Scribd Logo
Upload

Welcome to Scribd!

  • 67 views

    SRT 605 - Topic (7) ANOVA

    Uploaded by

    pug
    Kruskal-Wallis H test Reason: 1) There are 3 independent groups (badminton player groups) 2) The dependent variable is weight (numerical) 3) One of the groups is not normally distributed 4) Kruskal-Wallis H test is a non-parametric method used to compare 3 or more independent groups when the dependent variable is numerical but not normally distributed. So the appropriate analysis method for this data is Kruskal-Wallis H test since one of the assumption of ANOVA (normal distribution) is violated.

    Copyright:

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

    SRT 605 - Topic (7) ANOVA

    Uploaded by

    pug
    67 views23 pages
    Kruskal-Wallis H test Reason: 1) There are 3 independent groups (badminton player groups) 2) The dependent variable is weight (numerical) 3) One of the groups is not normally distributed 4) Kruskal-Wallis H test is a non-parametric method used to compare 3 or more independent groups when the dependent variable is numerical but not normally distributed. So the appropriate analysis method for this data is Kruskal-Wallis H test since one of the assumption of ANOVA (normal distribution) is violated.

    Original Description:

    ANOVA

    Copyright

    © © All Rights Reserved

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Kruskal-Wallis H test Reason: 1) There are 3 independent groups (badminton player groups) 2) The dependent variable is weight (numerical) 3) One of the groups is not normally distributed 4) Kruskal-Wallis H test is a non-parametric method used to compare 3 or more independent groups when the dependent variable is numerical but not normally distributed. So the appropriate analysis method for this data is Kruskal-Wallis H test since one of the assumption of ANOVA (normal distribution) is violated.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    67 views23 pages

    SRT 605 - Topic (7) ANOVA

    Uploaded by

    pug
    Kruskal-Wallis H test Reason: 1) There are 3 independent groups (badminton player groups) 2) The dependent variable is weight (numerical) 3) One of the groups is not normally distributed 4) Kruskal-Wallis H test is a non-parametric method used to compare 3 or more independent groups when the dependent variable is numerical but not normally distributed. So the appropriate analysis method for this data is Kruskal-Wallis H test since one of the assumption of ANOVA (normal distribution) is violated.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    of 23

    SRT 605: ANOVA

    Dr Tan Chee Hian


    Profesor Madya
    Sports Management Department
    FSR, UiTM Shah Alam
    03-55442949
    A4046
    [email protected]
    ANALYSIS OF VARIANCE (ANOVA)
    Introduction
    Apply for more than 2 groups/category independent variable.
    - If only one classifying variable, then we have one-way ANOVA
    (parametric test) or Kruskal Wallis (nonparametric test). e.g.
    treatment groups.
    - If two classifying variables are present, then we have two-ways
    ANOVA. e.g. treatment groups and route of taking medication.
    ANOVA

    Dependent Variables *(numerical)


    Liked: age, BP, glucose level, weight, height etc
    Independent Variables * (Categorical)
    Liked: Group A, Group B and Group C
    Anova

    a) Independent variable (IV):


    - more than two groups
    b) Dependent variable (DV):
    – numerical
    c) Observations/sample: Independent
    d) Distribution of data – Normal (Parametric)
    One- way ANOVA
    One-way ANOVA is conducted to access whether population means
    are significantly different among groups.

    - If the overall ANOVA test is statistically significant, then, pairwise


    comparison test should be conducted to investigate which of the two
    populations means are significantly different.
    Assumptions for ANOVA
    1) Random sample.
    2) The observations are independent.
    3) The distribution is normal.
    4) The variance for the observations are equal.

    If assumption not met?


    used Non-parametric Method
    Kruskal Wallis Test
    Calculation Manually
    Suppose the result of the treatment for three different races, namely
    Malay 1, Chinese 2 and Indian 3, and we are interested in testing
    whether these population means are different. These sample are
    randomly selected.
    Data
    Malay1 Chinese2 Indian3
    77 75 89
    68 68 75
    80 82 68
    75 85 82
    68 79 89
    82 75 90
    79 68 75
    83 82 68
    65 85 96
    86 91 93
    Steps
    Step 1: Generate the hypothesis
    Ho: μ1 = μ2 = μ3
    Ha: At least two of the treatment groups are different.

    Step 2: Set the significance level


    α = 0.05
    Con’t
    Step 3: Checking the assumptions
    Assumptions for ANOVA
    1. Random sample: from statement
    2. The observations are independent: Each observation refer to
    different group.
    3. The distribution is normal.
    SPSS - Computer
    Select Analyze=>Descriptive Statistics=> Explore
    Insert treatment in the Dependent List box.
    Insert group in the Factor List box.
    In the Explore: Plot, click on the Normality
    Plots with Test and Histogram.
    SPSS
    Con’t
    4. The variance for the observations are equal.

    A= ∑X² B = (∑X )²
    N
    C= (∑X 1 )² + (∑ X 2 )² + …… + (∑X k )²
    n1 n2 nk
    Table used
    One Way ANOVA
    Test statistics using SPSS:
    Analyze => compare mean = > One-way ANOVA
    Insert Treatment in Dependent List box
    Insert Group in Factor box
    Select Post Hoc and click Bonferroni and continue
    Select Option and click Descriptive and Homogeneity of variance test and
    continue
    Select Ok
    SPSS
    Report of result APA
    Table 1: Descriptive Results

    Treatment n Mean SD 95% CI


    1 10 76.30 7.15 81.41
    2 10 79.00 7.51 84.37
    3 10 92.50 10.36 89.91
    Con’t
    Table 2: One way ANOVA for the treatment

    Sum of Square df mean square F


    Between groups 193.267 2 96.633 1.349
    Within groups 1934.600 27 71.652
    Total 2127.867 29

    Conclusion : An analysis of variance showed that the effect of treatment was non
    significant, F(2,27) = 1.35, p = 0.277
    Kruskal Wallis Tests
    Use the same dataset and run for non parametric test.
    - Analyze => Non parametric Tests => Legacy dialogue => k Independent
    sample
    - Insert Treatment in Test variable list box
    - Insert Group in Grouping variable box
    - Click Define range, insert Min and Max and continue
    - Click Option and Descriptive and continue
    - Click Ok
    Con’t – Non para
    Example
    Shary wants to study a mean difference for sport performance score
    between 3 groups of player. A number of 150 FSR students was
    randomly sample from football player, netball player and badminton
    player.

    What is the appropriate method used in the data given.


    Solution
    1. dependent variable– sport performance score
    (numerical)
    2. independent variable – football player, netball player and badminton player (3
    groups)
    3. Groups observation – independent
    4. Distribution – n=150, sample size is large enough to CLT, so consider normal
    distribution

    ANALYSIS METHOD: One-way ANOVA


    Example
    Ahmad would like to study is there any difference in weight between
    3 groups of badminton player. A sample of 60 player was randomly
    selected. One of the group was normal distribution .

    What is analysis method that appropriate for this data?

    You might also like