Statistics help please

Discussion in 'Technology' started by andy_rocks, Jan 30, 2006.

Users Viewing Thread (Users: 0, Guests: 0)

  1. andy_rocks

    andy_rocks Registered User

    Joined:
    Dec 2, 2003
    Messages:
    8,705
    Likes Received:
    0
    Statistics help please

    Right, I'm doing an audit at the moment, comparing a procedure we do to others that have been published.

    From 19 cases, we have a success rate of 73.3 %.

    Statistically, is this significantly different to 90% given the small numbers, and, what are the confidence intervals? Is Student T test the right one to use?

    Cheers guys :D
  2. 1615634792921.png
  3. ManofScience

    ManofScience Guest

  4. andy_rocks

    andy_rocks Registered User

    Joined:
    Dec 2, 2003
    Messages:
    8,705
    Likes Received:
    0
    I'm no statistician (as you may have guessed) but is that not just for doing the test though?

    My real problem is knowing which one to do....T test is for groups of data between population, and I don't think that's appropriate in this case :confused:
  5. andy_rocks

    andy_rocks Registered User

    Joined:
    Dec 2, 2003
    Messages:
    8,705
    Likes Received:
    0
    Hang on, could be chi squared :D
  6. nouvelle fille

    nouvelle fille Registered User

    Joined:
    Aug 10, 2005
    Messages:
    437
    Likes Received:
    0
    Location:
    Tyneside Riviera
    You've prob sorted it now, but from what i remember it is chi square ......
  7. andy_rocks

    andy_rocks Registered User

    Joined:
    Dec 2, 2003
    Messages:
    8,705
    Likes Received:
    0
    It wasn't chi squared :(

    This should be such a simple thing to do, and I'm totally stumped.

    Wish I'd done A level maths now.
  8. claire

    claire Beautifully Bad

    Joined:
    Dec 8, 2005
    Messages:
    1,568
    Likes Received:
    0
    Location:
    Leamington Spa
    Unpaired t-test?

    Need more info to be certain

Share This Page