Results 1 to 3 of 3

Thread: t-test Dependent samples vs Correlation

  1. #1
    Newbie
    Joined
    Jul 2017
    From
    Johannesburg
    Posts
    4

    t-test Dependent samples vs Correlation

    I know that t tests are primarily used when n<=30 and/or sigma is unknown.
    Which statistics should be used if n>30 or sigma is known.

    E.g. which statistics I should use to assess the effect of a medication on 45 patients
    when week-1: received the medication, week-2 received a Placibo.
    Applying a measurement scale, we know mean of general pop. is 30 and Std-dev=10.

    Would a correlation statistics between the two sets of results be the correct way to go?
    appreciate your earliest response
    Shemtov
    Follow Math Help Forum on Facebook and Google+

  2. #2
    Senior Member
    Joined
    Mar 2008
    From
    Pennsylvania, USA
    Posts
    334
    Thanks
    45

    Re: t-test Dependent samples vs Correlation

    When n>30, the critical z and the critical t are essentially the same. So make life easier on yourself and use Z when n>30.

    The Pearson correlation coefficient does NOT represent the line of best fit; a correlation coefficient of +1 does not mean that a unit increase in one variable represents a unit increase in another variable. A correlation of +1 just means that there is no variation between the data points and the line of best fit. You should always have a reason for testing correlation. Since you are testing for effects, do not test for correlation.
    Follow Math Help Forum on Facebook and Google+

  3. #3
    Senior Member
    Joined
    Mar 2008
    From
    Pennsylvania, USA
    Posts
    334
    Thanks
    45

    Re: t-test Dependent samples vs Correlation

    If sigma is unknown but n>30, you can approximate the population standard deviation with the sample standard deviation.
    Last edited by abender; Sep 17th 2017 at 07:27 PM.
    Follow Math Help Forum on Facebook and Google+

Similar Math Help Forum Discussions

  1. Dependent Samples Comparing Means
    Posted in the Statistics Forum
    Replies: 0
    Last Post: Oct 8th 2015, 04:36 AM
  2. Two dependent samples problem
    Posted in the Statistics Forum
    Replies: 2
    Last Post: Jul 22nd 2012, 06:14 AM
  3. Correlation of samples
    Posted in the Advanced Statistics Forum
    Replies: 1
    Last Post: Mar 5th 2010, 10:13 PM
  4. How to construct statistics from dependent samples
    Posted in the Advanced Statistics Forum
    Replies: 0
    Last Post: Jul 13th 2009, 02:43 AM

/mathhelpforum @mathhelpforum