+ Reply to Thread
Results 1 to 4 of 4

Thread: "Easy" regression/hypothesis test questions

  1. #1
    Points: 1,426, Level: 21
    Level completed: 26%, Points required for next Level: 74

    Posts
    6
    Thanks
    0
    Thanked 0 Times in 0 Posts

    "Easy" regression/hypothesis test questions




    Dear community, I'm having trouble with some basic regression questions.

    1)

    Test the null H0: Intercept = 0.

    If -95% confidence interval = -0.004, +95% confidence invterval = 0.0041, do I accept the null hypothesis at ALPHA = 0.05 that Interval = 0?

    2)

    Is the error term = the difference between YActual and YResidual (which is provided by the excel regression output).

    ---

    Thankyou for any help.

  2. #2
    TS Contributor
    Points: 18,889, Level: 87
    Level completed: 8%, Points required for next Level: 461
    CowboyBear's Avatar
    Location
    New Zealand
    Posts
    2,062
    Thanks
    121
    Thanked 427 Times in 328 Posts

    Re: "Easy" regression/hypothesis test questions

    Hi, welcome to the forum. Please have a look over posting guidelines: http://www.talkstats.com/announcement.php?f=2

    What work have you done so far to try and figure these questions out?

  3. #3
    Points: 1,426, Level: 21
    Level completed: 26%, Points required for next Level: 74

    Posts
    6
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: "Easy" regression/hypothesis test questions

    What I have done:

    Calculated the 95% confidence interval for each Intercept.

    The outputted intercepts are all extraordinarily close to 0 (such as 1.8E-18).

    "0" lies within the intercept's confidence intervals. Since this is the case, does that mean that

    H0: Intercapt = 0 is accepted?

  4. #4
    TS Contributor
    Points: 18,889, Level: 87
    Level completed: 8%, Points required for next Level: 461
    CowboyBear's Avatar
    Location
    New Zealand
    Posts
    2,062
    Thanks
    121
    Thanked 427 Times in 328 Posts

    Re: "Easy" regression/hypothesis test questions


    Quote Originally Posted by imiyakawa View Post
    What I have done:

    Calculated the 95% confidence interval for each Intercept.

    The outputted intercepts are all extraordinarily close to 0 (such as 1.8E-18).

    "0" lies within the intercept's confidence intervals. Since this is the case, does that mean that

    H0: Intercapt = 0 is accepted?
    You're correct in thinking that this indicates a (just) non-significant result; you should not reject the null hypothesis. (Unless CI's and p values can provide differing conclusions in the one sample regression case - any takers?)

    But it's not ever really correct to say that you "accept" a null hypothesis. A non-significant result only means that one has not found evidence to reject the null hypothesis. It doesn't mean the null is true.

+ Reply to Thread

           




Similar Threads

  1. Replies: 2
    Last Post: 09-14-2010, 04:54 PM
  2. Replies: 0
    Last Post: 08-02-2010, 07:57 PM
  3. "significance metrics" for hypothesis testing
    By p05 in forum Statistical Research
    Replies: 0
    Last Post: 08-02-2010, 04:58 PM
  4. Replies: 4
    Last Post: 01-25-2010, 11:57 AM
  5. two questions for "poisson process"
    By smallkai in forum Probability
    Replies: 1
    Last Post: 08-11-2008, 05:11 AM

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