The 12th Invest in ME Research Conference June, 2017, Part 2
MEMum presents the second article in a series of three about the recent 12th Invest In ME International Conference (IIMEC12) in London.
Discuss the article on the Forums.

Way More Americans May Be Atheists Than We Thought (on questionaires)

Discussion in 'Other Health News and Research' started by RogerBlack, May 18, 2017.

  1. RogerBlack

    RogerBlack Senior Member

    Messages:
    891
    Likes:
    2,889
    https://fivethirtyeight.com/features/way-more-americans-may-be-atheists-than-we-thought/

    An interesting approach to people wanting to please the experimenter (or feeling compelled to)

    Take ten random statements of no great controversy 'I own a dog'.
    For half of the people, swap out one of the uncontroversial questions randomly with 'I believe in god'.

    Now, do not ask them to state the result directly, but just how many of the statements apply to them.

    The fact that this causes a very different answer than if they are asked explicitly is interesting for any aspect of research which may be biased by wanting to please the experimenter.
     
    Woolie, TiredSam, barbc56 and 6 others like this.
  2. Jonathan Edwards

    Jonathan Edwards "Gibberish"

    Messages:
    5,230
    Likes:
    31,933
    Neat
     
    barbc56, trishrhymes and Barry53 like this.
  3. Barry53

    Barry53 Senior Member

    Messages:
    2,277
    Likes:
    13,705
    UK
    Interesting. I suppose the "I believe in god" question has a built-in loading factor, a person's response influenced by things like cultural/family expectations, personal experiences, etc. Even if the questionnaire is guaranteed anonymous, there may still be a significant self-judgemental element that influences answering the question; not having to answer the question directly, but only obliquely and by omission, possibly alleviates some of the "pressure to conform".

    How you could ever achieve something similar in clinical trial I cannot imagine. The trial itself, no matter how well run, must engender some need to conform, even if only self imposed. Let alone when badly run and unethically run.
     
    Woolie likes this.
  4. RogerBlack

    RogerBlack Senior Member

    Messages:
    891
    Likes:
    2,889
    It would eliminate the fear of nebulous consequences or them not liking you, or ... in relation to your answers.
    It would not eliminate changes in answers due to convincing someone that black is white.
    In principle.

    In practice, I can't see how you'd do this without a significantly larger required effect size or trial group - this technique while possibly useful statistically dilutes your answers.

    It would be interesting to do the above trial with four arms.
    Ask the questions 'neutrally' - and ask them with a person of religion presenting the form and in the room.
    Perhaps after some church activity of some secular sort.
     
    Barry53 likes this.

See more popular forum discussions.

Share This Page