Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

grantcart

(53,061 posts)
10. I hope your reading skills are better than your math skills
Fri Nov 9, 2012, 05:49 PM
Nov 2012

Unfortunately I have only a few minutes to respond so I will go to the meat of your reply



"I am not a poster"



Before my article in October Doug Kaplan had gone on a long extended campaign of serial lying and self promoting that had created a completely fraudulent public image of a national pollster with years of experience and lots of field experience.

Listen to this interview:

http://english.ruvr.ru/2012/01/20/64250633.html

In it he talks about polls he has done and polls he is releasing that day, about field work he did in Iowa. My favorite however is at the 22:09 he states verbatim

"I have run a number of campaigns both at the Presidential and local level. If I had been in charge of that campaign then the next day we would have had a press conference and declared victory.

He kept this lie going for several months when in August he started to issue polls and these polls were accepted into the mainstream as being from a pollster.

After we exposed multitudes of lies about Kaplan he then states the following in the interview

1) He has no formal education in statistics or political science
2) He has not in fact worked as a campaign manager for Presidential or local campaigns
3) He has not worked for any polling company

He has how ever

A) conceded that he evolved his business from making the calls to asking the questions
B) admitted that early polls (which would have been the polls we were complaining about in September) "were not great"
C) had to hire a "statistician"
D) hired a Washington DC insider as a political adviser


A pollster is someone who either through academic or professional experience conforms to the standards and practices established by one of the professional peer groups. In the context of Gravis Wang referred to the particular standards and practices that he would expect him to comply with.

Kaplan is a guy who lied for months about his background and experience and only when we called him out on it did he admit that he, himself doesn't have any of that.

Kaplan admits that he doesn't have any Academic, or professional experience. He isn't a member of any professional polling associations for pollsters. Hiring a 'statistician' and a 'political' adviser doesn't make you a 'pollster' it is simply a guy that hired a 'statistician' and a 'political' adviser. And the fact that he now admits that some of his polls are 'not very good' raises the obvious question "if you know the business why would you release a poll that is not very good, and how did exactly when and how have you gotten the ability to know what is a good poll". Now it is entirely possible that he is getting help from a competent pollster/statistician now. That still would not make HIM a pollster just somebody that hired a pollster.

Finally the ultimate test of a professional polling firm is that you get paid for your work. Kaplan has stated that all of the polls are on his dime and that he makes no money from them.

So my snarky pithy "I am not a pollster" is not a verbatim quote its what you get when you add a person who says that he has

no academic experience
+
no professional experience
+
is not compensated for his polls
+
doesn't subscribe to any peer review associations or up hold their practices
+
has to hire statistical and political competencies because they are outside of his own
+
admits that some of his polls "are not very good".

Now if you place his statement that he "ran a number of campaigns both at the presidential and local level" and the dozens of lies that he has made about his polling prior to August in juxtaposition to my "I am not a pollster" I think my slightly exaggerated snarky simplification stands up.

But if you are unhappy with it then that it is a criticism of my approach, which is fine. It does not change the fact that Kaplan lied extensively about who he was and what he did and his professional abilities.

And if it hadn't been for my blogging (not reporting, I would have been happy to have a reporter take the leads that we had and the leads we still have about other allegations by companies that Kaplan worked for and his associates). But if they won't do it then we will continue to expose the lies and misdeads. Kaplan did not say that any of the charges that we made were untrue, just that they were unfair.

As for Kaplan's references to where he studied you missed the point completely. It isn't that he doesn't have formal education (Neither does Rasmussen, atleast in statistics, but he has learned the business and I frankly don't have that big of a problem with Ras. He has a bias but if you factor that in atleast you know what you have. The criticism about Kaplan's 'education record' is that In various CVs he lists completely different schools and apparently cannot keep his lies apart. Also if you are going to lie about what school you are going to why would you pick one that is itself the target of investigations.

A pollster is providing information. If everything that you can check out about that pollster can be proven to be not true then why would anybody rely on the things that you cannot check to be true?



New Subject



Paleologos' alleged crime is that his polls favored Obama so he stopped polling. What you haven't explained is why a small time pollster ceasing polls in a highly competitive marketplace consisting of dozens of pollsters is suspicious or important at all. Furthermore, why would you be upset that a person you think is ideologically compromised stopped polling?


You must have skimmed the post very quickly. I don't think that Suffolk University is compromised, I thought that they had quite a reputation.

The charge against David Paleologos is not that he stopped polling because Obama was ahead. The charge is that he went from reporting what the people are saying via the polls to trying to help create a meme of what the polls were saying in order to influence what people should think. He mentions that the data that he has seen allows them to "paint the states red" when in fact his own polls show the contrary. If he has the data then release the data.

In any case an academic pollster should be encouraging polling and more information not try to discourage it. Then to go on to partisan media and try to deliver slanted analysis under the guise of an academic standing is highly unethical. The fact that you didn't pick that up from the extremely strong reaction from his peers is interesting in itself.


Finally



Polls can and should be judged on their statistical merits. You have not done so with Gravis. Notably, you linked to an analysis you claimed came from a statistician that contained a statistical criticism of a Gravis poll. The anonymous author never claimed to be a statistician and his or her analysis was not statistical but arithmetic and demonstrated a severe misunderstanding of poll weightings.



I don't have a statistical degree and never offered an opinion on the statistical merits of Gravis polls. I assumed that if a person was going to completely fabricate a poll they would still create a viable spread sheet and that the numbers would be real, even if the people were not asked so I never addressed the statistical question. Sam Wang did. He did say that he thought that there might be an insignificant weighting problem that would account for the questions people asked and that would account for it. That the statistical questions raised could not rule the poll a fraud by a simple statistical analysis.

But more to the point is your "Polls can and should be judge on their statistical merits". This is frankly a very sophmoric view of polls. I don't have a statistical degree but I do have a political science degree. Polls are a blend of political science ( or sociology or psychology depending on the subject). Polls cannot be evaluated simply on the basis of their 'math'. Even in the first year 101 class you can see how you can get strongly you can influence how the question is structured, what vocabulary is used and so on.

For example this NYT article explains that the question "Who do you think will win" is a better question to ask people than "Who are you voting for



http://www.nytimes.com/2012/11/02/us/politics/a-better-poll-question-to-predict-the-election.html

A version of that question has produced similarly telling results throughout much of modern polling history, according to a new academic study. Over the last 60 years, poll questions that asked people which candidate they expected to win have been a better guide to the outcome of the presidential race than questions asking people whom they planned to vote for, the study found




I am sure that have lots of other points to bring up that you think are relevant but I will probably not be back much before Tuesday, so don't interpret the timing of my response as my lack of interest.

Here is the bottom line:

Kaplan was and is untruthful. If you can disprove virtually everything about a person that you have access to why in the world would you believe any information he has about things you cannot check?

David Paleologos' actions in November clearly violate ethical questions on both the academic and the polling front and it shocked people when he did it. He needs to answer for them. I don't think we should look the other way. If you don't agree on that point we aren't going to have much to talk about.
Latest Discussions»General Discussion»DU Exclusive: Paleologos'...»Reply #10