13.anova1.06
DESCRIPTION
AnovaTRANSCRIPT
![Page 1: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/1.jpg)
Statistics for the Social SciencesPsychology 340
Fall 2006
Analysis of Variance (ANOVA)
![Page 2: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/2.jpg)
Statistics for the Social Sciences
Outline (for week)
• Basics of ANOVA• Why• Computations• ANOVA in SPSS• Post-hoc and planned comparisons• Assumptions • The structural model in ANOVA
![Page 3: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/3.jpg)
Statistics for the Social Sciences
Outline (for week)
• Basics of ANOVA• Why• Computations• ANOVA in SPSS• Post-hoc and planned comparisons• Assumptions • The structural model in ANOVA
![Page 4: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/4.jpg)
Statistics for the Social Sciences
Example
• Effect of knowledge of prior behavior on jury decisions– Dependent variable: rate how innocent/guilty
– Independent variable: 3 levels• Criminal record
• Clean record
• No information (no mention of a record)
![Page 5: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/5.jpg)
Statistics for the Social Sciences
Statistical analysis follows design
• The 1 factor between groups ANOVA:– More than two
– Independent & One score per subject
– 1 independent variable
![Page 6: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/6.jpg)
Statistics for the Social Sciences
Analysis of Variance
XB XAXC
Criminal record Clean record No information
10 5 4
7 1 6
5 3 9
10 7 3
8 4 3
XA =8.0
SSA =18.0
XB =4.0
SSB =20.0
XC =5.0
SSC =26.0
• More than two groups– Now we can’t just
compute a simple difference score since there are more than one difference
Generic test statistic
observed difference
difference expected by chance
![Page 7: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/7.jpg)
Statistics for the Social Sciences
Analysis of Variance
XB XAXC
Criminal record Clean record No information
10 5 4
7 1 6
5 3 9
10 7 3
8 4 3
XA =8.0 XB =4.0 XC =5.0
SSA =18.0 SSB =20.0 SSC =26.0
– Need a measure that describes several difference scores
– Variance• Variance is
essentially an average squared difference
Test statistic
Observed variance
Variance from chanceF-ratio =
• More than two groups
![Page 8: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/8.jpg)
Statistics for the Social Sciences
Testing Hypotheses with ANOVA
– Step 1: State your hypotheses
• Hypothesis testing: a five step program
• Null hypothesis (H0)
– All of the populations all have same mean
• Alternative hypotheses (HA)
– Not all of the populations all have same mean
– There are several alternative hypotheses
– We will return to this issue later
H0 :μA =μB =μC
![Page 9: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/9.jpg)
Statistics for the Social Sciences
Testing Hypotheses with ANOVA
– Step 2: Set your decision criteria
– Step 3: Collect your data
– Step 4: Compute your test statistics • Compute your estimated variances
• Compute your F-ratio
• Compute your degrees of freedom (there are several)
– Step 5: Make a decision about your null hypothesis
• Hypothesis testing: a five step program– Step 1: State your hypotheses
– Additional tests• Reconciling our multiple alternative hypotheses
![Page 10: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/10.jpg)
Statistics for the Social Sciences
Step 4: Computing the F-ratio
• Analyzing the sources of variance– Describe the total variance in the dependent measure
• Why are these scores different?
XB XAXC
• Two sources of variability– Within groups
– Between groups
![Page 11: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/11.jpg)
Statistics for the Social Sciences
Step 4: Computing the F-ratio
• Within-groups estimate of the population variance – Estimating population variance from variation from
within each sample• Not affected by whether the null hypothesis is true
XB XAXC
Different people within each group
give different ratings
Different people within each group
give different ratings
![Page 12: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/12.jpg)
Statistics for the Social Sciences
• Between-groups estimate of the population variance – Estimating population variance from variation between
the means of the samples• Is affected by the null hypothesis is true
Step 4: Computing the F-ratio
XB XAXC
There is an effectof the IV, so the
people in differentgroups give different
ratings
There is an effectof the IV, so the
people in differentgroups give different
ratings
![Page 13: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/13.jpg)
Statistics for the Social Sciences
Partitioning the variance
Total variance
Stage 1
Between groups variance
Within groups variance
Note: we will start with SS, but willget to variance
Note: we will start with SS, but willget to variance
![Page 14: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/14.jpg)
Statistics for the Social Sciences
Partitioning the variance
Total varianceCriminal record Clean record No information
10 5 4
7 1 6
5 3 9
10 7 3
8 4 3
GM =X∑
N=8515
=5.67
SSTotal = X−GM( )∑ 2= 10 −5.67( )2 + ...+ 3−5.67( )2 =107.33
• Basically forgetting about separate groups– Compute the Grand Mean
(GM)– Compute squared deviations
from the Grand Mean
dfTotal =N−1=15 −1=14
![Page 15: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/15.jpg)
Statistics for the Social Sciences
Partitioning the variance
Total variance
Stage 1
Between groups variance
Within groups variance
SSTotal = X−GM( )∑ 2
dfTotal =N−1
![Page 16: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/16.jpg)
Statistics for the Social Sciences
Partitioning the variance
Within groups varianceCriminal record Clean record No information
10 5 4
7 1 6
5 3 9
10 7 3
8 4 3
SSWithin = SSeach group∑ =SSA +SSB +SSC =64
• Basically the variability in each group– Add up of the
SS from all of the groups
dfWithin = dfeach group∑ =4 + 4 + 4 =12
XA =8.0 XB =4.0 XC =5.0
SSA =18.0 SSB =20.0 SSC =26.0
![Page 17: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/17.jpg)
Statistics for the Social Sciences
Partitioning the variance
Total variance
Stage 1
Between groups variance
Within groups variance
SSTotal = X−GM( )∑ 2
dfTotal =N−1
SSWithin = SSeach group∑dfWithin = dfeach group∑
![Page 18: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/18.jpg)
Statistics for the Social Sciences
Partitioning the variance
Between groups varianceCriminal record Clean record No information
10 5 4
7 1 6
5 3 9
10 7 3
8 4 3
SSBetween = n X−GM( )∑ 2
• Basically how much each group differs from the Grand Mean– Subtract the GM
from each group mean
– Square the diffs
– Weight by number of scores
dfbetween =#groups−1=3−1=2
XA =8.0 XB =4.0 XC =5.0
SSA =18.0 SSB =20.0 SSC =26.0
=5 8 − 5.67( )2
+ 5 4 − 5.67( )2
+ 5 5 − 5.67( )2
=43.3
![Page 19: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/19.jpg)
Statistics for the Social Sciences
Partitioning the variance
Total variance
Stage 1
Between groups variance
Within groups variance
SSTotal = X−GM( )∑ 2
dfTotal =N−1
SSWithin = SSeach group∑dfWithin = dfeach group∑
SSBetween = n X−GM( )∑ 2
dfbetween =#groups−1
![Page 20: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/20.jpg)
Statistics for the Social Sciences
Partitioning the variance
Total variance
Stage 1
Between groups variance
Within groups variance
SSTotal = X−GM( )∑ 2
dfTotal =N−1
SSWithin = SSeach group∑dfWithin = dfeach group∑
SSBetween = n X−GM( )∑ 2
dfbetween =#groups−1
Now we return to variance. But, we call it Means Square (MS)
Now we return to variance. But, we call it Means Square (MS)
Recall:Recall:
variance =SSdf
MSWithin =SSWithindfWithin
MSBetween =SSBetweendfBetween
![Page 21: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/21.jpg)
Statistics for the Social Sciences
Partitioning the variance
Mean Squares (Variance)
SSBetween =43.3
dfbetween =2SSWithin =64dfWithin =12
MSBetween =43.32
=21.67 MSWithin =6412
=5.33
Within groups variance
Between groups variance
![Page 22: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/22.jpg)
Statistics for the Social Sciences
• The F ratio– Ratio of the between-groups to the within-groups
population variance estimate
Step 4: Computing the F-ratio
• The F distribution• The F table
Observed variance
Variance from chanceF-ratio = =
MSBetween
MSWithin
=21.67
5.33= 4.07
Do we reject or failto reject the H0?
Do we reject or failto reject the H0?
![Page 23: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/23.jpg)
Statistics for the Social Sciences
Carrying out an ANOVA
• The F distribution • The F table– Need two df’s
• dfbetween (numerator)
• dfwithin (denominator)
– Values in the table correspond to critical F’s
• Reject the H0 if your computed value is greater than or equal to the critical F
– Separate tables for 0.05 & 0.01
![Page 24: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/24.jpg)
Statistics for the Social Sciences
Carrying out an ANOVA
• The F table– Need two df’s
• dfbetween (numerator)
• dfwithin (denominator)
– Values in the table correspond to critical F’s
• Reject the H0 if your computed value is greater than or equal to the critical F
– Separate tables for 0.05 & 0.01
F =MSBetweenMSWithin
=21.67
5.33= 4.07
Do we reject or failto reject the H0?
Do we reject or failto reject the H0?
– From the table (assuming 0.05) with 2 and 12 degrees of freedom the critical F = 3.89.
– So we reject H0 and conclude that not all groups are the same
![Page 25: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/25.jpg)
Statistics for the Social Sciences
Summary of Example ANOVA
Criminal record Clean record No information
10 5 4
7 1 6
5 3 9
10 7 3
8 4 3
XA =8.0
SSA =18.0
XB =4.0
SSB =20.0
XC =5.0
SSC =26.0
GM =X∑
N=8515
=5.67
SSTotal = X−GM( )∑ 2=107.33
dfTotal =N−1=15 −1=14
SSWithin = SSeach group∑ =64
dfWithin = dfeach group∑ =4 + 4 + 4 =12
SSBetween = n X−GM( )∑ 2=43.3
dfbetween =#groups−1=3−1=2
F =MSBetweenMSWithin
=21.67
5.33= 4.07
MSWithin =SSWithindfWithin
=5.33
MSBetween =SSBetweendfBetween
=21.67
Fcrit(2,12) = 3.89, so we reject H0
![Page 26: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/26.jpg)
Statistics for the Social Sciences
Next time
• Basics of ANOVA• Why• Computations• ANOVA in SPSS• Post-hoc and planned comparisons• Assumptions • The structural model in ANOVA
![Page 27: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/27.jpg)
Statistics for the Social Sciences
Assumptions in ANOVA
• Populations follow a normal curve
• Populations have equal variances
![Page 28: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/28.jpg)
Statistics for the Social Sciences
Planned Comparisons
• Reject null hypothesis– Population means are not all the same
• Planned comparisons– Within-groups population variance estimate– Between-groups population variance estimate
• Use the two means of interest
– Figure F in usual way
![Page 29: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/29.jpg)
Statistics for the Social Sciences
1 factor ANOVA
Null hypothesis: H0: all the groups are equal
XBXA XC
XA = XB = XCAlternative hypotheses
HA: not all the groups are equal
XA ≠ XB ≠ XC XA ≠ XB = XC
XA = XB ≠ XC XA = XC ≠ XB
The ANOVA tests this one!!
![Page 30: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/30.jpg)
Statistics for the Social Sciences
1 factor ANOVA
Planned contrasts and post-hoc tests:
- Further tests used to rule out the different Alternative hypotheses
XA ≠ XB ≠ XC
XA ≠ XB = XC
XA = XB ≠ XC
XA = XC ≠ XB
Test 1: A ≠ B
Test 2: A ≠ C
Test 3: B = C
![Page 31: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/31.jpg)
Statistics for the Social Sciences
Planned Comparisons
• Simple comparisons
• Complex comparisons
• Bonferroni procedure– Use more stringent significance level for each
comparison
![Page 32: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/32.jpg)
Statistics for the Social Sciences
Controversies and Limitations
• Omnibus test versus planned comparisons– Conduct specific planned comparisons to
examine• Theoretical questions
• Practical questions
– Controversial approach
![Page 33: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/33.jpg)
Statistics for the Social Sciences
ANOVA in Research Articles
• F(3, 67) = 5.81, p < .01
• Means given in a table or in the text
• Follow-up analyses– Planned comparisons
• Using t tests
![Page 34: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/34.jpg)
Statistics for the Social Sciences
1 factor ANOVA
• Reporting your results– The observed difference– Kind of test – Computed F-ratio– Degrees of freedom for the test– The “p-value” of the test– Any post-hoc or planned comparison results
• “The mean score of Group A was 12, Group B was 25, and Group C was 27. A 1-way ANOVA was conducted and the results yielded a significant difference, F(2,25) = 5.67, p < 0.05. Post hoc tests revealed that the differences between groups A and B and A and C were statistically reliable (respectively t(1) = 5.67, p < 0.05 & t(1) = 6.02, p <0.05). Groups B and C did not differ significantly from one another”
![Page 35: 13.ANOVA1.06](https://reader031.vdocuments.net/reader031/viewer/2022032600/55cf92df550346f57b9a3abb/html5/thumbnails/35.jpg)
Statistics for the Social Sciences
The structural model and ANOVA
• The structural model is all about deviations
Score
(X)
Group mean
(M)
Grand mean
(GM)
Score’s deviation
from group mean
(X-M)
Group’s mean’s
deviation from
grand mean
(M-GM)
Score’s deviation from grand mean
(X-GM)
SStotal =SSWithin +SSBetweenX −GM( )∑ 2
= X−M( )∑ 2+ M −GM( )∑ 2