1. CAUSAL-COMPARATIVE
RESEARCH
Prepared for:
Dr.Johan @ Eddy Luaran
Prepared by:
Nur Hazwani Mohd Nor (2013833994)
Noriziati Abd Halim (2013277906)
Noor fadzilah binti Adnan (2013663406)
Abdul Aqib Iqbal bin Abdul Aziz (2013210324)
Muhammad Azizan bin Rozman(2013446662)
2. What is Causal-Comparative Research?
O Determined cause or consequences to the
existing research.
O Something referred as ‘ex post facto’
O Two different type:
O can be manipulated
O manipulate
3. O example of type:
O exploration of effects
O exploration of causes
O exploration of consequent
O Similarity & differences between
correlational research and experimental
research to causal-comparative research
6. STEP
EXAMPLE
Interest in Student
Creativity
5W 1H
Questions
1. Who is the target person?
2. What cause the creativity?
3. Why do only certain student
got the creativity while other
don’t?
4. When the students are
creative?
5. How do they show their
7. STEP
INSTRUMENTATION
No limits of using
instrumentation
Example :
1.
2.
3.
4.
5.
Questionnaires
Achievement Test
Interview Schedule
Attitudinal Measures
Observational
Devices
9. STEP
Example of Basic Causal
Comparative Design
Group
1
Group
:
Independent
Variable
Dependent
Variable
Independent
Variable
Dependent
Variable
C
O
C
O
Dropout
Student
Independent
Level of
Creativity
Dependent
Art
Student
Independent
Level of
Creativity
Dependent
Variable
Variable
Variable
Variable
-C
O
-C
O
Non
Dropout
Student
Level of
Creativit
y
Non
Art
Student
Level of
Creativit
y
Group
2
Group
10. Threats to Internal Validity in CausalComparative Research
O Divided into two threats:
O Subject Characteristics
O Other threats
O Have two weaknesses:
O Lack of randomization – since the groups
are already formed.
O Inability to manipulate an independent
variable – the groups have already been
exposed to the independent variable.
11. O Subject Characteristics:
O The major threats to the internal validity of
a causal-comparative study
O The researcher has had no say in either
the selection or formation of the
comparison groups, there is always the
likelihood that the groups are not
equivalent on one or more important
variables other than the identified group
membership variable.
O Three types of procedures can be use to
reduce the chance of this threats which is:
O Matching of Subjects
O Finding or Creating Homogeneous
Subgroups
O Statistical Matching
12. O Matching of Subjects:
O To control for an extraneous variable is to
match subjects from the comparison
groups on that variable.
O Pairs of subjects, one from each group, are
found that are similar on that variable.
O Eliminate/reduced the particular subject if
match cannot be found.
13. O Finding or Creating Homogenous
Subgroups:
O Create groups that are relatively
homogenous on that variables – to control
for an extraneous variable.
O Find two groups that have similar subject –
form subgroups that represent various
levels of the extraneous variable (eg.
high, middle, low) – compare the
comparable subgroups.
14. O Statistical Matching:
O To control for an important extraneous
variable.
O Adjusts scores on a posttest for initial
differences on some other variable that
assumed to be related to performance on
the dependent variable.
15. O Other Threats:
O Depends on the type of study being
considered.
O Eg. In non invention studies, If the persons
who are lost to data collection are different
from those who remain (as is often
probable) and if more are lost from one
group than the other(s), internal validity is
threatened.
O If unequal numbers are lost, an effort
should be made to determine the probable
reasons.
16. O Conclusion:
O Subject Characteristics:
O Deal with only four – socioeconomic level of
the family, gender, ethnicity, and marketable
job skills.
17. Evaluating threats to internal Validity
in Causal-Comparatives Studies
O -involves a set of steps similar for
experimental studies
O Step 1: the researcher need to be concerned
with factors unrelated to what is being studied.
O Step 2 : What is the likelihood of comparison
groups differing on each of these factors?
(that different between group cannot be
explained away by factor that is the same for
all group)
O Step 3 : Evaluate the threats on the basis of
how likely they are to have an effect and plan
to control for them.
18. Subject characteristics
O Ex:
-gender
-ethnicity
Mortality
Step 1:probable refusing to be interview is
related the hypothesis causal variable
Step 2: more student in the dropout refuse to
interview
Step 3: likelihood of having an effect unless
control:high
19. Instrumentation
Instrument decay
O Step 1 -this study means interview fatigue
O Step 2 -the fatigue could be different for the two groups
O Step 3 -likelihood of having an effect unless control:moderate
Data Collector characteristics
O Step 1-Can be expected to influence the information obtained
on
the hypothesis causal variable
O Step 2 -Interview should be balance across the two groups
O Step 3 - Likelihood of having an effect unless control :moderat
Data collector bias
O Step 1 -bias might be related to information obtained on the
hypothesis
O Step 2 -bias might differ for the two groups
O Step 3 - likelihood of having an effect unless control: high
20. Other treats
O -implementation, history, maturation, attitudinal and
regression threats
O -trick to identifying threats to internal validity in
causal study
O -based on evidence or experience
O -can be greatly reduced if causal comparative are
replicated
22. Analyzing data
O
O
First step in analyzing data in causal
comparative study is :
To construct frequency polygons and
then calculate the mean and standard
deviation of each group.
Means and standard deviation are
usually calculated if the variables
involved are quantitative.
23. O Commonly used test in causal-
comparative studies is a :
O t –test : its for differences between means.
O When more than 2 groups are used, then
either an analysis of variance or an
analysis of covariance is the appropriate
test.
24. • Analysis of covariance
• -The Analysis of Covariance (generally
known as ANCOVA) is a technique that
sits between analysis of variance and
regression analysis.
• Particularly helpful in causal-comparative
research.
• Its provide a way to match group on such
variable as age, socioeconomic, status
and so on.
25. O Before analysis of covariance can be used
the data involved need to satisfy certain
assumptions.
O The result must be interpreted with
caution.
O Causal-comparative studies are good at
indentifying relationship between variable
but do not prove cause and effect.
26. 2 ways to strengthen the interpretability of
casual-comparative studies
O First, alternative hypothesis should be
formulated and investigated.
O Second, if the dependent variable
involved are categorical the study should
be examined using the technique of
discriminant function analysis.
O The most powerful way to check on
possible causes is perform an experiment.
27. Steps Involved in CausalComparative Research
O Problem Formulation
O The first step is to identify and define the particular
phenomena of interest and consider possible causes
O Sample
O Selection of the sample of individuals to be studied
by carefully identifying the characteristics of select
groups
O Instrumentation
O There are no limits on the types of instruments that
are used in Causal-comparative studies
O Design
O The basic design involves selecting two or more
groups that differ on a particular variable of interest
and comparing them on another variable(s) without
manipulation (see Figure 16.1)
28. Threats to Internal Validity in
Causal-Comparative Research
O Subject Characteristics
O The possibility exists that the groups are not
equivalent on one or more important variables
O One way to control for an extraneous variable
is to match subjects from the comparison
groups on that variable
O Creating or finding homogeneous subgroups
would be another way to control for an
extraneous variable
O The third way to control for an extraneous
variable is to use the technique of statistical
matching.
30. Other Threats
O Loss of subjects
O Location
O Instrumentation
O History
O Maturation
Data collector bias
Instrument decay
Attitude
Regression
Pre-test/treatment
interaction effect