chapter 11 hypothesis testing using the one-way analysis of variance
TRANSCRIPT
Chapter 11
HYPOTHESIS TESTING USING THEONE-WAY ANALYSIS OF VARIANCE
Moving Forward
Your goals in this chapter are to learn:• The terminology of analysis of variance• When and how to compute • Why should equal 1 if H0 is true, and why
it is greater than 1 if H0 is false
• When and how to compute Tukey’s HSD• How eta squared describes effect size
obtF
obtF
Analysis of Variance
• The analysis of variance is the parametric procedure for determining whether significant differences occur in an experiment with two or more sample means
• In an experiment involving only two conditions of the independent variable, you may use either a t-test or the ANOVA
An Overview of ANOVA
One-Way ANOVA
• Analysis of variance is abbreviated as ANOVA• An independent variable is also called a factor• Each condition of the independent variable is
called a level or treatment• Differences produced by the independent
variable are a treatment effect
Between-Subjects
• A one-way ANOVA is performed when one independent variable is tested in the experiment
• When an independent variable is studied using independent samples in all conditions, it is called a between-subjects factor
• A between-subjects factor involves using the formulas for a between-subjects ANOVA
Within-Subjects Factor
• When a factor is studied using related (dependent) samples in all levels, it is called a within-subjects factor
• This involves a set of formulas called a within-subjects ANOVA
Diagram of a Study Having ThreeLevels of One Factor
Assumptions of the ANOVA
1. All conditions contain independent samples
2. The dependent scores are normally distributed, interval or ratio scores
3. The variances of the populations are homogeneous
Experiment-Wise Error
• The probability of making a Type I error somewhere among the comparisons in an experiment is called the experiment-wise error rate
• When we use a t-test to compare only two means in an experiment, the experiment-wise error rate equals
Comparing Means
• When there are more than two means in an experiment, the multiple t-tests result in an experiment-wise error rate much larger than the we have selected
• Using the ANOVA allows us to make all our decisions and keep the experiment-wise error rate equal to
Statistical Hypotheses
kH 210 :
equalaresallnot:a H
The F-Test
• The statistic for the ANOVA is F
• When Fobt is significant, it indicates only that somewhere among the means at least two of them differ significantly
• It does NOT indicate which specific means differ significantly
• When the F-test is significant, we perform post hoc comparisons
Post Hoc Comparisons
• Post hoc comparisons are like t-tests
• We compare all possible pairs of level means from a factor, one pair at a time to determine which means differ significantly from each other
Components of the ANOVA
Mean Squares
• The mean square within groups describes the variability in scores within the conditions of an experiment. It is symbolized by MSwn.
• The mean square between groups describes the differences between the means of the conditions in a factor. It is symbolized by MSbn.
The F-Ratio
• The F-ratio equals the mean square between groups divided by the mean square within groups
• When H0 is true, Fobt should equal 1
• When H0 is false, Fobt should be greater than 1
wn
bnobt MS
MSF
Performing the ANOVA
Sum of Squares
• The computations for the ANOVA require the use of several sums of squared deviations
• The sum of squares is the sum of the squared deviations of a set of scores around the mean of those scores
• It is symbolized by SS
Summary Table of a One-way ANOVA
Computing Fobt
1. Compute the sums and means• • •
for each level. Add the from all levels to get . Add together the from all levels to get . Add the ns together to get N.
X2X
X
X
totX 2X2totX
Computing Fobt
2. Compute the total sum of squares (SStot)
N
XXSS
2tot2
tottot
)(
Computing Fobt
3. Compute the sum of squares between groups (SSbn)
N
X
n
XSS
2tot
2
bn
)(
columnin
)columnin(
Computing Fobt
4. Compute the sum of squares within groups (SSwn)
bntotwn SSSSSS
Computing Fobt
Compute the degrees of freedom• The degrees of freedom between groups
equals k – 1 where k is the number of levels in the factor
• The degrees of freedom within groups equals N – k
• The degrees of freedom total equals N – 1
Computing Fobt
Compute the mean squares
•
•
bn
bnbn df
SSMS
wn
wnwn df
SSMS
Computing Fobt
Compute Fobt
wn
bnobt MS
MSF
Sampling Distribution of FWhen H0 Is True
Degrees of Freedom
The critical value of F (Fcrit) depends on
• The degrees of freedom (both the dfbn = k – 1 and the dfwn = N – k)
• The selected
• The F-test is always a one-tailed test
Tukey’s HSD Test
When the ns in all levels of the factor are equal, use the Tukey HSD multiple comparisons test
where qk is found using Table 5 in Appendix B
n
MSqHSD k
wn)(
Tukey’s HSD Test
• Determine the difference between each pair of means
• Compare each difference between the means to the HSD
• If the absolute difference between two means is greater than the HSD, then these means differ significantly
Effect Size and Eta2
Proportion of Variance Accounted For
Eta squared indicates the proportion of variance in the dependent variable scores that is accounted for by changing the levels of a factor
)( 2
tot
bn2
SS
SS
Example
Using the following data set, conduct a one-way ANOVA. Use = 0.05.
Group 1 Group 2 Group 3
14 14 10 13 11 15
13 10 12 11 14 13
14 15 11 10 14 15
Example
611.5518
2292969
)(
2
2tot2
tottot
N
XXSS
Example
111.2218
229
6
82
6
67
6
80
)(
columnin
)columnin(
2222
2tot
2
bn
N
X
n
XSS
Example
50.33
111.22611.55bntotwn
SSSSSS
Example
• dfbn = k – 1 = 3 – 1 = 2
• dfwn = N – k = 18 – 3 = 15
• dftot = N – 1 = 18 – 1 = 17
Example
055.112
111.22
bn
bnbn
df
SSMS
233.215
50.33
wn
wnwn
df
SSMS
951.4233.2
055.11
wn
bnobt
MS
MSF
Example
• Fcrit for 2 and 15 degrees of freedom and = 0.05 is 3.68
• Since Fobt = 4.951, the ANOVA is significant
• A post hoc test must now be performed
Example
242.26
233.2675.3)( wn
n
MSqHSD k
334.0333.13667.13
500.2167.11667.13
166.2167.11333.13
13
23
21
XX
XX
XX
Example
Because 2.50 > 2.242 (HSD), the mean of sample 3 is significantly different from the mean of sample 2.