You are on page 1of 6

Q1.

A re-motivation team in psychiatric hospital conducted an experiment to compare four


methods for re-motivating patients. Patient in each group were randomly assigned to the four
methods. At the end of experimental period, the patient were evaluated by a team composed of
psychiatric, psychologist, a Nurse and a social worker, none of whom were aware of method to
which patients has been assigned. The team assigned each patient a composite score as a measure
of his level of motivation. The result were as follows.
Assumptions of ANOVA
Checking Assumption #1
Level of measurement:
1. Our dependent variable Score is quantitative or continuous
2. Our independent variables (factor) have more than two level which can be
either continuous or categorical.

Checking Assumption #2
Independence of observation:
Scatter Plot:

Data points are scattered throughout the plot, it means that observation of 16 patients with
composite score as measure of his level of motivation Y are independent.
Checking Assumption #3 Apply Shapiro Wilk test on all continuous variables.
Normality: Response variable should be normally distributed
H0: Data is normally distributed

H1: Data is not normally distributed

Tests of Normality
Method

Kolmogorov-Smirnova

Groups Statistic
Composite Score as measure
of his level of motivation

Shapiro-Wilk

df

Sig.

Statistic

df

Sig.

.198

.973

.857

.252

.903

.444

.198

.958

.764

.273

.947

.699

a. Lilliefors Significance Correction

Result:
The p-value is greater than > 0.05, this insignificant result by Shapiro Wilk test shows that
dependent variable Y is normally distributed.
Checking Assumption # 4 Apply levene Test
Homogeneity: variances of population group should be equal
H0: Population variances are equal
H1: Population variances are not equal
Test of Homogeneity of Variances
Composite Score as measure of his level of motivation
Levene Statistic

df1

df2

Sig.

.467

12

.710

Here p-value is greater than 0.05, this insignificant result by Levenes test indicates that
population variances are equal.

Hypothesis testing Steps for ANOVA.


1. Statement of hypothesis
H0: All means are equal.
H1: At least one mean is different.
2. Level of significance :

: 0.05

3. Test Statistics: f-test


4. Critical region: Reject H0: if p-value is = 0.05
5. Computation: by SPSS ver 16.

ANOVA
Composite Score as measure of his level of motivation
Sum of Squares

df

Mean Square

Sig.

Between Groups

623.688

207.896

5.162

.016

Within Groups

483.250

12

40.271

Total

1106.938

15

6. Conclusion:
Since p-value is less than (0.05), therefore we reject the H0: and conclude that at
least one group mean is different.

Here we can only find out the at least one mean is different but we dont know which group is
different. But luckily we can find out the difference with multiple comparison (Post Hoc) test.
Checking Assumption # 5 Apply Shapiro Wilk test on residuals
Residuals should be normally distributed.
H0: Residuals are normally distributed
H1: Residuals are not normally distributed

Tests of Normality
Kolmogorov-Smirnova

Shapiro-Wilk

Statistic

df

Sig.

Statistic

df

Sig.

.231

16

.022

.938

16

.330

Residual for Score

a. Lilliefors Significance Correction

Result:
The p-value is greater than

= 0.05, this insignificant result by Shapiro Wilk test indicates

That residuals are following normal distribution.

Post Hoc
Which Post Hoc test to be used:
1. Assumption of homogeneity is true, then
For equal sample size
Tukey Honesty significant difference ( HSD) test.
For Unequal sample size
Fishers least significant difference (LSD) test
2. If assumption of homogeneity is not true, then

For equal / unequal sample size


Games-Howell
Multiple Comparisons
Composite Score as measure of his level of motivation
Tukey HSD
(I)

(J)

95% Confidence Interval

Method Method Mean Difference


Groups Groups
A

(I-J)

Std. Error

Sig.

Lower Bound

Upper Bound

-10.000

4.487

.171

-23.32

3.32

-1.500

4.487

.986

-14.82

11.82

-15.250*

4.487

.024

-28.57

-1.93

10.000

4.487

.171

-3.32

23.32

8.500

4.487

.281

-4.82

21.82

-5.250

4.487

.656

-18.57

8.07

1.500

4.487

.986

-11.82

14.82

-8.500

4.487

.281

-21.82

4.82

-13.750*

4.487

.042

-27.07

-.43

15.250*

4.487

.024

1.93

28.57

5.250

4.487

.656

-8.07

18.57

13.750*

4.487

.042

.43

27.07

*. The mean difference is significant at the 0.05 level.

H0: Two population group means are not different.


H1: Two population group means are different.

Comparison

P-Value

Decision

Population A with B
Population A with C
Population A with D
Population B with C
Population B with D
Population C with D

0.171

Accept H0 :
Accept H0 :
Reject H0 :
Accept H0 :
Accept H0 :
Reject H0 :

Conclusion:

0.986
0.024
0.281
0.656
0.042

The above table is showing multiple comparison test result by Tukeys test. Here we can observe
that mean composite score as measure of his level of motivation are different in between Group
A & D and Group C & D.
There is no significant different in mean composite score of his motivational level in between
Group A & B, Group A & C , Group B & C and B & D.

Q2. The following data provides the measure mean upper central incisor width (mm) in
individuals
From our four different ethnic Groups.
Assumptions of ANOVA
Checking Assumption #1
Level of measurement:
3. Our dependent variable mean upper central incisor width (mm) is
quantitative or continuous variable
4. Our independent variables (factor) have more than two level which can be
either continuous or categorical.

Checking Assumption #2
Independence of observation:
Scatter Plot:

You might also like