Scribd Logo
Upload

Welcome to Scribd!

  • Exercise 1

    Uploaded by

    Rong NG
    55 views1 page
    The document discusses the sampling distribution of ordinary least squares (OLS) estimators. It states that OLS estimators are random variables with a probability distribution called their sampling distribution. The sampling distribution describes the values the estimators could take on across different samples. The document also notes that under the Gauss-Markov assumptions, the OLS estimator is the best, linear, and unbiased (BLUE) estimator, meaning it has lower variance than other estimators and is an unbiased estimator of the parameter. The full sampling distribution of the OLS estimator is normally distributed with a mean of β and variance of σ2(X'X)-1.

    Original Description:

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses the sampling distribution of ordinary least squares (OLS) estimators. It states that OLS estimators are random variables with a probability distribution called their sampling distribution. The sampling distribution describes the values the estimators could take on across different samples. The document also notes that under the Gauss-Markov assumptions, the OLS estimator is the best, linear, and unbiased (BLUE) estimator, meaning it has lower variance than other estimators and is an unbiased estimator of the parameter. The full sampling distribution of the OLS estimator is normally distributed with a mean of β and variance of σ2(X'X)-1.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Flag for inappropriate content
    Download as docx, pdf, or txt
    55 views1 page

    Exercise 1

    Uploaded by

    Rong NG
    The document discusses the sampling distribution of ordinary least squares (OLS) estimators. It states that OLS estimators are random variables with a probability distribution called their sampling distribution. The sampling distribution describes the values the estimators could take on across different samples. The document also notes that under the Gauss-Markov assumptions, the OLS estimator is the best, linear, and unbiased (BLUE) estimator, meaning it has lower variance than other estimators and is an unbiased estimator of the parameter. The full sampling distribution of the OLS estimator is normally distributed with a mean of β and variance of σ2(X'X)-1.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Flag for inappropriate content
    Download as docx, pdf, or txt
    of 1

    Exercise 1 NG KUANG YONG 825934

    Sampling distribution of the OLS estimators


     The sampling distribution of the OLS estimators describes the values which could take on
    over different samples. This is because b0 and b1 are computed from a sample, hence the
    estimators themselves are random variables with a probability distribution.

     In the statistic, sampling distribution of the sample mean shows that the mean of the
    sample means is exactly equal to the population mean. Besides that, the dispersion of the
    sampling distribution of the sample mean than the population distribution and tends to
    become bell-shaped and to approximate a normal probability distribution.

     The sampling distribution of the OLS estimator as following:


    y = xβ + e
    which the e is normally distributed with zero mean and variance.
    b = (X’X)-1X’y = f(y)
    which the b is an estimator of β and function of the random sample data. b has a
    probability distribution which called its sampling distribution.

     Based on the Gauss-Markov, the OLS estimator used is the best, linear and unbiased
    which in the short form of the BLUE.

     The best means that the variance of the OLS estimator is always smaller than variances
    for other estimator which can be shown by:
    var (βi)OLS < var (βi)other
    For example, the variance of OLS will get the v(b) = σ 2(X’X)-1 but other estimator will
    have the variance that v(b) = σ 2(X’X)-1 + σ 2CC’.

     Secondly is the linear which the estimator of OLS is linear function to the dependent
    variable (Yi), which is the
    b = (X’X)-1X’y = Ay, which the A is a non-random matrix. For example:

    ()[ ]( )
    b1 a11 ⋯ a1 n y 1
    ⋮ = ⋮ ⋯ ⋮ ⋮
    bk ak 1 ⋯ a kn y 2

    b1 = [a11y2 + a12y2 +…+a1nyn]

     And the lastly is the unbiased which mean that the value estimated for OLS estimator or
    the E( βi ) same with the value of parameter ( βi ).
    Hence, y is an unbiased estimator of parameter, μ. If assumed that X is non-random,
    hence the E(b) will get the β and this show that the OLS estimator is unbiased.
    Finally, because the error term, e is assumed to be normally distributed which y = xβ + e,
    implied that y is normal distributed while b = (X’X)-1X’y = A(y) implies that b is also
    normally distributed.

     Hence the full sampling distribution of the OLS estimator, b is the b ~ N[β, σ 2(X’X)-1].

    You might also like