William M. Briggs

Statistician to the Stars!

Stream: AP’s LGBT ‘Rights’ Survey: Stacked, Biased, Bogus

pn

Today’s post is at The Stream: AP’s LGBT ‘Rights’ Survey: Stacked, Biased, Bogus.

AP is reporting young Americans “overwhelmingly support LGBT rights”, a proposition deduced from a GenForward survey “of Americans ages 18-30”. Results?…

Everybody knows that question wording affects the outcome of surveys. In politics, adjusting phrasing to solicit a desired answer is called push-polling. Push-polls make use of ancient prosecutorial trick questions, such as “Have you stopped beating your wife?” For example, asked of the same people, “Are you in favor of killing, chopping up, and selling for parts unwanted people?” will give vastly different results than “Are you in favor of a woman’s right to choose?”, even though both questions are about the same thing, abortion.

GenForward acknowledges this form of bias in their own survey. Half their panel were asked if they supported “Coverage of transgender health issues by health insurance”, of which 47% of blacks and 40% of whites strongly supported. But the other half were asked a much more specific question, whether they supported “Coverage of transgender health issues, such as hormone treatments and sexual reassignment surgery, by health insurance”. To this only 34% of blacks and 34% of whites strongly favored, a substantial reduction. The same drop was found for Asians (49-35%) and Hispanics (53-32%)….

The GenForward survey purports to be a representation across all Americans. Is it? A good chunk of its respondents were not employed, from 42% of blacks to 32% of whites. Yet the Bureau of Labor Statistics reports unemployment rates for people these ages at about 16% for 18-19-year-olds and 9% for 20-24-year-olds, with rates declining for older folks (until retirement, of course). Many more GenForward panelists were unemployed than average Americans.

Panelists also report anomalously large rates of deviant sexuality. Among blacks, 6% say they are “Gay/Lesbian”, 7% “Bisexual”, 1% “Transgender, transsexual, or gender non-conforming”, for a total of 14% or less across all LGBT categories (it’s “or less” because the survey allowed respondents to choose more than one identification)…

Did you go to The Stream and read the rest? Circle one: (Yes) (No).

21 Comments

  1. Politics has taught “science” well. Who needs data? For global warming, LGBT, who’s leading in the presidential polls? Just create the outcome you want and publish it. No one checks, no one who reads it is smart enough to realize you fabricated the outcome. It’s like ghost stories around a campfire. The kids just eat them up and jump at every noise for a week thereafter, at least. The fact that GenForward can openly state the bias in their studies and KNOW that it will not matter is very, very telling. (They state: “We present our data in an accessible form to multiple constituencies with the hope that our findings will contribute to a call to action to bring about change rooted in the ways young people of color experience contemporary America.” This clearly indicates they are activists presenting research that is used to push their political agenda. Not scientific research.)

    It may not be possible to overestimate the stupidity of the American people. (Or to avoid the consequences thereof.)

    It is unlikely that asking about freedom of religion or speech would have helped—Briggs stated these are the “safe-space denizens” and said persons do not believe in freedom of religion or speech and won’t until their speech is limited. They appear to have little or no religion, so that may not matter.

    In this vein:
    http://ktla.com/2016/08/09/mother-and-son-in-love-face-jail-time-for-incestuous-relationship/
    Other headlines on this story:
    Mother and son charged with incest, fight for relationship
    Mother and Son going to jail for Love Relationship in New Mexico
    Mother and Son Lovers in New Mexico Claim Incest Should Be Legal
    New Mexico mother and son fell in love and will go to JAIL to defend relationship
    Wonder how the GenForward group rates this activity…..

  2. No. I didn’t go to the Stream, an unbiased and trusted media outlet.

  3. JH: Don’t worry, it is fine to stick with CNN. No hurtful deviation from the party line there.

  4. Briggs is only partly right about the abuse of the survey tool & the subsequent “weighing of the evidence.”

    Its not that they’re doing it wrong — because from a certain perspective (theirs) they’re doing it right. They want a particular answer, so they “torture the data until it confesses” and they get the answer they want. Some of this is done with willful unscrupulousness, and some out of naive self-delusion (both are intrinsic elements of the ‘human condition’). Either way, they’re like the dowser that flicks the rod subconsciously when they need to indicate water (an example of the naive self-delusional), or, the butcher that unscrupulously sneaks their thumb on the scale to increase the weight & inflate the sale price.

    Philosophically, what a wonderful world it would be if those with political & philosophical agendas and/or rigid value systems could truly objectively “weigh the evidence” without “putting their thumbs on the scale” to force out a desired, instead of truthful, finding — and then accept and apply the actual truthful finding(s) and apply only those to actual policy decisions … why … it would be just, no exactly, like, N. deG Tyson’s imaginary world of Rationalia…but, as Briggs made clear earlier this week, that is also unacceptable.

  5. Trigger Warnings

    August 12, 2016 at 12:06 pm

    It’s astonishing, really, that there are so many people doing surveys and so few have failed to read the canonical work in survey question construction, “The Art of Asking Questions”, first published by Stanley Payne in 1951. Although the argot of 1951 is dated and somewhat different than that of today (it was a wonderful life, not an awesome life), the principles Payne lays out are as timely today as they were in 1951.

    As a statistical and psychometric consultant for many years, I was ever-disappointed by the sheer ignorance of the survey industry. And don’t get me started on “focus groups”. The current white whale, Frank Luntz, should be shuttled immediately down to the flensing deck.

  6. Anon, our TV seems stuck on CNBC, and I’d definitely adopt your advice if I care to learn how to work our complicated remote control. My favorite anchor woman is Melisa Lee.

  7. Surveying like this is a way of viewing political attitudes but not on specific issues. It has some use. You can use it to see the effects of propaganda, for instance. Ask people about all the various components of Obamacare without ever mentioning or conferring Obamacare, and compare the answers to questions directly mentioning and about Obamacare, and the results look very different. From this we Roger Ailes was good at his last job. Combine the results and you may be able to find a political path in there somewhere.

    JMJ

  8. Jersey, at last we agree on something! It’s another example of politics being used to sell an ideology… i.e. the canny production, sifting and censorship of “data” to provoke a desired irrational response.

  9. Well, it’s not what the survey does, but what it can tell you. This one can tell you a little about the boundaries of some political positions. But the political, like the sexual, is a very complex subject.

    JMJ

  10. Ye Olde Statistician

    August 12, 2016 at 8:21 pm

    Watch this to learn about polls:
    https://www.youtube.com/watch?v=G0ZZJXw4MTA

  11. Exactly. Take, for instance, the ridiculously misleading and misrepresenting question about abortion, “Are you in favor of killing, chopping up, and selling for parts unwanted people?” Though, even that question can tell you something rather frightening about the respondent, if their answer is anything other than “What the hell are you talking about?”

    JMJ

  12. JMJ: The answers to many survey questions can reveal frightening things about people that have nothing to do with the question.

  13. a rather odd, and biased, definition of abortion.

  14. Take, for instance, the ridiculously misleading and misrepresenting question about abortion, “Are you in favor of killing, chopping up, and selling for parts unwanted people?”

    which parts do you find mislead you?

  15. I did not say I was mislead. If I were, how would I even know to point it out? Did I find out later? Well, if you’re curious, no. It seems you, however, are deeply mislead.

    JMJ

  16. JMJ,

    You claimed to find the description misleading. Naturally, sharp fellow that you are, you were not misled yourself, for you spotted Briggs’ clever trick before being led astray by it. Not being as sharp as you, I don’t see the trick.

    I get that he is talking about abortion, which is something inflicted on unwanted people. They are killed by being cut up, and the resulting parts sold, when there are buyers, and dumped in the trash the rest of the time.
    As far as I can tell these are all accurate statements. I don’t see where I was misled and you were not. Can you help me out?

  17. Trump wins: How would you define abortion?

    JMJ: You could know you were mislead if afterward you discovered the truth. I would agree that your statement that if you were mislead you couldn’t point it out because that would imply you had discovered the truth. Interesting, though, that you then just repeat Andrew’s accusation back to him. For a bit here, you were making valid points. Guess it’s just too hard to sustain that kind of thing for you. Sad. Sometimes, you actually occasionally show signs of intellligent life. Your hatred of conservatives keep that from happening too often, I know. (A life sustained by hatred…….One would think you were one of those conservatives you hate.)

    Andrew: Some will object to the idea that a person is killed in abortion, as they believe in the magical transformation of a lump of tissue to a human being upon passing through the birth canal. Some believe that lump of tissue becomes magically self-aware at a different point, but that exact point varies depending upon when the person believes women will be pacified by the timing.

  18. In Fairfax Virginia the school board came out with a policy on transgender students. They started the paper with some definitions. I discovered that sex is what you are assigned on your birth certificate and there was no definition of gender. Stupid me, I thought that sex was determined at conception by chromosomes and gender was a property of declined languages. I never knew that sex was assigned at birth and humans have gender. Gender was proposed as a human property by a quack psychiatrist Dr. Money in the 1960 and was debunked, I thought. Now quackery is back as government policy.

  19. Ye Olde Statistician

    August 13, 2016 at 1:24 pm

    They started the paper with some definitions. I discovered that sex is what you are assigned on your birth certificate and there was no definition of gender. Stupid me, I thought that sex was determined at conception by chromosomes and gender was a property of declined languages.

    Doncha just love dem science deniers?

  20. Milton Hathaway

    August 13, 2016 at 2:44 pm

    Sometimes I wonder what ever happened to the “smell test”. In this particular survey, there are two parts to the smell test:

    1) Play a “what-if” game and invert the results. Would the media have still reported the story with such apparent glee? Or reported the story at all? What stories do we never see?

    2) Play a slightly different what-if game, and suppose that the survey is in error, and that reality is the inverse of the reported results. Will the authors of the survey suffer any negative consequences from their error? I.e., is there any accountability?

    As a side note, if you apply the smell test to Hillary/Trump polls, you come to the realization that such polls should be ignored until we get close to the election, where accountability starts to kick in. Until then, they are just push-polls, accountable only to the entity that commissioned them.

    In these days of information overload, the smell test is more indispensable than ever.

  21. Hmm, I’d say that gender is a property of language grammars. English has three genders. Genders don’t apply to specimens of sexually reproducing species. Sexes apply to those and there are only two. Male and Female created He them. Sex is not a dirty word.

    @YOS, yes, we do love them, whether science deniers or not ;).

Leave a Reply

Your email address will not be published.

*

© 2016 William M. Briggs

Theme by Anders NorenUp ↑