Scribd Logo
Upload

Welcome to Scribd!

  • 4 views

    Hypotheses Testing Hand-Outs EN

    Uploaded by

    Awp DLore

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Hypotheses Testing Hand-Outs EN

    Uploaded by

    Awp DLore
    4 views2 pages

    Original Title

    hypotheses testing hand-outs EN

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    4 views2 pages

    Hypotheses Testing Hand-Outs EN

    Uploaded by

    Awp DLore

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 2

    Hypotheses testing

    Tatyana Zhuravleva
    May 11, 2024

    1 Type I and Type II errors


    Type I error: Rejecting H0 when it is true, which is a false positive. We denote the probability of this
    type of error by α

    Type II error: Failing to reject H0 when it is false, which is a false negative. We denote the probability
    of this type of error by β

    2 How do we perform tests?


    Assume θ̂ has the sampling distribution

    θ̂ ∼ N (θ, V )

    1. Define the hypotheses. In this case, suppose we test:


    H0 : θ = θ0 Versus H1 : θ > θ0
    2. State the test statistic and its distribution, then compute its value. In this case, due to the known standard
    error, standardisation of θ̂ yields:
    θ̂−θ0
    ∼ N (0, 1)
    E.S.E(θ̂)

    3. Define the critical region for a given significance level, α. Obtain critical values from statistical tables to de-
    fine the critical region (also known the rejection region)
    4. Decide whether or not to reject H0 . If the statistical value lies in the critical region than we reject H0 , otherwise
    we do nor reject H0 .
    5. Draw conclusions. It is always important to draw conclusions in the context of the original problem.

    3 Significance Levels
    Definition: Significance level of a test is the probability of committing a Type I error.

    UPPER-TAILED TEST TWO-TAILED TEST


    α = 0.10 =⇒ z0.10 = 1.2816 α = 0.10 =⇒ zα/2 = z0.05 = 1.6449
    giving critical value of ±1.6449
    α = 0.05 =⇒ z0.05 = 1.6449 α = 0.05 =⇒ zα/2 = z0.025 = 1.9600
    giving critical value of ±1.9600
    α = 0.01 =⇒ z0.01 = 2.3263 α = 0.01 =⇒ zα/2 = z0.005 = 2.5758
    giving critical value of ±2.5758

    4 Hypothesis test for a single mean (σ known)


    Test statistic is:

    X̂−µ0
    Z= √σ ∼ N (0, 1)
    n

    1
    5 Hypothesis test for a single mean (σ unkown)
    The test statistic is:
    X̂−µ0
    T = S
    √
    ∼ tn−1
    n

    6 Hypothesis Test for a Single Proportion


    p
    The standard error for a proportion is S.E.(P ) = π(1 − π)/n. When testing H0 : π = π0 , under H0 the standard error is
    p
    π0 (1 − π0 )/n leading to the following test statistic, achieved by standardising the sample proportion estimator P .

    The test statistic is:

    P −π0
    Z∼
    = √ ∼ N (0, 1)
    π0 (1−π0 )/n

    7 Hypothesis Test for the Difference between two population pro-


    portions
    In this case, the test statistic is:

    P1 −P2 −(π1 −π2 )


    Z∼
    = √
    P (1−P )(1/n1 +1/n2 )

    R1 +R2
    where P = n1 +n2
    , R1 and R2 represent the number of favourable responses.

    8 Hypothesis Test for the Difference between two population means


    Variances known
    ¯ ¯
    X −X −(µ1 −µ2 )
    Z = √1 2 2 2
    ∼ N (0, 1)
    σ1 /n1 +σ2 /n2

    Hence critical values are obtained from the standard normal distribution tables.
    Note: if testing for the equality of means, then µ1 − µ2 = 0 under H0 . Hence, in the above equation, we set the term (µ1 −
    µ2 ) = 0

    Variances unknown and unequal


    If the population variances σ1 2 and σ2 2 are unknown, provided the sample sizes n1 and n2 are large (greater than 30) then:

    X¯ −X¯ −(µ1 −µ2 )


    Z∼
    = √1 2 2 2
    ∼ N (0, 1)
    S1 /n1 +S2 /n2

    Hence critical values are obtained from the standard normal distribution tables.
    Note: if testing for the equality of means, then µ1 − µ2 = 0 under H0 . Hence, in the above equation, we set the term (µ1 −
    µ2 ) = 0

    Variances unknown and equal


    If the population variances σ1 2 and σ2 2 are unknown but assumed equal, then:
    ¯ ¯
    X −X −(µ1 −µ2 )
    T = √1 2 2 ∼ tn1 +n2 −2
    Sp (1/n1 +1/n2 )

    Where Sp 2 is the pooled variance estimator, given in

    (n1 −1)S1 2 +(n2 −1)S2 2


    Sp 2 = n1 +n2 −2

    Hence critical values are obtained from the Student’s t distribution with n1 + n2 − 2 degrees of freedom.
    Note: if testing for the equality of means, then µ1 − µ2 = 0 under H0 . Hence, in the above equation, we set the term (µ1 −
    µ2 ) = 0

    You might also like