2nd ASSIGNMENT

ANALYSIS OF GROUP DIFFERENCES

EDU6950 SEM1 2009

Lecturer:

ASSOCIATE PROFESOR
rohaniat@[Link]

By
ELENCHOTHY A/P DAVRAJOO
Institute for Mathematical Research
University Putra Malaysia, 43400 UPM, Serdang, Selangor.
echothyat@[Link]
GS 21907
01.09.2009

1
QUESTION NO 1
To test the significance of effectiveness of negotiation for meaning towards the improvement of
students’ classroom interaction skill.
a. Hypotheses
H o: There are no significant differences in the students’ classroom interaction skill before
and after the treatment (negotiation for meaning)
H A : There are significant differences in the students’ classroom interaction skill before
and after the treatment (negotiation for meaning)
¿
H o: μ1=¿ μ2=¿ μ3
H A : μ1 ≠ μ2 ≠ μ 3

b. Decision and justification

By using One way Repeated Measures in ANOVA

Descriptive Statistics
Mean Std. Deviation N
Pre 7.9000 2.02485 10
Post1 8.6000 2.06559 10
Post2 11.4000 1.77639 10
Post3 15.0000 1.05409 10

Multivariate Tests(b)
Partial
Hypothesis Eta
Effect Value F df Error df Sig. Squared
Time Pillai's Trace .970 74.560(a) 3.000 7.000 .000 .970
Wilks' Lambda .030 74.560(a) 3.000 7.000 .000 .970
Hotelling's Trace 31.954 74.560(a) 3.000 7.000 .000 .970
Roy's Largest
31.954 74.560(a) 3.000 7.000 .000 .970
Root
a Exact statistic
b Design: Intercept
Within Subjects Design: TIME

Mauchly's Test of Sphericity(b)

Measure: MEASURE_1
Epsilon(a)
Approx. Chi- Greenhouse Lower
Within Subjects Effect Mauchly's W Square df Sig. -Geisser Huynh Feldt Bound
TIME .567 4.377 5 .499 .751 1.000 .333
Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variables is proportional
to an identity matrix.
a May be used to adjust the degrees of freedom for the averaged tests of significance. Corrected tests are displayed in the Tests
of Within-Subjects Effects table.
b Design: Intercept
Within Subjects Design: TIME

2
 Pairwise Comparisons

Measure: MEASURE_1
Mean
Difference 95% Confidence Interval for
(I) TIME (J) TIME (I-J) Std. Error Sig.(a) Difference(a)

Upper Bound Lower Bound

1 2 -.700 .260 .149 -1.576 .176
3 -3.500(*) .428 .000 -4.940 -2.060
4 -7.100(*) .433 .000 -8.558 -5.642
2 1 .700 .260 .149 -.176 1.576
3 -2.800(*) .359 .000 -4.008 -1.592
4 -6.400(*) .427 .000 -7.836 -4.964
3 1 3.500(*) .428 .000 2.060 4.940
2 2.800(*) .359 .000 1.592 4.008
4 -3.600(*) .340 .000 -4.744 -2.456
4 1 7.100(*) .433 .000 5.642 8.558
2 6.400(*) .427 .000 4.964 7.836
3 3.600(*) .340 .000 2.456 4.744
Based on estimated marginal means
* The mean difference is significant at the .05 level.
a Adjustment for multiple comparisons: Bonferroni.

c. Calculation and interpret an appropriate effect size

One Way repeated measures ANOVA was conducted to compare scores on interaction
skill in coping with Statistics test at PRE Test (prior to the intervention), POST 1, POST
2, POST 3 (following the intervention), in certain time of follow up. The means and the
standard deviation are presented in Descriptive Statistics Table. The mean of score at
PRE was 7.9 (SD = 2.02) and the mean of POST 1 was 8.6 (SD = 2.061) followed by
11.4 (SD = 1.78) and 15 (SD = 1.05). Therefore we can conclude that, there was
improvement in score of test from PRE to POST 3. There was a significant effect for
time, Wilk’s Lambda = 0.030; F (3, 7) = 74.56, p < 0.005, multivariate partial eta squared
= 0.970. Using the commonly used guidelines proposed by Cohen (1988, pp 284-7)
(0.01 = small, 0.06 = moderate, 0.14 = large effect). This result suggests a very large
effect size. The information provided by Comparisons Table which compare each pair of
time points and indicate the differences are significant (all Sig values are less than 0.005).

3
QUESTION NO 2
To test the significance of effectiveness of a new teaching method to improve sudents’knowledge
on statistics.
a. Hypotheses
H o: There are no significant differences in the students’ knowledge on statistics before
and after the treatment (new teaching method)
H A : There are significant differences in the students’ knowledge on statistics before and
after the treatment (new teaching method)
¿
H o: μ Pre =¿ μ Post
H A : μ Pra ≠ μ Post

b. Decision and justification

Group Statistics

Std. Error
Group N Mean Std. Deviation Mean
Pre2 1 6 9.8333 1.47196 .60093
2 8 11.0000 1.85164 .65465
Post 1 6 13.1667 1.72240 .70317
2 8 12.2500 2.25198 .79620

Independent Samples Test

Levene's
Test for
Equality of
Variances t-test for Equality of Means
95% Confidence
Interval of the
Sig. Mean Std. Error Difference
(2- Differen Differenc
F Sig. t df tailed) ce e Lower Upper
Pre2 Equal
-
variances .527 .482 -1.268 12 .229 .92013 -3.17146 .83813
1.16667
assumed
Equal
-
variances not -1.313 11.919 .214 .88864 -3.10431 .77097
1.16667
assumed
Post Equal
variances .792 .391 .829 12 .423 .91667 1.10606 -1.49324 3.32658
assumed
Equal
variances not .863 11.977 .405 .91667 1.06225 -1.39826 3.23160
assumed

4
 Between-Subjects Factors

N
Group 1 6
2 8

Descriptive Statistics

Dependent Variable: Post

Group Mean Std. Deviation N
1 13.1667 1.72240 6
2 12.2500 2.25198 8
Total 12.6429 2.02322 14

Levene's Test of Equality of Error Variances(a)

Dependent Variable: Post

F df1 df2 Sig.
2.478 1 12 .141
Tests the null hypothesis that the error variance of the dependent
variable is equal across groups.
a Design: Intercept+Pre2+Group

Tests of Between-Subjects Effects

Dependent Variable: Post

Type III Sum Partial Eta
Source of Squares df Mean Square F Sig. Squared
Corrected Model 44.700(a) 2 22.350 28.875 .000 .840
Intercept .526 1 .526 .679 .427 .058
Pre2 41.819 1 41.819 54.027 .000 .831
Group 14.567 1 14.567 18.820 .001 .631
Error 8.514 11 .774
Total 2291.000 14
Corrected Total 53.214 13
a R Squared = .840 (Adjusted R Squared = .811)

Group
Dependent Variable: Post
95% Confidence Interval
Group Mean Std. Error Upper Lower
1 13.897(a) .373 13.077 14.717
2 11.702(a) .320 10.998 12.406
a Covariates appearing in the model are evaluated at the following values: Pre2 = 10.5000.

5
 c. Calculation and interpret an appropriate effect size
A One way between-groups analysis of covariances technique was conducted to evaluate
the impact of intervention (or experimental manipulation) to increase the the
participants’s score o f statistics while controling the pretest score. The independent
variable was the the type of intervention and the dependent variable consisted of scores
on the post test which was administred after the intervention completed. Participant’s
scores on the pre intervention administration of the improvement of Statistics Test were
used as the covariate in this analysis.
Preliminary checks were conducted to ensure that there was no violotion of assumptions
as normanilty, linearity, homogeneity of variances and influence of covariate. Levene's
Test for Equality of Variances in the Independent Sample test table shows significant
value (0.482) is larger than 0.05. Therefore Equal variances assumed. The table of Test
Between Subject Effects shows the significant value of group different is less than 0.05,
which indicated that the improvement on treatment group (with the mean = 13.16
compared to mean scoresof control group = 12.25) of statistics score for subjects in the
new teaching method. After adjusting the for pre-intervention scors, there was a
significant difference between the two groups on post test scores F (1, 11) = 18.20, p =
0.001, . Partial Eta Squared = 0.631. There was a strong relationship between pretest
and post as indicated by partial eta squared value of 0.84 (big effect size according to
Cohen’s 1988, guideliness).

QUESTION NO 3
Using the SURVEY. SAV file, select a set of variables suitable to conduct a one-way
MANOVA.
a. Hypotheses
H o: There are no significant differences between males and females on total life
satisfaction, the perceved stress and self esteem
H A : There are significant differences between males and females on total life satisfaction,
the perceved stress and self esteem
¿
H o: μmale=¿ μfemale
H A : μmale ≠ μ female

b. Decision and justification

Between-Subjects Factors

Value Label N
sex 1 MALES 183
2 FEMALES 248

Descriptive Statistics
sex Mean Std. Deviation N

6
 Total life satisfaction MALES 21.64 6.548 183
FEMALES 22.85 6.934 248
Total 22.34 6.792 431
Total perceived stress MALES 25.80 5.428 183
FEMALES 27.44 6.074 248
Total 26.74 5.858 431
Total Self esteem MALES 33.99 4.905 183
FEMALES 33.24 5.647 248
Total 33.56 5.351 431

Box's Test of Equality of Covariance Matrices(a)

Box's M 8.520
F 1.409
df1 6
df2 1061856.10
0
Sig. .207
Tests the null hypothesis that the observed covariance matrices of the dependent variables are equal across groups.
a Design: Intercept+sex

Multivariate Tests(b)

Partial Eta
Effect Value F Hypothesis df Error df Sig. Squared
Intercept Pillai's Trace .993 21167.272(a) 3.000 427.000 .000 .993
Wilks' Lambda .007 21167.272(a) 3.000 427.000 .000 .993
Hotelling's Trace 148.716 21167.272(a) 3.000 427.000 .000 .993
Roy's Largest Root 148.716 21167.272(a) 3.000 427.000 .000 .993
sex Pillai's Trace .053 8.027(a) 3.000 427.000 .000 .053
Wilks' Lambda .947 8.027(a) 3.000 427.000 .000 .053
Hotelling's Trace .056 8.027(a) 3.000 427.000 .000 .053
Roy's Largest Root .056 8.027(a) 3.000 427.000 .000 .053
a Exact statistic
b Design: Intercept+sex

Levene's Test of Equality of Error Variances(a)

F df1 df2 Sig.

Total life satisfaction .678 1 429 .411
Total perceived stress 1.856 1 429 .174
Total Self esteem 2.977 1 429 .085
Tests the null hypothesis that the error variance of the dependent variable is equal across groups.
a Design: Intercept+sex

Tests of Between-Subjects Effects

7
 Type III Sum Partial Eta
Source Dependent Variable of Squares df Mean Square F Sig. Squared
Corrected Model Total life
154.542(a) 1 154.542 3.369 .067 .008
satisfaction
Total perceived
282.409(b) 1 282.409 8.370 .004 .019
stress
Total Self esteem 58.779(c) 1 58.779 2.058 .152 .005
Intercept Total life
208426.653 1 208426.653 4543.522 .000 .914
satisfaction
Total perceived
298395.485 1 298395.485 8843.952 .000 .954
stress
Total Self esteem 475953.721 1 475953.721 16660.665 .000 .975
sex Total life
154.542 1 154.542 3.369 .067 .008
satisfaction
Total perceived
282.409 1 282.409 8.370 .004 .019
stress
Total Self esteem 58.779 1 58.779 2.058 .152 .005
Error Total life
19679.677 429 45.873
satisfaction
Total perceived
14474.487 429 33.740
stress
Total Self esteem 12255.462 429 28.568
Total Total life
234867.000 431
satisfaction
Total perceived
322937.000 431
stress
Total Self esteem 497714.000 431
Corrected Total Total life
19834.218 430
satisfaction
Total perceived
14756.896 430
stress
Total Self esteem 12314.241 430
a R Squared = .008 (Adjusted R Squared = .005)
b R Squared = .019 (Adjusted R Squared = .017)
c R Squared = .005 (Adjusted R Squared = .002)
sex
95% Confidence Interval
Dependent Variable sex Mean Std. Error Lower Bound Upper Bound
Total life satisfaction MALES 21.639 .501 20.655 22.623
FEMALES 22.851 .430 22.005 23.696
Total perceived stress MALES 25.798 .429 24.954 26.642
FEMALES 27.435 .369 26.711 28.160
Total Self esteem MALES 33.989 .395 33.212 34.766
FEMALES 33.242 .339 32.575 33.909

c. Calculation and interpretation on effect size

8
 Three dependent variables were used; Total Life Satisfaction, Total Perceived Stress and
Total Self esteem. The independent variable was gender. There was a statistically
significant difference between males and females on the combined dependent variables;
F (3, 427) = 8.02, p = 0.000 (less than 0.005); Wilk’ Lamda = 0.993, partial eta squared =
0.53. When the results for the dependent variable were considered separately, the only
difference to reach Statistical significance, using Bonferroni adjusted alpha level 0f 0.17,
was Total Perceived Stress F(1, 429) = 8.34, p = 0.004 (less than 0.005); Wilk’ Lamda =
0.993, partial eta squared = 0.02.
An inspection of the mean scores indicated that females reported slightly higher levels of
Total perceived Stress (mean = 27.42, SD = 6.08) than males (mean = 25.8, SD = 5.4).

QUESTION NO 4
Using [Link] file, select a set of variables suitable to conduct a MANCOVA.
Task was carried out to test are there any difference in mean of Systolic BP Cold Pressor,
Diastolic BP Cold Pressor and Heart Rate Cold Pressor between the races controlled their
age.

a. Hypotheses
H o: There are no significant differences in mean of Systolic BP Cold Pressor, Diastolic
BP Cold Pressor and Heart Rate Cold Pressor between races when controlled for
age
H A : There are significant differences in in mean of Systolic BP Cold Pressor, Diastolic
BP Cold Pressor and Heart Rate Cold Pressor between races when controlled for
age
¿
H o: μi=¿ μi’
H A : μi ≠ μi ’

H o: μ j=¿ μ j’
H A: μ j ≠ μ j’

H o: μi j =¿ μi j ’
H A : μi j ≠ μ i j’

b. Decision and justification

9
OUTPUT OF MANCOVA ANALYSIS
Between-Subjects Factors

Value Label N
Race 2 Asian or
Pacific 2
Islander
3 Black, Not of
Hispanic 5
Origin
4 Hispanic 4
5 White, not of
Hispanic 165
origin
8 1

Descriptive Statistics

Race Mean Std. Deviation N

systolic bp cold pressor Asian or Pacific Islander 116.0000 .00000 2
Black, Not of Hispanic
Origin 131.4000 18.98157 5
Hispanic 129.7500 9.74252 4
White, not of Hispanic
origin 131.8970 15.55503 165
8 132.0000 . 1
Total 131.6554 15.43350 177
diastolic bp cold pressor Asian or Pacific Islander 89.0000 5.65685 2
Black, Not of Hispanic
Origin 69.4000 26.55748 5
Hispanic 76.0000 9.38083 4
White, not of Hispanic
origin 78.0061 14.46168 165
8 65.0000 . 1
Total 77.7684 14.73314 177
heart rate cold pressor Asian or Pacific Islander 92.0000 7.07107 2
Black, Not of Hispanic
Origin 79.0000 21.14237 5
Hispanic 79.5000 15.35144 4
White, not of Hispanic
origin 84.0485 15.49028 165
8 88.0000 . 1
Total 83.9153 15.49298 177

Box's Test of Equality of Covariance Matrices(a)

10
 Box's M 16.523
F .889
df1 12
df2 312.818
Sig. .559
Tests the null hypothesis that the observed covariance matrices of the dependent variables are equal across groups.
a Design: Intercept+age+race

Multivariate Tests(d)
Partial
Hypothesis Eta Noncent. Observed
Effect Value F df Error df Sig. Squared Parameter Power(a)
Intercept Pillai's Trace .131 8.487(b) 3.000 169.000 .000 .131 25.462 .993
Wilks' Lambda .869 8.487(b) 3.000 169.000 .000 .131 25.462 .993
Hotelling's
.151 8.487(b) 3.000 169.000 .000 .131 25.462 .993
Trace
Roy's Largest
.151 8.487(b) 3.000 169.000 .000 .131 25.462 .993
Root
age Pillai's Trace .072 4.347(b) 3.000 169.000 .006 .072 13.041 .863
Wilks' Lambda .928 4.347(b) 3.000 169.000 .006 .072 13.041 .863
Hotelling's
.077 4.347(b) 3.000 169.000 .006 .072 13.041 .863
Trace
Roy's Largest
.077 4.347(b) 3.000 169.000 .006 .072 13.041 .863
Root
race Pillai's Trace .043 .625 12.000 513.000 .822 .014 7.498 .367
Wilks' Lambda .957 .621 12.000 447.423 .825 .014 6.562 .318
Hotelling's
.044 .617 12.000 503.000 .829 .015 7.402 .362
Trace
Roy's Largest
.028 1.204(c) 4.000 171.000 .311 .027 4.816 .372
Root
a Computed using alpha = .05
b Exact statistic
c The statistic is an upper bound on F that yields a lower bound on the significance level.
d Design: Intercept+age+race

Levene's Test of Equality of Error Variances(a)

F df1 df2 Sig.

systolic bp cold pressor 1.174 4 172 .324
diastolic bp cold pressor 2.854 4 172 .025
heart rate cold pressor .742 4 172 .565
Tests the null hypothesis that the error variance of the dependent variable is equal across groups.
a Design: Intercept+age+race

Tests of Between-Subjects Effects

11
 Observ
Partial ed
Type III Sum Mean Eta Noncent. Power
Source Dependent Variable of Squares df Square F Sig. Squared Parameter (a)
Corrected systolic bp cold pressor
2238.777(b) 5 447.755 1.929 .092 .053 9.647 .643
Model
diastolic bp cold pressor 2090.228(c) 5 418.046 1.979 .084 .055 9.897 .656
heart rate cold pressor 396.986(d) 5 79.397 .324 .898 .009 1.622 .131
Intercept systolic bp cold pressor 2517.70
2517.703 1 10.849 .001 .060 10.849 .906
3
diastolic bp cold pressor 328.367 1 328.367 1.555 .214 .009 1.555 .236
heart rate cold pressor 2821.54
2821.544 1 11.529 .001 .063 11.529 .922
4
age systolic bp cold pressor 1723.99
1723.998 1 7.429 .007 .042 7.429 .773
8
diastolic bp cold pressor 1302.92
1302.920 1 6.169 .014 .035 6.169 .695
0
heart rate cold pressor 47.869 1 47.869 .196 .659 .001 .196 .072
race systolic bp cold pressor 752.124 4 188.031 .810 .520 .019 3.241 .256
diastolic bp cold pressor 704.877 4 176.219 .834 .505 .019 3.338 .263
heart rate cold pressor 328.055 4 82.014 .335 .854 .008 1.340 .125
Error systolic bp cold pressor 39683.200 171 232.065
diastolic bp cold pressor 36113.274 171 211.189
heart rate cold pressor 41848.743 171 244.729
Total systolic bp cold pressor 3109887.000 177
diastolic bp cold pressor 1108685.000 177
heart rate cold pressor 1288639.000 177
Corrected systolic bp cold pressor
41921.977 176
Total
diastolic bp cold pressor 38203.503 176
heart rate cold pressor 42245.729 176
a Computed using alpha = .05
b R Squared = .053 (Adjusted R Squared = .026)
c R Squared = .055 (Adjusted R Squared = .027)
d R Squared = .009 (Adjusted R Squared = -.020)
Estimates
95% Confidence Interval

Dependent Variable Race Mean Std. Error Lower Bound Upper Bound
systolic bp cold pressor Asian or Pacific Islander 113.043(a) 10.826 91.672 134.413
Black, Not of Hispanic
Origin 132.583(a) 6.827 119.108 146.058
Hispanic 128.271(a) 7.636 113.198 143.345
White, not of Hispanic
origin 131.951(a) 1.186 129.609 134.292
8 129.043(a) 15.272 98.896 159.189
diastolic bp cold pressor Asian or Pacific Islander 86.429(a) 10.328 66.043 106.816
Black, Not of Hispanic
Origin 70.428(a) 6.512 57.574 83.283
Hispanic 74.715(a) 7.285 60.335 89.094
White, not of Hispanic
origin 78.053(a) 1.131 75.819 80.286

12
 8 62.429(a) 14.569 33.671 91.188
heart rate cold pressor Asian or Pacific Islander 91.507(a) 11.118 69.561 113.453
Black, Not of Hispanic
Origin 79.197(a) 7.010 65.359 93.035
Hispanic 79.254(a) 7.842 63.775 94.733
White, not of Hispanic
origin 84.057(a) 1.218 81.653 86.462
8 87.507(a) 15.683 56.549 118.465
a Covariates appearing in the model are evaluated at the following values: subject age = 19.00.

Pairwise Comparisons
95% Confidence Interval
Mean for Difference(a)
Dependent Difference Lower Upper
Variable (I) Race (J) Race (I-J) Std. Error Sig.(a) Bound Bound
systolic bp Asian or Black, Not of Hispanic
-19.540 12.836 1.000 -56.043 16.963
cold pressor Pacific Islander Origin
Hispanic
-15.229 13.204 1.000 -52.779 22.322
White, not of Hispanic
-18.908 10.893 .844 -49.887 12.071
origin
8
-16.000 18.657 1.000 -69.060 37.060
Black, Not of Asian or Pacific Islander
19.540 12.836 1.000 -16.963 56.043
Hispanic Origin
Hispanic 4.311 10.266 1.000 -24.883 33.506
White, not of Hispanic
.632 6.928 1.000 -19.069 20.334
origin
8 3.540 16.757 1.000 -44.115 51.195
Hispanic Asian or Pacific Islander
15.229 13.204 1.000 -22.322 52.779
Black, Not of Hispanic
-4.311 10.266 1.000 -33.506 24.883
Origin
White, not of Hispanic
-3.679 7.729 1.000 -25.660 18.302
origin
8
-.771 17.040 1.000 -49.233 47.690
White, not of Asian or Pacific Islander
18.908 10.893 .844 -12.071 49.887
Hispanic origin
Black, Not of Hispanic
Origin -.632 6.928 1.000 -20.334 19.069
Hispanic 3.679 7.729 1.000 -18.302 25.660
8 2.908 15.320 1.000 -40.660 46.476
8 Asian or Pacific Islander
16.000 18.657 1.000 -37.060 69.060
Black, Not of Hispanic
-3.540 16.757 1.000 -51.195 44.115
Origin
Hispanic .771 17.040 1.000 -47.690 49.233
White, not of Hispanic
origin -2.908 15.320 1.000 -46.476 40.660

13
diastolic bp Asian or Black, Not of Hispanic
16.001 12.245 1.000 -18.822 50.824
cold pressor Pacific Islander Origin
Hispanic
11.715 12.596 1.000 -24.107 47.537
White, not of Hispanic
8.376 10.392 1.000 -21.176 37.929
origin
8
24.000 17.798 1.000 -26.617 74.617
Black, Not of Asian or Pacific Islander
-16.001 12.245 1.000 -50.824 18.822
Hispanic Origin
Hispanic -4.286 9.793 1.000 -32.137 23.564
White, not of Hispanic
origin -7.624 6.609 1.000 -26.419 11.170
8 7.999 15.985 1.000 -37.461 53.460
Hispanic Asian or Pacific Islander
-11.715 12.596 1.000 -47.537 24.107
Black, Not of Hispanic
4.286 9.793 1.000 -23.564 32.137
Origin
White, not of Hispanic
-3.338 7.373 1.000 -24.307 17.631
origin
8 12.285 16.256 1.000 -33.945 58.516
White, not of Asian or Pacific Islander
-8.376 10.392 1.000 -37.929 21.176
Hispanic origin
Black, Not of Hispanic
Origin 7.624 6.609 1.000 -11.170 26.419
Hispanic 3.338 7.373 1.000 -17.631 24.307
8 15.624 14.614 1.000 -25.938 57.186
8 Asian or Pacific Islander
-24.000 17.798 1.000 -74.617 26.617
Black, Not of Hispanic
-7.999 15.985 1.000 -53.460 37.461
Origin
Hispanic
-12.285 16.256 1.000 -58.516 33.945
White, not of Hispanic
-15.624 14.614 1.000 -57.186 25.938
origin
heart rate Asian or Black, Not of Hispanic
12.310 13.181 1.000 -25.176 49.796
cold pressor Pacific Islander Origin
Hispanic
12.254 13.559 1.000 -26.308 50.815
White, not of Hispanic
7.450 11.186 1.000 -24.363 39.263
origin
8 4.000 19.160 1.000 -50.489 58.489
Black, Not of Asian or Pacific Islander
Hispanic Origin -12.310 13.181 1.000 -49.796 25.176

Hispanic -.057 10.542 1.000 -30.037 29.924

White, not of Hispanic
origin -4.860 7.114 1.000 -25.092 15.372
8 -8.310 17.208 1.000 -57.248 40.627
Hispanic Asian or Pacific Islander
-12.254 13.559 1.000 -50.815 26.308
Black, Not of Hispanic
.057 10.542 1.000 -29.924 30.037
Origin
White, not of Hispanic
-4.804 7.937 1.000 -27.377 17.769
origin

14
 8
-8.254 17.499 1.000 -58.020 41.513
White, not of Asian or Pacific Islander
-7.450 11.186 1.000 -39.263 24.363
Hispanic origin
Black, Not of Hispanic
Origin 4.860 7.114 1.000 -15.372 25.092
Hispanic 4.804 7.937 1.000 -17.769 27.377
8 -3.450 15.732 1.000 -48.191 41.291
8 Asian or Pacific Islander -4.000 19.160 1.000 -58.489 50.489
Black, Not of Hispanic
8.310 17.208 1.000 -40.627 57.248
Origin
Hispanic
8.254 17.499 1.000 -41.513 58.020
White, not of Hispanic
3.450 15.732 1.000 -41.291 48.191
origin
Based on estimated marginal means
a Adjustment for multiple comparisons: Bonferroni.

Estimated Marginal Means of systolic bp cold pressor

135.00
Estim ated Marginal M eans

130.00

125.00

120.00

115.00

110.00

Asian or Pacific Black, Not of Hispanic White, not of 8

Islander Hispanic Origin Hispanic origin
Race

15
 Estimated Marginal Means of diastolic bp cold pressor

90.00
Estim ated M arginal M eans

85.00

80.00

75.00

70.00

65.00

60.00

Asian or Pacific Black, Not of Hispanic White, not of 8

Islander Hispanic Origin Hispanic origin
Race

Estimated Marginal Means of heart rate cold pressor

90.00
Estimated Marginal Means

87.00

84.00

81.00

78.00

Asian or Pacific Black, Not of Hispanic White, not of 8

Islander Hispanic Origin Hispanic origin
Race

16
 c. Calculation and interpretation on effect size
Based on Multivariate Test table Wilks’ Lamda is 0.621, p value is 0.825 (> 0.05), the
effect size is 0.014 (< 0.14) and the power is 31.8% (< 80%). Hence there is no
difference between group (race) centroids when controlled for age.
Based on the Test of Between-Subjects Effect table, for race
Systolic bp Cold Pressor: p-value is 0.52 (> 0.05), the effect size is 0.19 (< 0.14) and the
power is 25.6% (< 80%);
Diastolic bp Cold Pressor: p-value is 0.51 (> 0.05), the effect size is 0.19 (< 0.14) and the
power is 26.3% (< 80%) and
Heart Rate Cold Pressor: p-value is 0.85 (> 0.05), the effect size is 0.008 (< 0.14) and the
power is 12.5% (< 80%). Therefore, race has no impact on Systolic bp Cold Pressor,
Diastolic bp Cold Pressor and Heart Rate Cold Pressor.
Again based on the Pairwise comparison tests, race has no impact on Systolic bp Cold
Pressor, Diastolic bp Cold Pressor and Heart Rate Cold Pressor.

QUESTION NO 5
Using the SURVEY. SAV file, select a set of valubles suitable to conduct a Factorial- ANOVA
(2 –factor).
a. Write approperiate null and alternative hypotheses
H o: There is significant effects on perceive stress level between gender and age.
H A : There is no significant effects on perceive stress level between gender and age
¿
H o: μ B 1=¿ μ B 2
H A : μ B 1 ≠ μB 2

b. Make your decision and justify your answer

Levene's Test of Equality of Error Variances(a)

Dependent Variable: Total perceived stress

F df1 df2 Sig.
1.026 9 423 .418
Tests the null hypothesis that the error variance of the dependent variable is equal across groups.
a Design: Intercept+sex+agegp5+sex * agegp5

Non significance indicates homogeneity of variance

17
 Tests of Between-Subjects Effects
Dependent Variable: Total perceived stress
Type III Sum Partial Eta
Source of Squares df Mean Square F Sig. Squared
Corrected Model 839.252(a) 9 93.250 2.831 .003 .057
Intercept 295968.489 1 295968.489 8985.743 .000 .955
sex 277.994 1 277.994 8.440 .004 .020
agegp5 503.367 4 125.842 3.821 .005 .035
sex * agegp5 64.874 4 16.219 .492 .741 .005
Error 13932.591 423 32.938
Total 324089.000 433
Corrected Total 14771.843 432
a R Squared = .057 (Adjusted R Squared = .037)
interaction between factors.

F ratios and p-values show no significant interaction between factors. Age groups and gender
are significant (< 0.05). The column of partial Eta Squared shows very small effect size.

Pairwise Comparisons
Dependent Variable: Total perceived stress
95% Confidence Interval for
Mean Difference(a)
Difference Upper Bound
(I) age 5 groups (J) age 5 groups (I-J) Std. Error Sig.(a) Lower Bound
18 - 24 25 - 32 2.933(*) .867 .008 .486 5.380
33 - 40 1.862 .876 .342 -.610 4.334
41 - 49 2.248 .858 .091 -.173 4.669
50+ 2.900(*) .893 .013 .380 5.419
25 - 32 18 - 24 -2.933(*) .867 .008 -5.380 -.486
33 - 40 -1.071 .890 1.000 -3.582 1.440
41 - 49 -.684 .872 1.000 -3.146 1.777
50+ -.033 .907 1.000 -2.591 2.525
33 - 40 18 - 24 -1.862 .876 .342 -4.334 .610
25 - 32 1.071 .890 1.000 -1.440 3.582
41 - 49 .387 .881 1.000 -2.100 2.873
50+ 1.038 .915 1.000 -1.544 3.621
41 - 49 18 - 24 -2.248 .858 .091 -4.669 .173
25 - 32 .684 .872 1.000 -1.777 3.146
33 - 40 -.387 .881 1.000 -2.873 2.100
50+ .652 .898 1.000 -1.882 3.185
50+ 18 - 24 -2.900(*) .893 .013 -5.419 -.380
25 - 32 .033 .907 1.000 -2.525 2.591
33 - 40 -1.038 .915 1.000 -3.621 1.544
41 - 49 -.652 .898 1.000 -3.185 1.882
Based on estimated marginal means
* The mean difference is significant at the .05 level.
a Adjustment for multiple comparisons: Bonferroni.

18
The Asterisk indicates which groups are significantly different.

Estimated Marginal Means of Total perceived stress

29 age 5 groups
18 - 24
25 - 32
33 - 40
41 - 49
28
50+
Estimated Marginal Means

27

26

25

24

MALES FEMALES
sex

c. Calculation and interpret an approperiate effect size

The two way analysis of variance was conducted to investigate Total Perceived Stress Level
in gender and age group respondents. ANOVA results, presented in Table 1, showed a
significant main effect for gender,( F (1, 423) = 277.99 p< 0.05, partial n2 = 0.020) and age
group, (F (4, 423) = 125.84, 4, p< 0.05, partial n2 = 0.035) . However, interaction between
factors are not significant, F (4, 423) = 0.492, 4, p= 0.741, partial n2 = 0.005). Calculated
effect size for each factor indicates a small proportion of perceived stress level variance is
accounted for each factor. The Bonferroni post hoc test was conducted to determine the age
group that significantly different. Result revealed that the age groups of 18-24 were
significantly different in Total Perceived Stress Level from those 25-32 years age and the
groups of age above 50+.

19