Welcome to PlagueFest.com! Log in or Sign up to interact with the Plague Fest community.
  1. Welcome Guest! to interact with the community and gain access to all the site's features.

Proper experimental design and collection data for behavioral correlation

Discussion in Everything & Anything started by epi1to10000, Jan 25, 2013

  1. Feb 4, 2011
    In the Psychological and Neuro-cognitive sciences for data to be valid and be of significant relevance to a proposed theory:

    The experimental design must account for all relevant variables:
    -Environment, Information input, record relevant previous biasing history, controls, ect.

    Be repeatable across many cultures

    Be repeatable across ethological conditions (across multiple natural environments)

    Be observable and coherent across multiple experimental measurement models (Concordance)

    The value of the signal - noise must be significant to interpretation

    What happens in political discourse is that people have an entrenched belief, use observer bias (Fundamental attribution error) and find and/or fabricate data to support their belief. In this endeavor they do not us any of the above mentioned scientific imperatives and use pure biased reason where garbage in equals garbage out.
    Edited: Change all variables to all relevant
    epi1to10000, Jan 25, 2013 Last edited by epi1to10000, Jan 25, 2013
  2. May 15, 2011
    • Funny Funny x 1
    • Aug 18, 2006
      This. I feel mentally unsophisticated now.
    • May 5, 2012
      I am confused.
    • Nov 11, 2011
      Your first point is invalid. Yes, you can account for many variables but you will never cover "all" variables. Your last point makes very little sense. SNR "must be significant to interpretation"? All experiments need to be weighed according to their basis, what the ratio has to do with anything is beyond me. As far as the discourse, it's not only in political circumstances (if I take it to be literal), but religious people hold the same beliefs.

      Believer: "I believe in Jesus because I found his picture on a piece of toast."
      Scientist: "The heating coil is shaped like this and since this one overheated, the picture you see is not really Jesus."
      Believer: "I don't care what you say, this is Jesus."

      TL: DR; People believe what they want to believe over many cultural and ethological conditions (i.e, strong religious conviction even though a scientist can explain an anomaly).
      • Agree Agree x 1
      • Feb 4, 2011
        Agreed (good call), all variables cannot be accounted for, I should have said all relevant variables. (Relevant variables is a whole nother ball of wax). I'm not saying signal to noise variable; I am trying to find a way to state that the measured outcome should be significant to the relevant hypothesis and measurement methods.
        Example-> If you can show an increase in IQ of 1 under certain environmental histories and have a sample size of 10,000; that is interesting but not anything that has "Cash Value" in making a decision under most circumstances. "Cash Value" as in Jamesian pragmatic decision.
        William James would agree there is much similarity between religious and political beliefs.
        epi1to10000, Jan 25, 2013 Last edited by epi1to10000, Jan 25, 2013
      • Apr 9, 2012
        Is this another 'I wanna show how smart I am' thread?
      • Jun 11, 2012
        I understand and I want to answer and write a really long complex theoretical argument approach thingy....But my work cup is missing and I am pissed off. But for others these information are to be understood with open minded approach to a non-linear thinking.

        For example: Normality is only dictated from the point of the view of the collective i.e what a mass amount of people find normal becomes the widely accepted normal.

        Also in terms of variables, yes you can't cover all possible outcomes that might influence. This is the one of the main reason the kilogram disc is always being recalibrate due to changing environments. Also variables are why the Pressure changes and we can't affix a fixed value.
      • Oct 29, 2010
        It's more of a "I wanna show how fucking loony I am" thread.