Вы находитесь на странице: 1из 17

Bradley Butler

Factorial Anova
#2
Means

Case Processing Summary


Cases
Included Excluded Total
N Percent N Percent N Percent
Score * HoursStudied 40 100.0% 0 0.0% 40 100.0%
Score * Time 40 100.0% 0 0.0% 40 100.0%

Score * HoursStudied
Score
HoursStudied Mean N Std. Deviation
2h 72.5500 20 5.67056
4h 83.5000 20 11.32068
Total 78.0250 40 10.43291

Score * Time
Score
Time Mean N Std. Deviation
am 71.5000 20 5.34593
pm 84.5500 20 10.25709
Total 78.0250 40 10.43291

Univariate Analysis of Variance

Between-Subjects Factors
Value Label N
HoursStudied 1.00 2h 20
2.00 4h 20
Time 1.00 am 20
2.00 pm 20
Descriptive Statistics
Dependent Variable: Score
HoursStudied Time Mean Std. Deviation N
2h am 70.0000 5.94418 10
pm 75.1000 4.25441 10
Total 72.5500 5.67056 20
4h am 73.0000 4.47214 10
pm 94.0000 2.35702 10
Total 83.5000 11.32068 20
Total am 71.5000 5.34593 20
pm 84.5500 10.25709 20
Total 78.0250 10.43291 40

Levene's Test of Equality of Error


Variancesa
Dependent Variable: Score
F df1 df2 Sig.
1.993 3 36 .132
Tests the null hypothesis that the error variance of
the dependent variable is equal across groups.
a. Design: Intercept + HoursStudied + Time +
HoursStudied * Time

Tests of Between-Subjects Effects


Dependent Variable: Score
Type III Sum of Partial Eta
Source Squares df Mean Square F Sig. Squared
Corrected Model 3534.075a 3 1178.025 59.655 .000 .833
Intercept 243516.025 1 243516.025 12331.660 .000 .997
HoursStudied 1199.025 1 1199.025 60.719 .000 .628
Time 1703.025 1 1703.025 86.241 .000 .706
HoursStudied * Time 632.025 1 632.025 32.006 .000 .471
Error 710.900 36 19.747
Total 247761.000 40
Corrected Total 4244.975 39
a. R Squared = .833 (Adjusted R Squared = .819)

Estimated Marginal Means

1. HoursStudied

Estimates
Dependent Variable: Score
95% Confidence Interval
HoursStudied Mean Std. Error Lower Bound Upper Bound
2h 72.550 .994 70.535 74.565
4h 83.500 .994 81.485 85.515

Pairwise Comparisons
Dependent Variable: Score
95% Confidence Interval for
Mean Difference Differenceb
(I) HoursStudied (J) HoursStudied (I-J) Std. Error Sig.b Lower Bound Upper Bound
2h 4h -10.950* 1.405 .000 -13.800 -8.100
4h 2h 10.950* 1.405 .000 8.100 13.800
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
b. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).

Univariate Tests
Dependent Variable: Score
Partial Eta
Sum of Squares df Mean Square F Sig. Squared
Contrast 1199.025 1 1199.025 60.719 .000 .628
Error 710.900 36 19.747
The F tests the effect of HoursStudied. This test is based on the linearly independent pairwise
comparisons among the estimated marginal means.

2. Time

Estimates
Dependent Variable: Score
95% Confidence Interval
Time Mean Std. Error Lower Bound Upper Bound
am 71.500 .994 69.485 73.515
pm 84.550 .994 82.535 86.565

Pairwise Comparisons
Dependent Variable: Score
95% Confidence Interval for
Mean Difference Differenceb
(I) Time (J) Time (I-J) Std. Error Sig.b Lower Bound Upper Bound
am pm -13.050* 1.405 .000 -15.900 -10.200
pm am 13.050* 1.405 .000 10.200 15.900
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
b. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).

Univariate Tests
Dependent Variable: Score
Partial Eta
Sum of Squares df Mean Square F Sig. Squared
Contrast 1703.025 1 1703.025 86.241 .000 .706
Error 710.900 36 19.747
The F tests the effect of Time. This test is based on the linearly independent pairwise comparisons
among the estimated marginal means.

3. HoursStudied * Time
Dependent Variable: Score
95% Confidence Interval
HoursStudied Time Mean Std. Error Lower Bound Upper Bound
2h am 70.000 1.405 67.150 72.850
pm 75.100 1.405 72.250 77.950
4h am 73.000 1.405 70.150 75.850
pm 94.000 1.405 91.150 96.850

Profile Plots
Results
A two-way analysis of variance was conducted on the influence of two independent variables
(hours studied, time of study) on the test scores. The hours studied included two levels (2 hours,
4 hours) and the time of study had two levels (am, pm). All effects were statistically significant
at the .05 significance level. The main effect for hours studied yielded an F ratio of F(1, 36) =
60.719, p <.001, indicating a significant difference between 2 hours(M=72.55, SD=5.67) and 4
hours(M=83.50, SD=11.32). The main effect for time of study yielded an F ratio of F(1, 36)
=86.24, p <.001, indicating that the effect for time of study was significant, am (M=71.50,
SD=5.35) and pm (M=84.55, SD=10.26). The interaction effect was significant F(1, 36) = 32.01,
p <.001.

#3

Univariate Analysis of Variance

Between-Subjects Factors
Value Label N
Treatment 1.00 Drug 30
2.00 Therapy 30
Time 1.00 Baseline 20
2.00 30 days 20
3.00 180 days 20

Descriptive Statistics
Dependent Variable: Score
Treatment Time Mean Std. Deviation N
Drug Baseline 24.0000 2.16025 10
30 days 15.0000 3.05505 10
180 days 19.0000 2.16025 10
Total 19.3333 4.45153 30
Therapy Baseline 22.0000 1.76383 10
30 days 12.0000 2.26078 10
180 days 4.0000 1.76383 10
Total 12.6667 7.72070 30
Total Baseline 23.0000 2.17643 20
30 days 13.5000 3.03488 20
180 days 11.5000 7.93062 20
Total 16.0000 7.09500 60

Levene's Test of Equality of Error


Variancesa
Dependent Variable: Score
F df1 df2 Sig.
1.035 5 54 .407
Tests the null hypothesis that the error variance of
the dependent variable is equal across groups.
a. Design: Intercept + Treatment + Time +
Treatment * Time

Tests of Between-Subjects Effects


Dependent Variable: Score
Type III Sum of Partial Eta
Source Squares df Mean Square F Sig. Squared
Corrected Model 2700.000a 5 540.000 108.000 .000 .909
Intercept 15360.000 1 15360.000 3072.000 .000 .983
Treatment 666.667 1 666.667 133.333 .000 .712
Time 1510.000 2 755.000 151.000 .000 .848
Treatment * Time 523.333 2 261.667 52.333 .000 .660
Error 270.000 54 5.000
Total 18330.000 60
Corrected Total 2970.000 59
a. R Squared = .909 (Adjusted R Squared = .901)

Estimated Marginal Means

1. Treatment
Dependent Variable: Score
95% Confidence Interval
Treatment Mean Std. Error Lower Bound Upper Bound
Drug 19.333 .408 18.515 20.152
Therapy 12.667 .408 11.848 13.485

2. Time
Dependent Variable: Score
95% Confidence Interval
Time Mean Std. Error Lower Bound Upper Bound
Baseline 23.000 .500 21.998 24.002
30 days 13.500 .500 12.498 14.502
180 days 11.500 .500 10.498 12.502

3. Treatment * Time
Dependent Variable: Score
95% Confidence Interval
Treatment Time Mean Std. Error Lower Bound Upper Bound
Drug Baseline 24.000 .707 22.582 25.418
30 days 15.000 .707 13.582 16.418
180 days 19.000 .707 17.582 20.418
Therapy Baseline 22.000 .707 20.582 23.418
30 days 12.000 .707 10.582 13.418
180 days 4.000 .707 2.582 5.418

Profile Plots
Results
A two-way analysis of variance was conducted on the influence of two independent variables
(treatment type, length) on the level of stress. The treatment type included two levels (therapy,
drug) and the length had three levels (baseline, 30 days, and 180 days). All effects were
statistically significant at the .05 significance level. The main effect for treatment type yielded an
F ratio of F(1, 54) = 133.33, p <.001, indicating a significant difference between
therapy(M=12.68, SD=7.72) and drug(M=19.33, SD=4.45). The main effect for length yielded
an F ratio of F(2, 54) =151.00, p <.001, indicating that the effect for time of study was
significant, baseline(M=23.00, SD=2.18), 30 days(M=13.50, SD=3.03) and 180 days(M=11.50,
SD=7.93). The interaction effect was significant F(2, 54) = 52.33, p <.001.

#5

Univariate Analysis of Variance

Between-Subjects Factors
Value Label N
Typesoftricks 1.00 rolling over 20
2.00 sitting 20
Animal 1.00 dog 20
2.00 cat 20

Descriptive Statistics
Dependent Variable: numberoftimes
Typesoftricks Animal Mean Std. Deviation N
rolling over dog 10.0000 2.58199 10
cat 5.0000 1.49071 10
Total 7.5000 3.28473 20
sitting dog 12.0000 2.16025 10
cat 6.0000 1.15470 10
Total 9.0000 3.50939 20
Total dog 11.0000 2.53398 20
cat 5.5000 1.39548 20
Total 8.2500 3.43996 40

Levene's Test of Equality of Error


Variancesa
Dependent Variable: numberoftimes
F df1 df2 Sig.
.618 3 36 .608
Tests the null hypothesis that the error variance of
the dependent variable is equal across groups.
a. Design: Intercept + Typesoftricks + Animal +
Typesoftricks * Animal

Tests of Between-Subjects Effects


Dependent Variable: numberoftimes
Type III Sum of Partial Eta
Source Squares df Mean Square F Sig. Squared
Corrected Model 327.500a 3 109.167 29.328 .000 .710
Intercept 2722.500 1 2722.500 731.418 .000 .953
Typesoftricks 22.500 1 22.500 6.045 .019 .144
Animal 302.500 1 302.500 81.269 .000 .693
Typesoftricks * Animal 2.500 1 2.500 .672 .418 .018
Error 134.000 36 3.722
Total 3184.000 40
Corrected Total 461.500 39
a. R Squared = .710 (Adjusted R Squared = .685)

Estimated Marginal Means


1. Typesoftricks

Estimates
Dependent Variable: numberoftimes
95% Confidence Interval
Typesoftricks Mean Std. Error Lower Bound Upper Bound
rolling over 7.500 .431 6.625 8.375
sitting 9.000 .431 8.125 9.875

Pairwise Comparisons
Dependent Variable: numberoftimes
95% Confidence Interval for
Mean Difference Differenceb
(I) Typesoftricks (J) Typesoftricks (I-J) Std. Error Sig.b Lower Bound Upper Bound
rolling over sitting -1.500* .610 .019 -2.737 -.263
sitting rolling over 1.500* .610 .019 .263 2.737
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
b. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).

Univariate Tests
Dependent Variable: numberoftimes
Partial Eta
Sum of Squares df Mean Square F Sig. Squared
Contrast 22.500 1 22.500 6.045 .019 .144
Error 134.000 36 3.722
The F tests the effect of Typesoftricks. This test is based on the linearly independent pairwise
comparisons among the estimated marginal means.

2. Typesoftricks * Animal
Dependent Variable: numberoftimes
95% Confidence Interval
Typesoftricks Animal Mean Std. Error Lower Bound Upper Bound
rolling over dog 10.000 .610 8.763 11.237
cat 5.000 .610 3.763 6.237
sitting dog 12.000 .610 10.763 13.237
cat 6.000 .610 4.763 7.237

3. Animal

Estimates
Dependent Variable: numberoftimes
95% Confidence Interval
Animal Mean Std. Error Lower Bound Upper Bound
dog 11.000 .431 10.125 11.875
cat 5.500 .431 4.625 6.375

Pairwise Comparisons
Dependent Variable: numberoftimes
95% Confidence Interval for
Mean Difference Differenceb
(I) Animal (J) Animal (I-J) Std. Error Sig.b Lower Bound Upper Bound
dog cat 5.500* .610 .000 4.263 6.737
cat dog -5.500* .610 .000 -6.737 -4.263
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
b. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).

Univariate Tests
Dependent Variable: numberoftimes
Partial Eta
Sum of Squares df Mean Square F Sig. Squared
Contrast 302.500 1 302.500 81.269 .000 .693
Error 134.000 36 3.722
The F tests the effect of Animal. This test is based on the linearly independent pairwise comparisons
among the estimated marginal means.

Profile Plots
Results
A two-way analysis of variance was conducted on the influence of two independent variables
(animal type, types of tricks) on the number of tricks completed. The animal type included two
levels (dog, cat) and the type of trick had two levels (rolling over, sitting). All effects were
statistically significant at the .05 significance level besides the interaction between the
independent variables. The main effect for animal type yielded an F ratio of F(1, 36) = 81.27, p
<.001, indicating a significant difference between dog (M=11.00, SD=2.53) and cat(M=5.50,
SD=1.40). The main effect for type of trick yielded an F ratio of F(1, 36) =6.05, p <.001,
indicating that the effect for type of trick was significant, rolling over(M=7.50, SD=3.28), and
sitting (M=9.00, SD=3.51). The interaction effect was not significant F(1, 36) = .672, p >.001.

Вам также может понравиться