When bad things happen to misguided emails

December 8, 2005

News websites routinely run online “polls” in which readers can weigh in on some controversy of the day. It’s just a web gimmick — the polls aren’t scientific, don’t even pretend to fine representative samples, and have zero predictive value. They’re just supposed to be fun.

Validity aside, these “polls” tend to get attention. For example, late Tuesday night, the Hartford Courant posed a question to online readers about whether they’d back Joe Lieberman or Lowell Weicker if the two squared off in Connecticut’s Senate race next year. Someone in Lieberman’s office thought it was time to rally her troops and make sure the results worked in the senator’s favor. Little did she know that Kos would end up posting her email.

From: Sherry Brown
To: ‘Sherry Brown’
Sent: Tue Dec 06 21:31:36 2005
Subject: Sorry for the blast email….

But the Courant is doing an online poll in which we have just fallen behind between Lowell Weicker and Joe Lieberman. We’ve been ahead all day, but one of the internet bloggers got hold of it and is promoting it among the extreme lefties.

Go to the Main Page – see Joe’s picture and then there is a link “Weicker – the antiwar candidate?” click onto that page – scroll down – right hand side there is a poll.

Let’s turn it around!!! Thanks. (and please note that I did NOT ask you for money, which may be a first)

Oops. As of this morning, Weicker leads Lieberman, 82.3% to 13.4%.

And as for the chief political aide of a Democratic senator casually dismissing “extreme lefties,” I can only say that people really should be more cautious about their choice of words in blast emails.

So that’s what Joementum means!

  • Bimbo. Obviously doesn’t understand that email are so much easier to get around. Guess she isn’t as nearly tech savvy as she thinks she is. Of course maybe she is a bit too trusting and that maybe some of her friends/Lieberman supporters are nearly as big supporters as they are supposed to be.

  • Also note the hypocrisy: She’s saying the poll is skewed because someone is stirring up votes for Joe’s competition WHILE trying to skew the poll by stirring up votes for Joe.

    This “extreme” lefty thinks it stinks like GOP.

  • Well, with luck, Lieberman can become SecDef in January, then fly to Baghdad where his airplane gets shot down by one of the 750,000 shoulder-fired anti-aircraft missiles currently available world-wide that BushCo doesn’t want to spend money to buy up and get off the market – and the legislation to force a change in this policy presented by Senators Obama and Lugar has been opposed by Lieberman. Thus, it’ll give a whole new meaning to “hoist on his own petard.”

    If there’s anyone who’s a DINO, it’s that tiresome whiner Lieberman.

  • Tom Cleaver – love the (accurate) reference to Leiberman’s petard.

    These “voluntary polls” shouldn’t even be called polls. I don’t what they measure – the willingness of groups of monkeys to punch the keys provided?

    Even “real” polls, designed to produce valid responses (whatever that means) from a representative sample (however obtained), are subject to sampling error (i.e., if you took an infinite number of such polls you’d get erroneous results x-percent of the time). And even “correct” results may be “statistically significant” but of no practical value in making a specific decision — most of us can’t resolve “to be, or not to be” decisions by rolling the dice an infinite number of times).

    My statistics professor in grad school (Univ of Oregon) was W.S. Robinson. He invented “fallacy of ecological correlations”, was the first to do statistical analyses of jury behavior, developed the advertising campaign (Listerine?) “your best friend won’t tell you (about halitosis)”, etc.

    He once demonstrated for us that the simple marketing question, “Why did you buy this dictionary?”, had some 26 different, valid interpretations – why did you BUY this dictionary, why did YOU buy this dictionary, why did you buy THIS dictionary, etc. Even for this simple question, there’s no way to know which meaning the respondent is responding to.

    While still a student I was hired by a researcher in the university’s PE department. He “believed in” somatotypes (classifaction by body type – endomorph, mesomorph, ectomorph) as a determinant of attitudes and behavior, and he wanted me to design that into questionnaires for students. As a pre-test, I used the lists of “personality traits” he suggested, but instead of having students pick which human sillhouette (fat, muscular, skinny) fit the characteristics, I used pictures of flower pots (round squat bowl, traditional semi-conical flower pot, bud vase). The students unhestitatingly attributed such traits as “lazy” or “heroic” or “frightened” to those innanimate figures.

    The point is that we often wind up creating measurements of things which didn’t even exist until we’ve measured them. The “ontological reality” of public opinion is, at best, very iffy.

    Beyond that, there’s the question of whether people know what they’re talking about. Back in 1956, when the New York Times was still a newspaper, they polled the public on Eisenhower’s “brinksmanship” policy (he stated that if one Russian foot soldier set foot in West Berlin, we would respond with thermonuclear weapons on the USSR). The public overwhelmingly (mid-80 percents) supported Ike. the NYT then asked the responents where Berlin was (a. in East Germany, b. on the border, c. in West Germany). Seventy-five percent of those supporting the president failed the test.

    And even if the public knows what it’s talking about, and expresses its will clearly on an unambiguous research instrument, a decision-maker can change that opinion overnight by simply acting (FDR closing the banks. Truman integrating the military). I wish the national Democratic Party understood, and had the courage to act upon, even a fraction of all this, but they don’t.

    • Comments are closed.