+ Reply to Thread
Results 1 to 4 of 4

Thread: Hypothesis testing : Help to formulate null hypothesis

  1. #1
    Points: 194, Level: 3
    Level completed: 88%, Points required for next Level: 6

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Hypothesis testing : Help to formulate null hypothesis




    Hi all

    I'm currently doing a hypothesis testing to test a set of sample data (large sample size). From my knowledge of statistics, the null hypothesis is usually the opposite of what the researcher is trying to prove.

    However for my case, it seems that I would need to formulate a null hypothesis and alternative hypothesis, whereby:

    Ho : x̄ 1 - x̄ 2 ≠ 0
    H1 : x̄ 1 - x̄ 2 = 0

    I understand that usually Ho uses the operator '='. Could someone advise me if the above formulation of the null and hypothesis is at all reasonable?

    Thanks in advance

    Regards
    Ben

  2. #2
    TS Contributor
    Points: 7,081, Level: 55
    Level completed: 66%, Points required for next Level: 69

    Location
    Copenhagen , Denmark
    Posts
    515
    Thanks
    71
    Thanked 123 Times in 116 Posts

    Re: Hypothesis testing : Help to formulate null hypothesis

    Assuming it was reasonable what would your next step be? How would carry out a test under the assumption that x_1 \not = x_2



    Simplest solution in youre case is probably just to assume equality, test for equality and if not rejected uphold the hypothesis of equality as working hypothesis...
    It would probably also be wise to take into consideration the precision of youre estimator and compare this to what you consider to be practically equal ( are 1.1=1.11 or 1.1=1.101 etc. )

  3. #3
    Points: 194, Level: 3
    Level completed: 88%, Points required for next Level: 6

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Hypothesis testing : Help to formulate null hypothesis

    Quote Originally Posted by JesperHP View Post
    Assuming it was reasonable what would your next step be? How would carry out a test under the assumption that x_1 \not = x_2



    Simplest solution in youre case is probably just to assume equality, test for equality and if not rejected uphold the hypothesis of equality as working hypothesis...
    It would probably also be wise to take into consideration the precision of youre estimator and compare this to what you consider to be practically equal ( are 1.1=1.11 or 1.1=1.101 etc. )
    yeah that was what I thought would be the issue..

    May I know what does it mean to test for equality? Do you mean to use > and < in the hypothesis?

  4. #4
    TS Contributor
    Points: 7,081, Level: 55
    Level completed: 66%, Points required for next Level: 69

    Location
    Copenhagen , Denmark
    Posts
    515
    Thanks
    71
    Thanked 123 Times in 116 Posts

    Re: Hypothesis testing : Help to formulate null hypothesis


    Quote Originally Posted by Corsa View Post
    May I know what does it mean to test for equality?
    I meant H_0 : \mu = \mu_0 \phantom{xx} H_a: \ \mu \not = \mu_0 which I guess is what you mean by
    use > and < in the hypothesis

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats