Scribd Logo
Upload

Welcome to Scribd!

  • 31 views

    Statistical Test Assumptions

    Uploaded by

    Jay Punayan

    Copyright:

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

    Statistical Test Assumptions

    Uploaded by

    Jay Punayan
    31 views28 pages

    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
    31 views28 pages

    Statistical Test Assumptions

    Uploaded by

    Jay Punayan

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 28

    Statistical Test

    Assumptions
    Angela Fel P. Gamalo
    Department of Mathematics
    Caraga State University

    Nov 29 - Dec 1, 2023 Basic Statistical Training and JASP Application Workshop
    Illustration

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    In Statistics…

    You should never assume that the statistical tests you are
    using are appropriate until you check your ASSUMPTIONS.

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Check your Assumptions

    DATA:
    You call me and I don’t answer.

    CONCLUSION:
    I am mad at you.

    ASSUMPTION:
    I had my phone with me and I don’t answer.

    What can we learn from this assumption?

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Check your Assumptions

    If your assumptions are wrong…


    - It prevents you from looking at the world accurately…
    - Your conclusions are going to be incorrect as well

    “… all models are limited by the validity of the assumptions


    on which they ride” Collier et al.

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Assumptions of One-way ANOVA

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Assumptions of Pearson’s Correlation

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Parametric Statistical Test Assumptions

    1. Level of Measurement of the Variable(s)


    2. Normality Test
    3. Homogeneity of Variance Test
    4. Linearity Test
    5. No Significant Outliers

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Level of Measurement of the Variable(s)
    1. Nominal Level
    ✓ the data collected are labels, names or categories
    ✓ frequencies or counts of observations belonging to the same category can be obtained
    ✓ every case must be nonoverlapping, mutually exclusive and exhaustive
    2. Ordinal Level
    ✓ data collected are labels with implied ordering
    ✓ does not indicate the magnitude of the differences between numbers.
    3. Interval Level
    ✓ the level of measurement that does not only indicate ordering but also the exact
    distance between them
    ✓ the data at this level has lack an absolute zero point
    4. Ratio Level
    ✓ the same as interval level, however, the number zero indicates the absence of the
    characteristics being measured (has absolute zero point)

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Level of Measurement

    In JASP,
    (Interval and Ratio)

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Level of Measurement

    Nominal

    Ordinal

    Ratio

    Nominal

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Normality Test
    The variable(s) should be approximately normally distributed.

    How to test for Normality?


    Shapiro-Wilk Test (Statistical Method)
    Hypotheses:
    𝐻0 : The data comes from a normal distribution.
    𝐻1 : The data deviates from normal distribution.
    Decision Rule:
    Result Decision Implication
    p-value ≤ 𝛼 Reject 𝐻0 The data deviates from normal distribution
    p-value > 𝛼 Fail to reject 𝐻0 The data comes from a normal distribution

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Homogeneity of Variance Test

    The variances of the variable across groups should be


    approximately equal; or
    The data shows homoscedasticity; or
    Homogeneity of Variances is satisfied.

    Note: This assumption is only necessary when comparing


    independent groups (not with repeated measure, that is,
    not for before and after data)
    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Homogeneity of Variance Test

    How to test for Homogeneity of Variances?


    Levene’s Test
    Hypotheses:
    𝐻0 : The variances are statistically homogeneous or equal.
    𝐻1 : The variances are statistically different.
    Decision Rule:
    Result Decision Implication
    p-value ≤ 𝛼 Reject 𝐻0 The variances are statistically different.
    p-value > 𝛼 Fail to reject 𝐻0 The variances are statistically homogeneous or equal.

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Linearity Test

    There is a linear relationship between the variables.


    How to test for linearity?
    Graphical using scatterplot

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Checking of Significant Outlier

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Checking of Significant Outlier

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Checking of Significant Outlier

    If there is an outlier,
    1. There may be a measurement error or data entry error.
    Correct the error if possible.
    2. That may not be a part of the population you are studying
    (i.e., unusual properties of conditions), you can
    legitimately remove the outlier.
    3. That may be a natural part of the population you are
    studying, you should not remove it.

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Violation of the assumption(s)

    May resort to…


    Nonparametric Test(s)
    ✓ used when assumptions are too badly violated for parametric
    tests
    ✓ do not require stringent assumptions about the population
    ✓ (Real Strength) can be used with skewed distributions
    Disadvantages
    ✓ more likely to commit Type II error (Fail to reject an incorrect 𝐻0 )
    ✓ cannot be used with complex research designs
    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    HANDS-ON
    APPLICATION IN JASP

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Level of Measurement

    Easily verified! Just look at your data.

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Normality Test

    The variable(s) should be approximately normally distributed.

    How to test for Normality?


    Shapiro-Wilk Test (Statistical Method)

    In JASP, Descriptives > Statistics > Distribution > Shapiro-Wilk


    Test

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Normality Test

    Hypotheses:
    𝐻0 : The data comes from a normal distribution.
    𝐻1 : The data deviates from normal distribution.

    Decision Rule:
    Result Decision Implication
    p-value ≤ 𝛼 Reject 𝐻0 The data deviates from normal distribution
    p-value > 𝛼 Fail to reject 𝐻0 The data comes from a normal distribution

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Homogeneity of Variance Test
    The variances of the variable across groups should be
    approximately equal; or
    The data shows homoscedasticity; or
    Homogeneity of Variances is satisfied.

    How to test for Homogeneity of Variances?


    Levene’s Test

    In JASP, ANOVA > Classical > ANOVA > Assumption Checks >
    Homogeneity tests > Homogeneity corrections > None
    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Homogeneity of Variance Test

    Hypotheses:
    𝐻0 : The variances are statistically homogeneous or equal.
    𝐻1 : The variances are statistically different.

    Decision Rule:

    Result Decision Implication


    The variances are statistically
    p-value ≤ 𝛼 Reject 𝐻0
    different.
    The variances are statistically
    p-value > 𝛼 Fail to reject 𝐻0
    homogeneous or equal.
    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Linearity Test

    In JASP, Descriptives > Customizable plots > Scatter plots

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    No significant outliers

    In JASP, Descriptives > Customizable plots > Boxplots

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions
    Workshop Datasets

    Basic Statistical Training and JASP Application Workshop Statistical Test Assumptions

    You might also like

    pFad - Phonifier reborn

    Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

    Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


    Alternative Proxies:

    Alternative Proxy

    pFad Proxy

    pFad v3 Proxy

    pFad v4 Proxy