Scribd Logo
Upload

Welcome to Scribd!

  • 19 views

    Lesson 16 Post-Hoc Tests: N MS Q HSD

    Uploaded by

    pbchanta
    The document discusses post-hoc tests, specifically Tukey's HSD post-hoc test. It explains that a post-hoc test is used after ANOVA to determine which group means differ significantly. Tukey's test compares the differences between all group mean pairs to a critical value called the HSD. The HSD is computed using the mean square within value from the ANOVA. Effect size is also discussed through eta-square, which measures the proportion of between-group variability to total variability, and omega-square, a more accurate measure of effect size.

    Copyright:

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

    Lesson 16 Post-Hoc Tests: N MS Q HSD

    Uploaded by

    pbchanta
    19 views3 pages
    The document discusses post-hoc tests, specifically Tukey's HSD post-hoc test. It explains that a post-hoc test is used after ANOVA to determine which group means differ significantly. Tukey's test compares the differences between all group mean pairs to a critical value called the HSD. The HSD is computed using the mean square within value from the ANOVA. Effect size is also discussed through eta-square, which measures the proportion of between-group variability to total variability, and omega-square, a more accurate measure of effect size.

    Original Description:

    estadistica

    Original Title

    Tukey

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses post-hoc tests, specifically Tukey's HSD post-hoc test. It explains that a post-hoc test is used after ANOVA to determine which group means differ significantly. Tukey's test compares the differences between all group mean pairs to a critical value called the HSD. The HSD is computed using the mean square within value from the ANOVA. Effect size is also discussed through eta-square, which measures the proportion of between-group variability to total variability, and omega-square, a more accurate measure of effect size.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    19 views3 pages

    Lesson 16 Post-Hoc Tests: N MS Q HSD

    Uploaded by

    pbchanta
    The document discusses post-hoc tests, specifically Tukey's HSD post-hoc test. It explains that a post-hoc test is used after ANOVA to determine which group means differ significantly. Tukey's test compares the differences between all group mean pairs to a critical value called the HSD. The HSD is computed using the mean square within value from the ANOVA. Effect size is also discussed through eta-square, which measures the proportion of between-group variability to total variability, and omega-square, a more accurate measure of effect size.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 3

    Lesson 16

    Post-hoc Tests

    Outline
    Tukeys HSD Post-hoc test
    -differences between means
    -studentized range statistic (q)
    -honestly significant difference (HSD)
    Example Tukey Problem
    Magnitude of the Effect
    -eta-square
    -omega-square

    Tukeys HSD Post-hoc test


    A post-hoc test is needed after we complete an ANOVA in order to determine which
    groups differ from each other. Do not conduct a post-hoc test unless you found an effect
    (rejected the null) in the ANOVA problem. If you fail to reject the null, then there are no
    differences to find.

    For the Tukeys post-hoc test we will first find the differences between the means of all
    of our groups. We will compare this difference score to a critical value to see if the
    difference is significant. The critical value in this case is the HSD (honestly significant
    difference) and it must be computed. It is the point when a mean difference becomes
    honestly significantly different.

    MSWithin
    HSD = q Note that q is a table value, and n is the number of values we are
    n
    dealing with in each group (not total n). The Mean Square value is from the ANOVA
    you already computed. To find q or the studentized range statistic, refer to your table
    on page A-32 of your text. On the table k or the number of groups is found along the
    top, and degrees of freedom within is down the side. Cross index the row and column to
    find the value you need to put in the formula above.

    Example
    This example is a continuation of the ANOVA problem we did in the last lesson. Here I
    show the groups, but have computed the average or mean of each group.

    Therapy A Therapy B Therapy C


    5 3 1
    2 3 0
    5 0 1
    4 2 2
    2 2 1
    X 1 = 3.6 X2 = 2 X3 =1

    The first step is to compute all possible differences between means:


    X 1 X 2 = 3.6 2 = 1.6 X 1 X 3 = 3.6 1 = 2.6 X 2 X 3 = 3.6 1 = 2.6

    We will only be concerned with the absolute difference, so, you can ignore any negative
    signs. Next we compute HSD.
    MSWithin 1.43
    HSD = q = 3.77 =
    n 5 3.77 .286 = 3.77(.53) = 1.99

    Now we will compare the difference scores we computed with the HSD value. If the
    difference is larger than the HSD, then we say the difference is significant.

    Groups 1 and 2 do not differ


    Groups 1 and 3 differ
    Groups 2 and 3 differ

    Magnitude of the Effect


    It is a common misconception that the size of the F-ratio you compute directly indicates
    how strongly the relationship is between the independent and dependent variable.
    However, a separate computation is needed to get a true idea of the strength of the
    relationship. Our test indicates only that there is difference in the treatments, and it is not
    a measure of a relationship between variables.

    Magnitude of the effect is the amount of variability that our independent variable can
    account for in the dependent variable. The easiest measure of the effect is eta-square
    (2). It measures the proportion of our between factor variability to the total variability.
    If we know what proportion of variability of the total is due to our treatment, we will
    have a rough idea of how strong the relationship might be.

    SS Between
    2 = Notice it is just a ratio of treatment effect variability to total variability
    SS total

    17.2
    For our example: 2 = = .50 50% of the variability in rated fear of spiders is
    34.4
    due to the type of therapy they received.

    One problem with eta-square is that it is a biased estimate and tends to overestimate the
    effect. A more accurate measure of the effect is omega-square (2).

    SS Between (k 1)MSWithin
    2 =
    SS total + MSWithin

    17.2 (3 1)1.43 17.2 (2)1.43 17.2 2.86 14.34


    2 = = = = = .40
    34.4 + 1.43 35.83 35.83 35.83

    You might also like