Please purchase "Trump the Press" through Create Space.

The book is on Kindle. Order here.

Tuesday, October 25, 2016

When I am polled

As one of the last landline owner (I have no cell phone) I take a lot of calls from pollsters, mainly local races. When I was working, I politely told the caller of my newspaper work and got excused. Retired, my opinion is viewed as worthy. My experiences offer why I am skeptical of polls.

I get a lot of calls. Most are about the local state elections.

On a national poll, the odds would be around one in 320,000 of being called in any poll of 1,000 people, which is usually the target size. Given 1,000 polls (an arbitrary number) by universities and news organizations in this election, that makes the odds one in 320 to be called this year.

Going back to what I learned when we did this at Cleveland State 35 years ago (the format has not changed over the years) for every call that counts, you make two or three calls.

That cuts the odds to one in 100 of being called once for a poll.

But given that I have taken ten or so national surveys, that means a lot of more people are being called -- and the people reached are not so random.

Now that does not invalidate the polls. But this is a problem. They may be overfishing the pond.

If it is by design -- and the Los Angeles Times uses the same pool of 2,500 people to catch the changes in the election campaign -- then disclose it. But I really doubt this is the plan.

The second thing I noticed is the calls now are all automatic. You respond by pressing numbers. That is a vast improvement because it eliminates one source of human error. You get the same questions asked the same way each time.

But punching numbers is a problem. I make mistakes and hang up. Why not use voice recognition? Expense? Seriously? ABC, CBS, CNN, Fox News, and NBC have so much money, they blow their noses in fifties.

The third thing I notice is that whoever is doing the Morgantown-area polling for Republicans needs to learn what an exchange is. I live on the other side of the state. My exchange is 755. That's the first three numbers after the area code. The people in the 755 exchange live in the Cross Lanes-Nitro-Poca area. The only thing we care about in Morgantown right now is the WVU football team. Unless you are asking me about the team, I don't care.

I also got a call from a Virginia pollster.

The fourth thing is push polls -- which are illegal -- live on. Democrats are using them against state Senator Chris Walters and state Attorney General Patrick Morrisey. Makes me more determined to vote for them.

@@@

Please read "Trump the Press," a fun romp through the Republican nomination that uses the deadliest weapon to skewer the media experts: their own words. "Trump the Press" is available as a paperback, and on Kindle.

13 comments:

  1. Households with landlines have decreased a third in the last 4 years.

    ReplyDelete
  2. Overfishing the pond is much like what is done in medical research where studies concentrate so much on people with diseases or physiological difficulties that they draw conclusions about what normal physiology is and create advice for healthy people based on their findings in a diseased population. That's how we got the "Salt Kills" campaign of the late sixties and on and on for years.
    It is also one reason for how we get medicines that do perfectly well in all phases of testing and then are found simply not to work in clinical practice.
    I'm not saying that I have a solution to this. When you are trying to come up with a solution to a problem you use the tools that are available to you. Creating a new tool is not easy, especially for someone not dealing directly with the problem. But keeping blinders on when you know you have a problem and not looking for help from the outside is stupid.
    A while back a guy who used to work with my dad said that one time they were running a code on a guy in the ER. They were doing everything they were supposed to be doing. He said my dad stepped back and said, "Things aren't responding normally. I don't like this. Does anyone have any suggestions?" He said he'd never seen a doctor ever do this before. I forgot to ask if the guy lived or not. He was making the point that my dad was not your typical doc.
    We need some atypical pollsters.

    ReplyDelete
    Replies
    1. Fact: Salt intake increases blood volume
      Fact: Blood volume is related to blood pressure
      Fact: True Linear relation BP/stroke from the non-walking dead*.
      Not Fact: Your high BP will lead to stroke if you're the walking living.

      Fact: That will not get you grant money for summer salary, reduced teaching load, and overscaled salary.

      *Glenn/Abraham : RIP

      Delete
  3. We also have caller ID, and don't pick up if we don't recognize the phone number or name, or they don't start talking to leave a message. I do kinda like the automated calls that start running during my message to leave a message.

    ReplyDelete
  4. Pollsters conduct polls to satisfy the requirements of the people who pay for the poll. (Quite an insight, eh?) If a lying-media organization (NBCABCCBSNPRFoxCNN) is paying for the poll, then you know it is being done to “skew the result in the Harpy’s favor to depress turnout for Trump”.

    The Harpy has her own pollsters and she is paying them to tell her the truth, which is that she is in mortal danger of losing this election in a landslide. That why we have so much money still being spent by the Harpy on battleground states in an election that the lying-media tells us the Harpy has already won.

    Steve in Greensboro

    ReplyDelete
    Replies
    1. A great piece by Big D, and a great reply by you, Steve! As a guy born in 1958, I can tell you that technology has improved my life in a whole bunch of ways. But in a few cases, Old Skool is still the Best Skool. How did they poll back in the 1930s? Door to door, if at all. You're much more likely to give straight answers face to face, as opposed to some robocall, IMHO. When Saira Blair, my WV state rep, ran and won two years ago, she sent out over 800 HANDWRITTEN notes to voters. That's Old Skool...

      I also think that all this early voting BS has gotten out of control, and that we should return to paper ballots. But perhaps Big D can tackle that topic next. But as far as your comments go, Steve, I am in total agreement. There's a Reality Disconnect here. DJT continues to draw MASSIVE crowds while Hillary is getting less people than a Thursday night b-ball game at Hedgesville High School. The polls are just not making sense to this country boy...

      Delete
  5. The Two Question Poll.

    First, truly random calls required; land and mobile, not selected by any party affiliation. Using Don's voice recognition. *This still excludes individuals on Do Not Call lists, but stay with me here.

    Q1. "On a scale of 1 to 100 how likely are you to vote in the presidential election November 8?"

    (exclude all responses below 75%)

    Q2. "From the individuals below please indicate who you will vote for. Press 1..X, 2..Y, 3...Z."

    3. Thank you message. Hang up. Now, analyze the data from 100,000.

    You'll never see that because of what Steve noted.

    ReplyDelete
    Replies
    1. The Do Not Call list does not apply to pollsters, any more than it does to people soliciting for charity. It's strictly for sales calls.

      Delete
  6. “Statistics are like loose women. You can do anything you want with them.”

    ReplyDelete
    Replies
    1. You can even grab them by the p***y. [/a thousand mea culpa]

      Delete
  7. The only thing inherently problematic about polling the same group over and over is that the initial selection might have been non-representative- it might track the trends accurately, but could be really wrong with respect to the true distribution of opinion.

    I think we will have to wait for the results to know what each of the different polls wanted. My suspicion is that the polls showing Clinton with a lead above the reported 95% confidence level are probably deliberately false, the ABC/WaPo poll being egregiously false.

    I have basically discounted all of the polls except for the LA Times poll, and I only use it for the trends. As such, Clinton started with a significant lead in that poll and Trump closed it during the Summer and took a reliable lead until just the last polling cycle (the poll only re-polls a portion of the selected cycle every few days), where Clinton has only a 1% lead. My observation is that the tape hurt Trump with the electorate, and probably cost hi around 3% support overall. What does that mean for November? I don't think we will even know until November the 9th.

    My reasons for being optimistic: turnout at the rallies indicates to me a vast difference in intensity that favors Trump; that it is politically incorrect to voice support for Trump makes it more likely his supporters are less likely to respond to strangers vocalizing that support; that all of the organizations showing the largest Clinton leads are clearly active supporters of Clinton, and I find it extremely unlikely that they haven't actively put their fingers on the polling scales to buttress that support; and, finally, the wrong track/right track polling of Gallup still indicates the change candidate should win this election.

    ReplyDelete
  8. The prospect of a Hillary Clinton administration sends the Kipling running through one's head.

    "The tares they had laughingly sown were ripe to the reaping.
    The trust they had leagued to disown was removed from their keeping.
    The eaters of other men’s bread, the exempted from hardship,
    The excusers of impotence fled, abdicating their wardship,
    For the hate they had taught through the State brought the State no defender,
    And it passed from the roll of the Nations in headlong surrender!"

    ReplyDelete
  9. C'mon, really, a landline? I'm 71 years old & haven't had a land line since the moon landing. I'm betting you've got an AOL email account you use occasionally. Right? ����

    ReplyDelete