21, 2012 at 6:03 pm said:
Here is a better idea: Google ‘Obama Deficit’ – do a little fact checking yourself if you are truly interested.
To give a link to a bogus graphs (given the popularity of the post) would have made those false graphs also go viral. Besides, they were numerous and easy to debunk.
Now if we wanted to be ‘propagandists’ it would have been easy enough to create our own graphs and throw up a few sites. That is the way of the internet – it can all be a hoax or a lie – and by the time anyone realizes it, their attention is already somewhere else – impression has already been made. I.E. such a scary time when the truth can mean absolutely nothing.
I am actually surprised in this era of ‘projection politics’ some false lefty charts haven’t already been created and distributed. I am sure they will.
The figures sited in the post are true – would love to debate those if you think there are any lies there.
Meep on May 21, 2012 at 8:19 pm said:
It’s not propaganda to take deceptive graphs in terms of actual amounts (as opposed to frame) and fix those so the scale is correct.
Heck, in this age of google docs, not only can you make graphs without having to buy fancy software, but you can share the spreadsheets and stats that went into making your graphs.
So, simply asserting that there are deceptive graphs out there doesn’t tell me much. Here’s a tip: it’s an election year. There are a lot of distortionary stats being trotted out by all sorts of interested parties.
In any case, in linking to the Obama campaign site (which has some graphs) and then stating various stats in your own post, you’ve given me material I can dig into. Because you didn’t show your work, I get to do so. So huzzah for lazy bloggers.
If you’re not actually lazy, maybe you can share your work with the rest of the class.
The Centrist Word on May 21, 2012 at 8:22 pm said:
BTW – about sending in links? That is pointless – it can all be made up. What needs to happen is that people need to check these things out themselves and report them – left, right, truth team or Romney’s campaign – doesn’t matter. if it is a lie – it should be exposed.
I personally have had habit of checking sources and data because of my work. What I did find – at this moment in time, this was a right-wing phenomenon. I went to liberal sites too and I didn’t find Romney job killing exaggerations or anything of kind that would draw a parallel.
Now, if this makes me a Communist, Leftist, Fascist, Socialist or some other ist in somebodies book – I’d highly suggest going to your local community college and signing up in a Poly Sci course. You will find yourself curled up with your copy of the Communist Manifesto. Then you will how utterly ridiculous you sound to anyone with an education. Sorry, I have have felt like saying that
Finally, I do not subscribe to this BS equalization “both sides are guilty” argument. This 50/50 guilty is rarely true and if one side is more guilty than the other, it is relavent.
Dan Collins on May 21, 2012 at 8:39 pm said:
I don’t believe that Meep posited any kind of 50/50 split, and I’d like to see your homework that shows which side is more likely to abuse statistics, because you seem to believe that one more routinely does.
And your argument by assertion sounds ridiculous, BTW, to this person with an education.
Lo! before you writes a statistician by training and by trade, who is willing to make demonstration if only you will afford her information, and what does she get? Snark and guff, and (more insultingly) snark and guff that asserts its superiority.
Shit, or get off the pot. We await your performance.
The Centrist Word on May 21, 2012 at 9:19 pm said:
I didn’t say that she said anything about a 50/50 split – read what I said again – I was speaking general terms. Kudos for thinking that you were rising to her defense though
You know how to use excel, right – congratulations my educated acquaintance – it is the only statistical tool you will need. Plug in the number and look at the graph – if it looks different – there you go. You dont need to be a statistician to type a few numbers in a spreadsheet.
Shit or get off the pot – thats pretty impressive. You should be really proud of yourself for putting that – it sound sounds very educated.
Dan Collins on May 21, 2012 at 9:27 pm said:
Actually, there are better programs for graphing than Excel, now. Some of them are even available as Web 2.0.
Let me ask you, my educated and refined friend, what you meant by this:
“This 50/50 guilty is rarely true and if one side is more guilty than the other, it is relavent [sic].”
Or did you mean anything at all? You’re a con, a flim-flam, someone who wants to take credit for knowledge and understanding and, above all, authority that they do not have, and do not deserve.
I was not defending Meep. She can do that quite well on her own, me bucko. I’m defending my readers from having to be bored by preening, shallow jackwagons, like you. Adios.
The Centrist Word on May 21, 2012 at 8:24 pm said:
Thanks for the response btw
jefferson101 on May 21, 2012 at 8:34 pm said:
Wasn’t it Disraeli? “Lies, Damned Lies, and Statistics!”
And, be it noted, I am in the Quality Department for a living. I can seldom pass a day at work without resorting to Minitab.
Such being the case, I know exactly what I can do with Statistics, and I’m working with stuff that generally (Let’s say 70% of the time) shows a normal distribution to 3 sigma with N=30 to 50.
I’m not “good” with Stat, in the sense that I can make it prove anything I want it to prove, but I do know how it’s done. If your first run isn’t what you want, change the parameters until it is, right?
That’s what I see done with some of my numbers, if someone has a lot invested in being right about some theory or other.
I don’t trust any of them very much that way, because I know enough about it.
‘Nough said.
meep on May 23, 2012 at 4:53 am said:
btw, in my day job, I do use govt stats (especially from Bureau of Labor Statistics, Centers for Disease Control, Treasury Dept, Census, etc…..). We know of various biases embedded in the data, but that’s fine. We’re mainly looking for trends. The absolute amounts are not necessarily the important thing.
And no, we’re not doing this for political stuff (unless you think trying to provide analysis to and of the insurance industry is political).
An example: the low interest rate environment. We don’t get into commentary as to whether the Fed should be pursuing such a strategy. We’re here to figure out the likely impacts on life insurance and what companies can do to respond. Lots of people do get into what the rates are like now (for whatever reasons), but I decided to look at U.S. rates all the way back to the earliest I could get (in the 1950s) and found an interesting secular trend. I do stuff like this all the time.
So I’m interested in those net job change numbers at the Obama site. People have been going on about the labor participation rate dropping, but I wonder how much of that is caused by long-term unemployment, or whether it’s the demographic bulge moving into retirement. So I will look at that.
Anyway, I’ve actually got two paying jobs so I probably won’t get to this til tomorrow at earliest. I’ve got publication deadlines at the day job, and I’ve got to kick off 5 online actuarial seminars this week, too. So see ya when I see ya.
Starless on May 23, 2012 at 6:50 am said:
We know of various biases embedded in the data
I immediately think of one of my favorite subjects (CAGW), how those who have deflated it the most are statistics nerds, and why those who have made their bones promoting doom get hemorrhoidal whenever they see a FOIA request for their raw data.
jefferson101 on May 23, 2012 at 7:36 pm said:
Trend analysis is easy for me, but as noted, I work with N equaling 30 to 50 samples, typically. Even across a 16 station machine, that’s only 800 values for any one parameter.
And I know my samples are sequential, too.
That’s one of the reasons I am doing what I do…I do my own sampling, and the data points will be sequential, or I will start over with the sample.
If I run 2 pc subgroups off of that? Trends stick out like a sore thumb.
Random samples from a population, though? That gets a bit trickier, particularly when you cannot be sure that your samples are actually “random”.
Other people’s data is always a bit suspect.
Ain’t it fun?
Starless on May 24, 2012 at 6:46 am said:
Other people’s data is always a bit suspect.
IMO, you need to be Reaganesque and Saganesque about that: “trust but verify” and “extraordinary claims require extraordinary proof”.
moish on May 23, 2012 at 6:38 pm said:
Now, here’s the thing. I followed Meep’s link and read the post in question. (I then went beyond that post and read a bunch of stuff on this so-called centrist blog, and I’ll get back to that later)
So the post claims that there are charts that there are charts that make Obama look bad, but when e looked into the numbers e found that the charts were lying. But there are no links. And when Meep asked for links so that she could look for herself, there was a reason not to provide such links. So we are left to basically take the Centrist’s word for it.
Sorry, but I’m not biting. Here’s where it matters that I perused the Centrist’s blog. There’s a link to the Democratic party and one to the “Centrist Party” but none to the Republican Party. And the “Donate” button solicits donations to the Democratic Party. The blog itself is a series of attacks on republican figures and conservatives. Some make valid points. But others are just the same usual unfounded attacks on the current crop of boogeymen like the Koch brothers and Joe Ricketts. It is not a centrist blog in any meaningful sense of the word “centrist.” To be blunt, it’s the liberal equivalent of Weaselzippers except that it’s more partisan. I will admit, though, that the tone on Centrist is less inflammatory than on WZ, even though it is more partisan.
So I’m left with two possibilities:
1) The Centrist is dishonest, using that self-applied label in order to fool the unsuspecting into believing that what’s posted there is fair, and that it will go after lies, dishonesty and hyperbole from wherever it comes, regardless of the
2) The Centrist really believes that the blog is fair. In that case, e is delusional.
Either way, I have little (OK, no) reason to believe Centrist’s claims regarding dishonest charts.
Dan Collins on May 23, 2012 at 6:55 pm said:
This is why I blocked Centrist. I got quickly sick of the twisting and turning to generate an issue entirely different from what he asserted.
moish on May 23, 2012 at 6:44 pm said:
Also, for the record, I’ll note “Centrist”‘s claim above that false lefty charts haven’t yet been distributed. Yeah, right.