Diapositive 1

1
Research methodology (459500)Lecture 12Dr
Jihad ABDALLAH

2
Pretest-posttest two group design(ANCOVA
Analysis)
3
Comparing Several Group means (One-way ANOVA)
4
Example1 (Employee data)
Dependent variable current salary Group
variable job category
5
Example (Alawni et al., 2007)
6
Factorial Analysis
Linear model
7
Employee data example
Factor 1 gender (two levels) Factor 2 job
category
8
Nonequivalent Groups Analysis
Pretest values should be adjusted
Then we proceed as in the ANCOVA
9
The linear model
10
Kruskal-Wallis test

• Non-parametric test.
• Used when the distribution of the dependent
  variable is not normal.
• Compares the mean ranks of groups.
• Normality can be tested using Kolmogorov-Smirnov
  test (K-S test) ? if P value lt 0.05 this means
  that the data is not normal.

11
(No Transcript)
12
Kruskal-Wallis test
13
Cross classification (contingency tables)

• Test of association between two categorical
  variables.
• Example gender and job category
• H0 there is no association between gender and
  job category
• H1 there is association between gender and job
  category (job category depend on gender).

14
(No Transcript)