Lecture 10 PY 427 Statistics 1 Fall 2006 Kin Ching Kong, Ph.D

Slides:



Advertisements
Similar presentations
Intro to ANOVA.
Advertisements

Chapter 10 Analysis of Variance (ANOVA) Part III: Additional Hypothesis Tests Renee R. Ha, Ph.D. James C. Ha, Ph.D Integrative Statistics for the Social.
PY 427 Statistics 1Fall 2006 Kin Ching Kong, Ph.D Lecture 9 Chicago School of Professional Psychology.
Independent Sample T-test Formula
ANOVA Analysis of Variance: Why do these Sample Means differ as much as they do (Variance)? Standard Error of the Mean (“variance” of means) depends upon.
Analysis of Variance: Inferences about 2 or More Means
Lecture 8 PY 427 Statistics 1 Fall 2006 Kin Ching Kong, Ph.D
Chapter 3 Analysis of Variance
Intro to Statistics for the Behavioral Sciences PSYC 1900
Lecture 11 Introduction to ANOVA.
Lecture 9: One Way ANOVA Between Subjects
Two Groups Too Many? Try Analysis of Variance (ANOVA)
Statistics for the Social Sciences Psychology 340 Spring 2005 Within Groups ANOVA.
One-way Between Groups Analysis of Variance
Anthony J Greene1 ANOVA: Analysis of Variance 1-way ANOVA.
Lecture 7 PY 427 Statistics 1 Fall 2006 Kin Ching Kong, Ph.D
PY 427 Statistics 1Fall 2006 Kin Ching Kong, Ph.D Lecture 6 Chicago School of Professional Psychology.
Comparing Several Means: One-way ANOVA Lesson 14.
Introduction to Analysis of Variance (ANOVA)
Chapter 9: Introduction to the t statistic
1 Chapter 13: Introduction to Analysis of Variance.
Analysis of Variance (ANOVA) Quantitative Methods in HPELS 440:210.
Psy B07 Chapter 1Slide 1 ANALYSIS OF VARIANCE. Psy B07 Chapter 1Slide 2 t-test refresher  In chapter 7 we talked about analyses that could be conducted.
Repeated ANOVA. Outline When to use a repeated ANOVA How variability is partitioned Interpretation of the F-ratio How to compute & interpret one-way ANOVA.
ANOVA Chapter 12.
AM Recitation 2/10/11.
Chapter 8 Introduction to Hypothesis Testing
Repeated Measures ANOVA
ANOVA Greg C Elvers.
1 1 Slide © 2006 Thomson/South-Western Slides Prepared by JOHN S. LOUCKS St. Edward’s University Slides Prepared by JOHN S. LOUCKS St. Edward’s University.
1 1 Slide © 2005 Thomson/South-Western Chapter 13, Part A Analysis of Variance and Experimental Design n Introduction to Analysis of Variance n Analysis.
COURSE: JUST 3900 INTRODUCTORY STATISTICS FOR CRIMINAL JUSTICE Instructor: Dr. John J. Kerbs, Associate Professor Joint Ph.D. in Social Work and Sociology.
Chapter 14: Repeated-Measures Analysis of Variance.
Chapter 13: Introduction to Analysis of Variance
One-Way Analysis of Variance Comparing means of more than 2 independent samples 1.
Chapter 11 HYPOTHESIS TESTING USING THE ONE-WAY ANALYSIS OF VARIANCE.
Comparing Several Means: One-way ANOVA Lesson 15.
Chapter 12: Introduction to Analysis of Variance
t(ea) for Two: Test between the Means of Different Groups When you want to know if there is a ‘difference’ between the two groups in the mean Use “t-test”.
Statistics 11 Confidence Interval Suppose you have a sample from a population You know the sample mean is an unbiased estimate of population mean Question:
Chapter 10: Analyzing Experimental Data Inferential statistics are used to determine whether the independent variable had an effect on the dependent variance.
Chapter 13 - ANOVA. ANOVA Be able to explain in general terms and using an example what a one-way ANOVA is (370). Know the purpose of the one-way ANOVA.
Chapter 14 Repeated Measures and Two Factor Analysis of Variance
Analysis of Variance (One Factor). ANOVA Analysis of Variance Tests whether differences exist among population means categorized by only one factor or.
Copyright © Cengage Learning. All rights reserved. 12 Analysis of Variance.
Chapter 10 The t Test for Two Independent Samples
Copyright (C) 2002 Houghton Mifflin Company. All rights reserved. 1 Understandable Statistics S eventh Edition By Brase and Brase Prepared by: Lynn Smith.
Chapter 12 Introduction to Analysis of Variance PowerPoint Lecture Slides Essentials of Statistics for the Behavioral Sciences Eighth Edition by Frederick.
Chapter 13 Repeated-Measures and Two-Factor Analysis of Variance
Psy 230 Jeopardy Related Samples t-test ANOVA shorthand ANOVA concepts Post hoc testsSurprise $100 $200$200 $300 $500 $400 $300 $400 $300 $400 $500 $400.
Psych 230 Psychological Measurement and Statistics Pedro Wolf November 18, 2009.
Hypothesis test flow chart frequency data Measurement scale number of variables 1 basic χ 2 test (19.5) Table I χ 2 test for independence (19.9) Table.
One-Way Analysis of Variance Recapitulation Recapitulation 1. Comparing differences among three or more subsamples requires a different statistical test.
Statistics for the Social Sciences Psychology 340 Spring 2009 Analysis of Variance (ANOVA)
1 Chapter 14: Repeated-Measures Analysis of Variance.
Introduction to ANOVA Research Designs for ANOVAs Type I Error and Multiple Hypothesis Tests The Logic of ANOVA ANOVA vocabulary, notation, and formulas.
Oneway/Randomized Block Designs Q560: Experimental Methods in Cognitive Science Lecture 8.
Outline of Today’s Discussion 1.Independent Samples ANOVA: A Conceptual Introduction 2.Introduction To Basic Ratios 3.Basic Ratios In Excel 4.Cumulative.
1 Statistics for the Behavioral Sciences (5 th ed.) Gravetter & Wallnau Chapter 13 Introduction to Analysis of Variance (ANOVA) University of Guelph Psychology.
Formula for Linear Regression y = bx + a Y variable plotted on vertical axis. X variable plotted on horizontal axis. Slope or the change in y for every.
1/54 Statistics Analysis of Variance. 2/54 Statistics in practice Introduction to Analysis of Variance Analysis of Variance: Testing for the Equality.
Stats/Methods II JEOPARDY. Jeopardy Estimation ANOVA shorthand ANOVA concepts Post hoc testsSurprise $100 $200$200 $300 $500 $400 $300 $400 $300 $400.
Independent Samples ANOVA. Outline of Today’s Discussion 1.Independent Samples ANOVA: A Conceptual Introduction 2.The Equal Variance Assumption 3.Cumulative.
©2013, The McGraw-Hill Companies, Inc. All Rights Reserved Chapter 4 Investigating the Difference in Scores.
Inferential Statistics Psych 231: Research Methods in Psychology.
Chapter 10: The t Test For Two Independent Samples.
Chapter 12 Introduction to Analysis of Variance
Chapter 14 Repeated Measures and Two Factor Analysis of Variance PowerPoint Lecture Slides Essentials of Statistics for the Behavioral Sciences Seventh.
I. Statistical Tests: Why do we use them? What do they involve?
Chapter 12: Introduction to Analysis of Variance
Presentation transcript:

Lecture 10 PY 427 Statistics 1 Fall 2006 Kin Ching Kong, Ph.D Chicago School of Professional Psychology Lecture 10 Kin Ching Kong, Ph.D

Agenda Analysis of Variance (continue) The Distribution of F-Ratios Review Intro. to ANOVA Hypotheses for ANOVA The Test Statistic for ANOVA: F The Logic of Analysis of Variance ANOVA Notation & Formulas The Distribution of F-Ratios Hypothesis Testing Measuring Effect Size Assumptions for the independent-measure ANOVA Post Hoc Tests

Analysis of Variance (ANOVA) ANOVA is used to compare two or more means The Question: Does the mean differences observed among the samples reflect mean differences among the populations? Two Possibilities: There really are no differences in the population means, the observed differences are due to chance (sampling error). The population means are truly different, and are partly responsible for differences in the sample means. Figure 13.1

Hypotheses for ANOVA The experiment: learning performance under three temperature: 50o, 70o, 90o Design: single-factor, between-subject (or independent-measures. The Hypotheses: H0: m1 = m2 = m3 (no differences in pop. means) H1: At least one pop. mean is different from the others, (or not all the pop. means are equal). There are many possible specific alternative hypotheses (e.g. all 3 means are different, first two means are identical but the third is different, the first is different and last two identical, etc.) So far we talked about inferential statistics using one sample to make inferences about an unknown population.

The Test Statistic for ANOVA The Test Statistic for t tests: t = obtained difference between sample means difference expected by chance (error) The Test Statistic for ANOVA, F-ratio: F = variance (differences) between sample means variance (differences) expected by chance (error) The F-ratio is based on variance rather than means Problem: how to define & calculate mean differences when there are more than two means. Solution: use variance to define and measure the size of differences among the sample means. E.g M1 = 20, M2 = 30, M3 =35 s2 = 58.33 M1 = 28, M2 = 30, M3 = 31 s2 = 2.33

The Logic of Analysis of Variance, The Data The experiment: learning under 3 different temperature The design: single factor between-subjects The data: DV = # of problems solved correctly. Temp. 50O Temp. 70O 4 1 3 2 6 M = 1 M = 4

The Logic of Analysis of Variance The Goal of ANOVA Measure the amount of variability in a data set and explain where it comes from. Step I: Measure Total Variance The variability in the whole data set (scores from all samples combined) Step II: Partition Total Variance into two components: Between-Treatment Variance Differences between treatment conditions Within-Treatment Variance: Differences within each treatment condition

The Between-Treatments Variance Measures how much difference exists between the treatment conditions (i.e. differences among treatment means). Two Sources For The Differences Between Treatments: Treatment effects Chance (unplanned & unpredictable difference) Individual differences Experimental error (Measurement error) To Demonstrate There Really Is A Treatment Effect: Show that differences between treatments are bigger than would be expected by chance alone.

The Within-Treatment Variance Measures differences due to chance, or when there is no treatment effect, H0 is true. Figure 13.2 Analysis of Variance

The F-Ratio: Test Statistic for ANOVA The F-Ratio compares the two component variance: F = Variance between treatments Variance within treatments = actual differences between treatments differences expected with no treatment effect F = treatment effect + differences due to chance differences due to chance (error) When there is no treatment effect: F = 0 + differences due to chance F is close to 1 differences due to chance When there is a treatment effect: The numerator significantly > the denominator F significantly > 1

ANOVA Notations SX2 = 106 G = 30 N = 15 k = 3 T1 = 5 T2 = 20 T3 = 5 SS1 = 6 SS2 = 6 SS3 = 4 n1 = 5 n2 = 5 n3 = 5 M1 = 1 M2 = 4 M3 = 1 Temp. 50O Temp. 70O 4 1 3 2 6 M = 1 M = 4

ANOVA Notations (Continue) k = number of treatment conditions (or number of levels of a factor) ni = number of scores in treatment i (i = 1 to k) N = total number of scores in the entire study. Ti = The total (SX) for treatment condition I G = The Grand Total = SX for all the scores, or G = ST For a one sample t test, the statistic is M, so the standard error is sM

ANOVA Formulas ANOVA Summary Table: Source SS df MS Between treatments F = MSbetween Within treatments MSwithin Total F = Variance between treatments Variance within treatments Variance, s2 = SS/df =MS F = MSbetween MSwithin To fill in the ANOVA summary table, need to calculate nine values: 3 SS, 3 df, 2 MS and F

ANOVA Formulas: SS Total Sum of Squares: SStotal = SX2 – (SX)2 = SX2 – G2 N N For our example: SStotal = SX2 – G2 = 106 -302/15 = 106 – 60 = 46 N Within-Treatment Sum of Squares: SSwithin = SSSwithin each treatment For our example: SSwithin = 6 + 6 + 4 = 16 Between-Treatment Sum of Squares: SSbetween = S T2 - G2 n N SSbetween = S T2 - G2 = 52 + 202 + 52 – 302 = 5 + 80 +5 -60 =30 n N 5 5 5 15

ANOVA Formulas, df Total Degrees of Freedom: dftotal = N - 1 For our example, dftotal = 15 – 1 = 14 Within-Treatment Degrees of Freedom: dfwithin = Sdfin each treatment = S(n-1) = N - k For our example, dfwithin = 15 – 3 = 12 Between-Treatment Degrees of Freedom: dfbetween = k – 1 For our example, dfbetween = 3 – 1 = 2

ANOVA Formulas, MS & F-Ratio Variance Between-Treatment, MSbetween: s2 = MSbetween = SSbetween/dfbetween For our example, MSbetween = 30/2 = 15 Variance Within-Treatment, MSwithin: s2 = MSwithin = SSwithin/dfwithin For our example, MSwithin = 16/12 = 1.33 The F-Ratio: F = MSbetween MSwithin For our example: F = 15/1.33 = 11.28

ANOVA Formulas Source SS df MS ANOVA Summary Table: Between treatments 30 2 15 F = 11.28 Within treatments 16 12 1.33 Total 46 14

The Distribution of F-Ratios Two characteristics of F values: F values are always positive because variances are always positive. When H0 is true, the numerator and denominator of the F-ratio estimate the same variance, thus, the ratio should be near 1. In other words, the distribution of F-ratios should pile up around 1.00 The Distribution of F-ratios: Cut off at zero (all positive values) Piles up around 1.00 Tapers off to the right. The exact shape of the F distribution depends on the df’s in the two variances. Figure 13.6 of your book

The F Distribution Table Table B.4 Find df of the numerator in first row. Find df of the denominator in first column The intersection of these two df’s is a pair of numbers: The smaller number is the critical value for a = .05 The larger number is the critical value for a = .01 Table 13.3 E.g. F = 4.18 with df = 2, 15. Is this value sufficient to reject H0 with a = .05? a =.01?

Hypothesis Testing (the experiment) Research Goal: Evaluate the effectiveness of three pain relievers (A, B, C) and a placebo. Experiment: Participants: four groups, n = 5 in each group Treatment (IV): Drug A, B, C and placebo Design: Single-factor, repeated-measures Dependent Variable (DV): Amount of time participants can withstand a painfully hot stimulus.

Hypothesis Testing (Data) SX2 = 262 T = 5 T = 10 T = 20 T = 25 SS = 8 SS = 8 SS = 6 SS = 10 Placebo Drug A Drug B Drug C 3 4 6 7 2 1 5

ANOVA Summary Table ANOVA Summary Table: Source SS df MS Between treatments F = Within treatments Total

ANOVA Calculations dftotal = N – 1 = 20 – 1 = 19 dfbetween = k – 1 = 4 – 1 =3 dfwithin = N – k = 16 SStotal = SX2 – G2/N = 262 – 602/20 = 82 SSwithin = SSSinside each treatment= 8 + 8 + 6 + 10 = 32 SSbetween = S T2/n – G2/N = 52/5 + 102/5 + 202/5 +252/5 – 602/20 = 50 MSbetween = SSbetween/ dfbetween = 50/3 = 16.67 MSwithin = SSwithin/ dfwithin = 32/16 = 2.00 F = MSbetween/ MSwithin = 16.67/2.00 = 8.33

ANOVA Formulas Source SS df MS ANOVA Summary Table: Between treatments 50 3 16.67 F = 8.33 Within treatments 32 16 2.00 Total 82 19

Hypothesis Testing with ANOVA Step 1: State the Hypotheses: H0: m1 = m2 = m3 = m4 (there is no treatment effect) H1: At least one of the treatment means are different. The level of significant is set at a = .05 Step 2: Locate the Critical Region: df = 3, 16, Fcritical = 3.24 Figure 13.7 Step 3: Calculate the test statistic: F = MSbetween / MSwithin = 8.33 Step 4: Make a decision: Since the test statistic, F = 8.33 falls in the critical region, reject H0 and concludes that there is a significant treatment effect.

Measuring Effect Size for ANOVA A significant difference Means that the difference observed in the samples is very unlikely to have occurred just by chance. Statistical significant does not necessarily mean large effect. Measuring effect size for ANOVA: r2: the percentage of variance accounted for by treatment r2 = SSbetween SStotal In published reports, the r2 value for ANOVA is usually call h2 (the Greek letter eta squared) For our example, h2 = 50/82 = 0.61 The pooled variance is used instead of the individual sample variances.

Assumptions Assumptions for the Independent-Measure ANOVA: The observations within each sample must be independent. The populations from which the samples are selected must be normal. The populations from which the samples are selected must have equal variances (homogeneity of variance)

Post Hoc Tests, Intro A significant F-ratio: Indicate that a significant difference exit, that not all the means are equal. Does not indicate which means are different and which are not. Example: M1 = 3, M2 = 5, M3 = 10 M2 – M1 = 2, M3 – M2 = 5, M3 – M1 = 7 A significant F indicate that at least one of these differences are significant, M3 = M1, but what about the other two? Post hoc tests are used to find out which of these difference are significant.

Post Hoc Tests & Type I Error are additional hypothesis test that are done after an analysis of variance revealed a significant difference They are performed to determine exactly which mean differences are significant and which are not. Post hoc test and Type I Error Post hoc tests compare two means at a time, i.e. pairwise comparisons. The process involve performing a series of separate hypothesis tests. Each of these tests includes the risk of a Type I error With more tests, the risk of a type I error accumulates Experimentwise alpha level: the overall probability of a Type I Error that accumulates over a series of separate hypothesis tests.

Post Hoc Tests, Planned Comparisons Controlling Type I Error (experimentwise alpha level) Whenever more than one test is done, need to be concerned about the experimentwise Type I Error. Planned Comparisons Planned Comparisons: specific mean differences are predicted by specific hypotheses before the study is conducted. Because a few specific comparisons were planned before the data were collected, many statisticians argue that planned comparisons can be conducted with a standard alpha, without concern about inflating the risk of a Type I error. Dunn Test: It is often recommended that researchers protect against an inflated alpha level by dividing alpha equally among the planned comparisons.

Post Hoc Tests, Tukey’s HSD Unplanned Comparisons sifting through the data by conducting a large number of comparisons. Tukey’s Honestly Significant Difference (HSD) Compute a single value (HSD) that determines the minimum difference that is necessary for significance. If a pairwise difference is > Tukey’s HSD, you conclude that there is a significant difference between the two means. HSD = q n = number of scores in each treatment, Tukey’s HSD test requires equal n’s Table B.5 to find value of q. (k = number of treatment conditions, df error term = df for the denominator of the F-ratio)

Post Hoc Tests, Tukey’s HSD, Example M1 = 3.00 M2 = 5.44 M3 = 7.00 ANOVA Summary Table Source SS df MS Between 73.19 2 36.60 F (2, 24) = 9.15 Within 96.00 24 4.00 Total 169.19 26 HSD = q = 3.53 = 2.36 M2 – M1 = 2.44, significant M3 – M1 = 4.00, significant M3 – M2 = 1.56, nonsignificant