Is it me, or is it anyone else?

Discussion in 'Off Topic Area' started by 47MartialMan, Mar 11, 2016.

  1. 47MartialMan

    47MartialMan Valued Member

    Nope......it will be like: I know someone who had got raped, do you know anyone?
     
  2. philosoraptor

    philosoraptor carnivore in a top hat Supporter

    Predict that murder is closer to national reporting rate than suicide. Your sampling method is flawed.
     
  3. holyheadjch

    holyheadjch Valued Member

    Do you seriously not see why asking a load of people from the same social group questions like this is going to produce data that is incredibly misleading?
     
  4. philosoraptor

    philosoraptor carnivore in a top hat Supporter

    47martialman doesn't care how his family thinks of him and they don't care how he thinks of them.
     
  5. philosoraptor

    philosoraptor carnivore in a top hat Supporter

    So for example, regarding statistical testing, when we look at something like this, it becomes a bit more clear.

    [​IMG]

    How do you show simple statements like "People who own a gun are more likely to defend themselves from homicide than people who do not own guns" is true?

    Already you have a problem with your question; a homicide is more likely to be reported than defense against a homicide. You can't rely on most formally collected data sets, so let's say you perform a better version of 47martialman's experiment and phone up randomized phone numbers in the United States.

    First let's test the data. We can make a prediction about the data - very few people will have witnessed many murders and very many will have witnessed few murders. We can test our data that way.

    Then we can compare our data to show that there is a significant difference between them, that the difference between the means cannot be explained by the variance within the samples - if, as in our picture, there is a very high degree of variance and a low difference between our samples, it is possible that the difference is due to methodological error as opposed to differences reflected in the populations (gun owners vs non gun owners) sampled.
     
  6. 47MartialMan

    47MartialMan Valued Member

    Nah......that side of the family are very forthrighteous
     
  7. philosoraptor

    philosoraptor carnivore in a top hat Supporter

    That's an assumption that compromises your data.

    Edit: worse it's an unfounded assumption.
     
    Last edited: Apr 2, 2016
  8. 47MartialMan

    47MartialMan Valued Member

    No assumption. Interviews went well.
     
  9. philosoraptor

    philosoraptor carnivore in a top hat Supporter

    I don't think you know what that word means. If before you begin your research you assume that your family members will be forthright, well, that's a huge assumption. Any data you collect is compromised and meaningless by that assumption because the question then becomes "What violence will my family tell me about?" rather than "What violence has my family encountered?"
     
  10. holyheadjch

    holyheadjch Valued Member

    Data is still trash. 10 points on offer if you can tell me why.
     
  11. 47MartialMan

    47MartialMan Valued Member

    Yep..that is what they stated
     

Share This Page