Introduction to Analysis of Variance (ANOVA) Limitations of the t-test



Similar documents
Independent t- Test (Comparing Two Means)

Chapter 5 Analysis of variance SPSS Analysis of variance

INTERPRETING THE ONE-WAY ANALYSIS OF VARIANCE (ANOVA)

SCHOOL OF HEALTH AND HUMAN SCIENCES DON T FORGET TO RECODE YOUR MISSING VALUES

UNDERSTANDING THE TWO-WAY ANOVA

Chapter 2 Probability Topics SPSS T tests

individualdifferences

Two Related Samples t Test

One-Way Analysis of Variance

Class 19: Two Way Tables, Conditional Distributions, Chi-Square (Text: Sections 2.5; 9.1)

1. What is the critical value for this 95% confidence interval? CV = z.025 = invnorm(0.025) = 1.96

Examining Differences (Comparing Groups) using SPSS Inferential statistics (Part I) Dwayne Devonish

1.5 Oneway Analysis of Variance

An Introduction to Statistics Course (ECOE 1302) Spring Semester 2011 Chapter 10- TWO-SAMPLE TESTS

Study Guide for the Final Exam

Analysis of Data. Organizing Data Files in SPSS. Descriptive Statistics

PSYCHOLOGY 320L Problem Set #3: One-Way ANOVA and Analytical Comparisons

UNDERSTANDING THE INDEPENDENT-SAMPLES t TEST

COMPARISONS OF CUSTOMER LOYALTY: PUBLIC & PRIVATE INSURANCE COMPANIES.

Section 13, Part 1 ANOVA. Analysis Of Variance

Research Methods & Experimental Design

UNDERSTANDING THE DEPENDENT-SAMPLES t TEST

General Method: Difference of Means. 3. Calculate df: either Welch-Satterthwaite formula or simpler df = min(n 1, n 2 ) 1.

THE KRUSKAL WALLLIS TEST

Chapter 7 Section 7.1: Inference for the Mean of a Population

UNDERSTANDING ANALYSIS OF COVARIANCE (ANCOVA)

7. Comparing Means Using t-tests.

Statistics Review PSY379

ANOVA ANOVA. Two-Way ANOVA. One-Way ANOVA. When to use ANOVA ANOVA. Analysis of Variance. Chapter 16. A procedure for comparing more than two groups

Recall this chart that showed how most of our course would be organized:

Descriptive Statistics

One-Way ANOVA using SPSS SPSS ANOVA procedures found in the Compare Means analyses. Specifically, we demonstrate

DDBA 8438: The t Test for Independent Samples Video Podcast Transcript

13: Additional ANOVA Topics. Post hoc Comparisons

Chapter 9. Two-Sample Tests. Effect Sizes and Power Paired t Test Calculation

Randomized Block Analysis of Variance

Chapter 7. One-way ANOVA

3.4 Statistical inference for 2 populations based on two samples

Unit 26 Estimation with Confidence Intervals

SPSS Resources. 1. See website (readings) for SPSS tutorial & Stats handout

Statistics. One-two sided test, Parametric and non-parametric test statistics: one group, two groups, and more than two groups samples

Profile analysis is the multivariate equivalent of repeated measures or mixed ANOVA. Profile analysis is most commonly used in two cases:

Lesson 1: Comparison of Population Means Part c: Comparison of Two- Means

Testing Group Differences using T-tests, ANOVA, and Nonparametric Measures

CHAPTER 11 CHI-SQUARE AND F DISTRIBUTIONS

Additional sources Compilation of sources:

Non-Parametric Tests (I)

A Basic Guide to Analyzing Individual Scores Data with SPSS

The Dummy s Guide to Data Analysis Using SPSS

" Y. Notation and Equations for Regression Lecture 11/4. Notation:

SPSS 3: COMPARING MEANS

An analysis method for a quantitative outcome and two categorical explanatory variables.

Two-sample hypothesis testing, II /16/2004

Statistiek II. John Nerbonne. October 1, Dept of Information Science

Nonparametric Statistics

EPS 625 INTERMEDIATE STATISTICS FRIEDMAN TEST

CONTINGENCY TABLES ARE NOT ALL THE SAME David C. Howell University of Vermont

1 Basic ANOVA concepts

HYPOTHESIS TESTING: POWER OF THE TEST

SPSS Guide How-to, Tips, Tricks & Statistical Techniques

SPSS Tests for Versions 9 to 13

Concepts of Experimental Design

One-Way Analysis of Variance (ANOVA) Example Problem

Two-Sample T-Tests Assuming Equal Variance (Enter Means)

Final Exam Practice Problem Answers

Part 2: Analysis of Relationship Between Two Variables

Good luck! BUSINESS STATISTICS FINAL EXAM INSTRUCTIONS. Name:

Two-Sample T-Tests Allowing Unequal Variance (Enter Difference)

Effects of Different Response Types on Iranian EFL Test Takers Performance

ABSORBENCY OF PAPER TOWELS

Mind on Statistics. Chapter 13

One-Way Independent ANOVA: Exam Practice Sheet

Outline. Topic 4 - Analysis of Variance Approach to Regression. Partitioning Sums of Squares. Total Sum of Squares. Partitioning sums of squares

UNIVERSITY OF NAIROBI

statistics Chi-square tests and nonparametric Summary sheet from last time: Hypothesis testing Summary sheet from last time: Confidence intervals

C. The null hypothesis is not rejected when the alternative hypothesis is true. A. population parameters.

Unit 31 A Hypothesis Test about Correlation and Slope in a Simple Linear Regression

Section 12 Part 2. Chi-square test

II. DISTRIBUTIONS distribution normal distribution. standard scores

Introduction. Hypothesis Testing. Hypothesis Testing. Significance Testing

Chi-square test Fisher s Exact test

Chapter Seven. Multiple regression An introduction to multiple regression Performing a multiple regression on SPSS

Outline. Definitions Descriptive vs. Inferential Statistics The t-test - One-sample t-test

Introduction to Quantitative Methods

Multivariate Analysis of Variance (MANOVA): I. Theory

Linear Models in STATA and ANOVA

CHAPTER 13. Experimental Design and Analysis of Variance

Regression Analysis: A Complete Example

Part 3. Comparing Groups. Chapter 7 Comparing Paired Groups 189. Chapter 8 Comparing Two Independent Groups 217

Chapter 8 Hypothesis Testing Chapter 8 Hypothesis Testing 8-1 Overview 8-2 Basics of Hypothesis Testing

Statistiek I. Proportions aka Sign Tests. John Nerbonne. CLCG, Rijksuniversiteit Groningen.

Standard Deviation Estimator

Introduction to Hypothesis Testing. Hypothesis Testing. Step 1: State the Hypotheses

Testing Research and Statistical Hypotheses

12: Analysis of Variance. Introduction

Lecture Notes Module 1

Nonparametric Two-Sample Tests. Nonparametric Tests. Sign Test

August 2012 EXAMINATIONS Solution Part I

Rank-Based Non-Parametric Tests

T-test & factor analysis

Transcription:

Introduction to Analysis of Variance (ANOVA) The Structural Model, The Summary Table, and the One- Way ANOVA Limitations of the t-test Although the t-test is commonly used, it has limitations Can only test differences between groups High school class? College year? Can examine ONLY the effects of 1 IV on 1 DV Limited to single group OR repeated measures designs 1

Limitations of the t-test Testing differences between group means IV: Gender (Male & Female) IV: High-school class (First-year, Sophomore, Junior, & Senior) Using the t-test, we must either collapse categories or not run the analysis Limitations of the t-test 1 Independent Variable Gender differences in depression IV: Gender (Male & Female) DV: Level of depression (BDI score) Independent Variables Gender and social support on depression IV 1 : Gender (Male & Female) IV : Social support (High, Medium, & Low) DV: Level of depression (BDI score)

Limitations of the t-test or more Independent Variables Simultaneously examine the impact of or more IVs on a single DV Examine how the effects of or more IVs COMBINE to affect a single DV Limitations of the t-test Single time point OR repeated measures designs 1 group at time points = repeated measures groups at 1 time point = independent groups Single time point AND repeated measures designs or more groups at or more time points 3

The Analysis of Variance (ANOVA) The ANOVA can test hypotheses that the t-test cannot Probably the most commonly abused statistical test Many varieties of ANOVA One-Way (between subjects) Factorial ANOVA (between or within subjects) Repeated Measures (within subjects) Mixed-Model (between & within subjects) One-Way ANOVA Varieties of ANOVA 1 continuous Dependent Variable 1 Independent Variable consisting of or more categorical groups The one-way ANOVA with groups is equivalent to the independent groups t-test 4

Varieties of ANOVA Factorial ANOVA 1 continuous Dependent Variable or more Independent Variables consisting of or more categorical groups for each IV IVs = Two-Way Factorial ANOVA 3 IVs = Three-Way Factorial ANOVA We call these factorial designs because EACH level of each IV is paired with EVERY level of ALL other IVs x Contingency Table Level 1 IV Level IV 1 Level 1 Level DATA DATA Col 1 DATA DATA Col Row 1 Row Note: Each Level 1 of IV 1 is paired with BOTH Level 1 and Level of IV 5

x Contingency Table Social Support High Low Gender Male DATA DATA Row 1 Female DATA DATA Row Col 1 Col Note: Each Level 1 of IV 1 is paired with BOTH Level 1 and Level of IV Repeated Measures ANOVA Varieties of ANOVA Time points = IV The DV is assessed at EACH time point 6

Mixed-Model ANOVA Varieties of ANOVA 1 continuous Dependent Variable 1 or more Independent Variables consisting of or more categorical groups (between) 1 Independent Variable consisting of or more categorical time points (within) The DV is assessed at EACH time point ANOVA models we will consider One-Way ANOVA Two- and Three-way Factorial ANOVA Repeated measures ANOVA Mixed-model ANOVA ANOVA 7

Choosing the Best Test The Underlying Model A statistical model by example: Assume: the average 18 year old human being weighs approximately 138 pounds Men, on average, weigh 1 pounds more than the average human weight Women, on average, weigh 10 pounds less than the average human weight 8

The Underlying Model For any given human being, I can break weight down into 3 components: Average weight for all individual 138 lbs Average weight for each group Men: +1 lbs Women: - 10 lbs The individual s unique difference The Underlying Model Male weight Weight = 138 lbs + 1 lbs + uniqueness Female weight Weight = 138 lbs 10 lbs + uniqueness If you understand this process, you understand the basic theory behind the ANOVA 9

Partitioning Variance The idea behind the ANOVA test is to divide or separate (partition) variance observed in the data into categories of what we CAN and what we CANNOT explain Variance Explained Unexplained The Structural Model Mathematically, we partition the total variance of our data using the structural form of the ANOVA model X ij = µ + τ j +ε ij The structural model translates as follows: The score for any single individual is equal to the sum of the population mean plus the mean of the group plus the individual s unique contribution 10

For our weight example: The Structural Model µ = population weight = 138 lbs τ = group difference in weight = 1 or 10 lbs ε = unique contribution of an individual s score µ& τ can be explained ε cannot be explained Uniqueness Oftentimes, we value our uniqueness In statistics, unique variance is BAD Since we can t explain unique variance, we call it error Thus, the ANOVA seeks to examine the relative proportion of explainable variance in our data to the unexplainable variance 11

Assumptions of the ANOVA Owing to the mathematical construction of the ANOVA, the underlying assumptions of the test are very important Homogeneity of variance Normality Independence of Observations The Null Hypothesis Homogeneity of Variance Homogeneity of variance refers to the variance for each group being equal to the variance of every other group Really, we mean that the variance of each group is equal to the variance of the error for the total analysis σ 1 = σ = σ 3 = σ j = σ e 1

Homogeneity of Variance Heterogeneity of variance is another BAD thing Heterogeneous variances can greatly influence the results you obtain, making it either more or less likely that you will reject H 0 Visual inspection of variances Tests of homogeneity of variance Normality The ANOVA procedure assumes that scores are normally distributed More accurately, it assumes that ERRORS are normally distributed Random sampling and random assignment Lacking normality, consider mathematical transformations Logarithmic & square root transformations 13

Independence of Observations Simple: The scores for 1 group are not dependent on the scores from another group Don t share subjects between groups If violated What is wrong with your experimental design? Are you using the appropriate test? The Null Hypothesis Less an assumption and more a theoretical point: H 0 : µ 1 = µ = µ 3 = µ 4 = µ 5 This is almost ALWAYS the basic form of your null hypothesis 14

Calculating the One-Way ANOVA In order to calculate the One-Way ANOVA statistic, we need to complete a number of intermediate steps Because there are several intermediate steps, we keep track of our progress with something called a summary table The Summary Table Source df Sum of Squares (SS) Mean Square (MS) F 15

One-Way ANOVA: Partitioning Variance The idea behind the ANOVA test is to divide or separate (partition) variance observed in the data into categories of what we CAN and what we CANNOT explain Variance The Summary Table Source df (k-1) Sum of Squares (SS) Mean Square (MS) F k(n-1) (N-1) Note: df + df =df 16

Note: x j Sum of Squares = The treatment group mean x.. = The grand mean (mean of all scores) X ij = Each individual score The Summary Table Source df (k-1) Sum of Squares (SS) SS n( X X ) = j.. Mean Square (MS) F k(n-1) SS = SS SS (N-1) SS = ( X ij X.. ) 17

The Summary Table Source df (k-1) Sum of Squares (SS) SS n( X X ) = j.. Mean Square (MS) SS MS = df F k(n-1) SS = SS SS SS MS = df (N-1) SS = ( X ij X.. ) The Summary Table Source df (k-1) Sum of Squares (SS) SS n( X X ) = j.. Mean Square (MS) SS MS = df F MS F = MS k(n-1) SS = SS SS SS MS = df (N-1) SS = ( X ij X.. ) 18

Example: Anorexia Nervosa 3 Group Tx Control -5-6 0-1.5 - -5-8 1-3 -10-4 IPT 0 1-1 - 0 0-1 4-1.5 CBT 4 6 4 0 4 3 3.5 6 9 0 6 3 Example: Anorexia Nervosa 3 Group Tx Change in Weight Descriptives Control IPT CBT 95% Confidence Interval for Mean N Mean Std. Deviation Std. Lower Bound Upper Bound 1-3.4583 3.6014 1.03985-5.7470-1.1696 11.318 1.7966.54051 -.8861 1.55 14 3.7500.45537.6563.333 5.1677 37.3919 4.0451.66501 -.9568 1.7406 19

Example: Anorexia Nervosa 3 Group Tx Test of Homogeneity of Variances Change in Weight Levene Statistic df1 df Sig..666 34.084 H 0 : σ 1 = σ = σ 3 H 1 : σ 1 σ σ 3 Fail to reject H 0... Example: Anorexia Nervosa 3 Group Tx Change in Weight ANOVA Between Groups Within Groups Sum of Squares df Mean Square F Sig. 335.87 167.914.544.000 53.41 34 7.448 589.068 36 F(,34) =.54, p <.05 0

Example: Anorexia Nervosa Group Tx Control -5-6 0-1.5 - -5-8 1-3 -10-4 CBT 4 6 4 0 4 3 3.5 6 9 0 6 3 Example: Anorexia Nervosa Group Tx Descriptives Change in Weight Control CBT 95% Confidence Interval for Mean N Mean Std. Deviation Std. Lower Bound Upper Bound Minimum Maximum 1-3.4583 3.6014 1.03985-5.7470-1.1696-10.00.00 14 3.7500.45537.6563.333 5.1677.00 9.00 6.431 4.7195.9557-1.483.393-10.00 9.00 1

Example: Anorexia Nervosa Group Tx Test of Homogeneity of Variances Change in Weight Levene Statistic df1 df Sig..81 1 4.144 H 0 : σ 1 = σ = σ 3 H 1 : σ 1 σ σ 3 Fail to reject H 0 Example: Anorexia Nervosa Group Tx Change in Weight ANOVA Between Groups Within Groups Sum of Squares df Mean Square F Sig. 335.74 1 335.74 36.443.000 1.104 4 9.13 556.846 5 F(1,4) = 36.44, p <.05

Example: Anorexia Nervosa Group Tx Independent Groups t-test Group Statistics Experimental Group Change in Weight Control CBT One-way ANOVA N Mean Std. Deviation Std. Mean 1-3.4583 3.6014 1.03985 14 3.7500.45537.6563 Descriptives Change in Weight Control CBT 95% Confidence Interval for Mean N Mean Std. Deviation Std. Lower Bound Upper Bound 1-3.4583 3.6014 1.03985-5.7470-1.1696 14 3.7500.45537.6563.333 5.1677 6.431 4.7195.9557-1.483.393 Example: Anorexia Nervosa Group Tx Independent groups t-test Independent Samples Test Levene's Test for Equality of Variances t-test for Equality of Means Change in Weight Equal variances assumed Equal variances not assumed One-way ANOVA F Sig. t df Sig. (-tailed) Mean Difference Std. Difference.81.144-6.037 4.000-7.0833 1.19406 ANOVA -5.86 18.96.000-7.0833 1.960 Change in Weight Sum of Squares df Mean Square F Sig. Between Groups 335.74 1 335.74 36.443.000 Within Groups 1.104 4 9.13 556.846 5 3

Example: Anorexia Nervosa Group Tx Shouldn t the results of the F- and t-tests be equal if the tests are equivalent? F(1,4) = 36.44, p <.05 t(4) = -6.04, p <.05 t = F 4