CHAPTER II
INTRODUCTION
A. Causal-Comparative Research.
Like correlational research, causal–comparative research is
sometimes treated as a type of descriptive research because it too describes
conditions that already exist. Causal–comparative research, however, also
attempts to determine reasons, or causes, for the existing condition.
Causal–comparative is thus a unique type of research, with its own research
procedures. In causal–comparative research the researcher attempts to determine
the cause, or reason, for existing differences in the behavior or status of
groups or individuals. In other words, established groups are already different
on some variable, and the researcher attempts to identify the major factor that
has led to this difference. Such research is sometimes called ex post facto,
which is Latin for “after the fact,” because both the effect and the
alleged cause have already occurred and must be studied in retrospect. For
example, a researcher may hypothesize that participation in preschool education
is the major factor contributing to differences in the social adjustment of
first graders. To examine this hypothesis, the researcher would select a sample
of first graders who had participated in preschool education and a sample of
first graders who had not and would then compare the social adjustment of the
two groups. If the children who participated in preschool education exhibited
the higher level of social adjustment, the researcher’s hypothesis would be
supported. Thus, the basic causal–comparative approach involves starting with
an effect (i.e., social adjustment) and seeking possible causes (i.e., did
preschool affect it).[1]
Moreover, According to Marguerite
G ,” Causal-comparative research, or ex–post facto research, is a
research approach that seeks to explain differences between groups by examining
differences in their experiences. Like experimental research, it examines the
effect of an independent variable (the past experience) on a dependent variable
while also trying to control extraneous variables. However, unlike experimental
research, the independent variable (the past experience) has either already
occurred or it would be unethical to manipulate. For example, let us say that
you are interested in what causes the differences in the readiness skills of
kindergarten students. After reading past research studies, you decide to
examine preschool attendance as an independent variable that might have
“caused” a difference in kindergarten
readiness (the dependent variable). Preschool attendance has already occurred
or happened; as a researcher, you cannot control or manipulate it. If you were
to conduct such a study, you will simply identify two groups, one group that
attended preschool and one group that did not, and then measure and compare
school readiness scores. If the groups differ on their readiness scores, the
researcher infers that preschool attendance caused the readiness scores to
differ. However, caution is warranted. Because no random assignment occurred,
the two groups being studied could be very different to begin with, which might
mean that other factors and not preschool attendance caused the difference in
readiness scores. For example, there may be differences in family income or
parental levels of education (or both). Therefore, making sure that the two
comparison groups are as similar as possible on all other extraneous variables
(other than the independent variable) is a critical part of designing a
causal-comparative study”.[2]
According to Ari Donald, “The designation ex post facto, from Latin for “after the
fact,” indicates that ex post facto
research is conducted after variation in the variable of interest
has already been determined in the natural course of events. This method is
sometimes called causal comparative because its purpose is to
investigate cause-and-effect relationships between independent and dependent
variables. Researchers use it in situations that do not permit the
randomization and manipulation of variables characteristic of experimental
research. Thus, much of the basic rationale for experimental and ex post facto
is the same. They both investigate relationships among variables and test
hypotheses. However, with an experiment it is possible to obtain much more
convincing evidence for a causal (functional)
relationship among
variables than can be obtained with ex post facto studies. The effects of
extraneous variables in an experiment are controlled by the experimental
conditions, and the antecedent independent variable is directly manipulated to
assess its effect on the dependent variable”.[3]
Like
experimental research, causal-comparative research involves comparing groups to
see if some independent variable has caused a change in a dependent variable.
Causal-comparative research also sets up studies so that possible extraneous
variables are controlled. However, the types of research questions addressed in
causal-comparative research involve variables that are difficult or impossible
to manipulate experimentally, often because they are experiences that have
already occurred. Rose’s study is one example of a causal-comparative study. Following
are some questions that might be addressed using causal-comparative research.
1.
Do children with a
history of abuse have lower levels of academic achievement than children with
no history of abuse?
2.
Do students who are
retained a grade have high school graduation rates different from those who are
not retained?
3.
Are women who attend a
same-sex college more likely to attain leadership positions after graduation
than women who attend coed colleges?
Note
that in these questions, we are attempting to see if one variable (abuse,
retention, working, or type of college) causes a change in another variable
(academic achievement, graduation rates, or leadership). However, we cannot
ethically or practically manipulate the variables that are thought to cause
change. Causal comparative research designs permit the study of the effects of
variables that have already occurred or are difficult to manipulate
experimentally with human research participants. In many causal-comparative
studies, the independent variable has already occurred, e.g., child abuse. This
is why the researcher cannot control or manipulate the independent variable; it
has already happened. In other studies, it might be possible to manipulate such
variables, but it would be unethical to do so. For example, researchers could
not ethically retain one group of research participants for a grade to study
the effect of retention on academic performance. Sometimes it may simply be
impractical to manipulate the independent variable. If students are already in
classrooms with teachers who have established instructional practices, it may
not be feasible to randomly assign classes to treatments or individual students
to treatments. In this case, a causalcomparative study might be required.
https://www.youtube.com/watch?v=zrPjBYEp8Sw
B.
..Steps in
Causal-Comparative Research
Causal-comparative
research often looks deceptively simple. One identifies two groups that had
different experiences and then measures how this affected them. However,
high-quality causal-comparative research requires careful thinking at each stage.
The steps involved in doing causal-comparative research are summarized below:
1.
Select
a topic.
In
causal-comparative research, the topic is likely to be based on past
experiences that are thought to have a strong effect on persons’ later
behaviors.
2.
Review
literature to identify important variables.
The
researcher reviews literature to identify what research has revealed about the
impact of the past experience on later behavior. Potential extraneous variables
might also be identified through the review of literature. For example, if one
was examining the leadership positions of women who attended same-sex versus
coed colleges, one might find that students at single-sex colleges tend to come
from families with higher levels of income and education. Also, the researcher
might find useful information about the methods used to select samples in past
studies or measure possible dependent variables. If one wanted to compare
children with a history of abuse and those with no history of abuse, studies
might reveal how these researchers were able to identify possible participants.
Based on a review of the literature, one would identify an independent variable
(prior experience or group difference that cannot or should not be manipulated)
and a dependent variable that might be affected by this independent variable.
3.
Developing
a research hypothesis.
Research
hypotheses for causal-comparative research take a form that is similar to
experimental research hypotheses because both types of research include an
independent and dependent variable. The research hypothesis would state the
expected causal relationship between the independent and dependent variables.
For example, the research hypothesis for the study of working part-time and
high school achievement might be It is hypothesized that students who are
employed 15 hours or more a week will have lower achievement than students who
are employed five hours or less a week. In this hypothesis, being employed or
not employed is the independent variable. The dependent variable would be
achievement as measured by high school grade point average.
4.
Clearly
define the independent variable.
In
causal-comparative research, the independent variable describes the different
past experiences of the participants. It is important to be clear about the
exact differences in the experiences of the two groups being compared. In our
opening example, employment was defined as working 15 hours or more per week.
In studying single-sex versus coed schools, one might want to indicate what
male-to-female ratios are included in the coed schools. In the study of
children with and without a history of abuse, one would discuss how the
information documenting the abuse was obtained. The definition of the
independent variable identifies the two populations from which participants
will be selected.
5.
Selecting
participants using procedures to control extraneous variables.
Unlike
experimental research, the participants in causal-comparative research already
belong to groups based on their past experiences, and so the researcher selects
participants from these preexisting groups. An important consideration in
designing causal-comparative studies is whether the two groups are similar
(comparable) except for the independent variable on which they are being
compared. If two groups are formed because they differ on the independent
variable, but they also happen to differ on other extraneous variables, the
researchers will not know whether group differences on the dependent variable
are caused by the independent or extraneous variables. If the employed students
were found to have lower scores on a measure of scholastic aptitude, we would
have to ask whether their lower academic achievement (the dependent variable)
is the result of their employment (independent variable) or their lower
academic aptitude (extraneous variable). To rule out the influence of the
extraneous variable, the counselor selected groups of students with different
levels of employment but with similar aptitudes (based on their freshman grade
point averages). Ideally, the two groups should be selected randomly, which
Rose did not do. Therefore, she cannot generalize the results of her sample to
the whole population of students at her school. Typically, researchers will try
to select participants who differ on the independent variable but are
comparable in other ways. Causal-comparative researchers use the same controls
for extraneous variables as those used in experimental research (except for
random assignment). These include matching, holding a variable constant,
comparing homogeneous subgroups, pretesting (when a researcher is comparing
intact groups who are about to receive a treatment that cannot be randomly
assigned, such as a new curriculum), use of factorial designs, and statistical
controls such as the use of analysis of covariance (ANCOVA) or multiple
regression. Although these were described in Chapter Eight, we should point
out that matching and ANCOVA are especially common in causal-comparative
designs because random assignment cannot be used to make sure that participants
are similar. To use these controls, the researcher must obtain measures of the
extraneous variables. If a researcher wants to use matching to make sure that
the group of participants who have been abused are similar in family income to
the group of participants who have not been abused, then information on family
income must also be obtained. The most common way that researchers today
control extraneous variables in causal-comparative studies is by statistically
estimating the effect of the extraneous variable on the dependent variable.
Some statistical tests, such as multiple regression, use correlation
coefficients to compare the size of effects of the independent variable and
extraneous variable on the dependent variable. Another statistical procedure, analysis
of covariance, or ANCOVA, compares the mean scores of the two groups after
the effect of the extraneous variable has been removed. This test estimates how
much the extraneous variable affects the dependent variable, and it
statistically adjusts the group means to take into account the initial
differences between the groups. However, again, to use these statistical
controls, there must be a reliable and valid measure of the extraneous
variable. Much of the work in designing a high-quality causal-comparative study
is focused on measuring and controlling possible extraneous variables. Note
that the random selection of students by itself does not control for
extraneous variables that might differ between the two groups. If students
working fewer than five hours a week generally have parents with higher levels
of education than students working 15 hours or more a week, randomly selecting
students may result in samples that still differ in parental education level.
The random selection would ensure that each group is representative of its
population. However, if the two populations differ in parental education, so
will the two samples randomly selected.
6.
Selecting
reliable and valid measuring instruments.
Selecting
appropriate instruments is an important issue in all types of quantitative
research. A researcher interested in the question of same-sex versus coed
colleges and leadership positions would certainly need to find or develop a
measure that accurately measured the dependent variable or types of leadership
positions participants had held.
7.
..Collect
data.
In
causal-comparative research, there is no treatment to administer. So once the
sample and measures have been selected, carrying out the study simply involves
obtaining data from the selected participants on the measures. If the measures
are archival data, then this may involve obtaining permission to access the
records. If a measure involves completion of a questionnaire, procedures must
be established to distribute these to the participants and have them returned
or the researcher could administer them in a group setting. Note that obtaining
permission or lack of return of the measures might change the sample and open
the possibility that extraneous variables have not been controlled.[4]
8.
Analyze
data to see if the groups differ.
Data
are usually reported as frequencies or means for each group. Inferential
statistical tests are used to determine whether the frequencies or means
reported for the groups are significantly different from each other. These are
the same statistical tests used in experimental research (listed in Chapter
Eight). Based on the results of these tests, the researcher would either accept
or reject the null hypothesis.
9.
Interpreting
the results.
If
the results of the statistical test are significant and extraneous variables
have been well controlled for, the researcher can conclude that the study
provides support for the research hypothesis. However, one should always be
cautious about stating that a causal-comparative study has “proved” that a
causal relationship exists. Causal-comparative research is valuable in
identifying possible causes or effects, but it usually cannot provide
definitive support for the hypothesis that one of the variables studied caused
the observed differences in the other variable. Evidence from causal-comparative
studies is considered to be weaker evidence of causality than experimental
studies, which show that a dependent variable changes only after the
researcher has manipulated the independent variable. When many
causal-comparative studies have been conducted by different researchers working
with different samples in different settings and consistent results emerge from
these studies, the combined evidence from these studies provides stronger
evidence of causality. This has been the case with research on smoking and lung
cancer. The probability that these results could occur by chance if smoking
does not cause lung cancer is so slight that most scientists who have worked in
the area have accepted the combined results as compelling evidence of a causal
relationship.
1. In causal–comparative research, the researcher attempts to
determine the cause, or reason, for existing differences in the behavior or
status of groups.
2. The basic causal–comparative approach is retrospective; that is,
it starts with an effect and seeks its possible causes. A variation of the
basic approach is prospective—that is, starting with a cause and investigating
its effect on some variable.
3. An important difference between causal– comparative and
correlational research is that causal–comparative studies involve two (or more)
groups of participants and one grouping variable, whereas correlational studies
typically involve two (or more) variables and one group of participants.
Neither causal–comparative nor correlational research produces true
experimental data.
4. The major difference between experimental research and
causal–comparative research is that in experimental research the researcher can
randomly form groups and manipulate the independent variable. In causal–comparative
research the groups are already formed and already differ in terms of the
variable in question.
5. Grouping variables in causal–comparative studies cannot be
manipulated, should not be manipulated, or simply are not manipulated but could
be.
6. Causal–comparative studies identify relations that may lead to
experimental studies, but only if a relation is established clearly. The
alleged cause of an observed causal– comparative effect may in fact be the
supposed cause, the effect, or a third variable that may have affected both the
apparent cause and the effect.[5]
[1] L. R. Gay, Geoffrey E.
Mills, Peter Airasian, Educational
Research: Competencies for Analysis and Applications, (United States:
Pearson Education, Inc 2012), 10th ed., p. 228.
[2] Marguerite G., Lodico,
Dean T., Spaulding, Katherine H. Voegtle, Methods
in Educational Research: From Theory to Practice, (San Francisco: John
Wiley & Sons, Inc 2006), p. 14.
[3] Donald Ary, Lucy Cheser
Jacobs, Chris Sorensen, Asghar Razavieh, Introduction
to Research in Education, (Canada: Nelson Education, Ltd 2010) 8th
ed., p. 332.
[4] Marguerite G., Lodico,
Dean T., Spaulding, Katherine H. Voegtle, Methods
in Educational Research: From Theory to Practice,…...........p. 213.
[5] L. R. Gay, Geoffrey E.
Mills, Peter Airasian, Educational
Research: Competencies for Analysis and Applications,..............p. 235.
No comments:
Post a Comment