There are three true experimental designs that involve forming control and experimental groups through randomization to establish equivalence. The post-test only equivalent group design is described as the most effective, where groups are randomly assigned to a treatment or no treatment condition and both groups are post-tested. An example is provided where students are randomly assigned to a reward or no reward condition to test the effect on retention. The pre-test, post-test control group design is also outlined, which involves pre-testing and post-testing both the experimental and control groups. Issues with internal and external validity are discussed. Finally, the Solomon four-group design is introduced as a way to control threats to validity by combining features of other designs.

1. True Experimental Designs
Submitted by
M.Jeya Kumari
M.phil.,

2. True experimental research
• There are three experimental designs, which are
called true experimental designs.
• In these designs the control group and the
experimental groups are formed, and their
equivalence established through randomization.
Comparison which is made is not between a
controlled intervention and nothing but between two
different types of experiences of two different
designs, the one belong planned and designed, and
the other being what would have happened anyway.
As with the static group comparison design, it is
important to be aware that the

3. As with the static group comparison design, it is
important to be aware that the comparison
which is made is not between a controlled
intervention and nothing but between two
different types of experiences of two different
designs, the one belong planned and designed,
and the other being what would have happened
anyway. As with the static group comparison
design, it is important to be aware that the

4. Post – test Only Equivalent Group
Design

5. • This design is the most effective and
useful experimental design, which
minimizes the threats to experimental
validity.
• Continues…
Randomly picked
experimental (R)
Special Treatment (X) Post-test (O1)
Randomly picked control
group (R)
No special treatment Post-test (O2)

6. Post-test only equivalent group designs
• In the above design there are two group. One
group is given treatment (X), usually called the
experimental group, and the other group is not
given any treatment called the control group.
Both groups are formed on the basis of random
assignment of the subject, and hence they are
equivalent.

7. Example
• Suppose a researcher randomly selects 100
students out of a total of 1000 students, using a
table of random numbers. These selected 100
students are then randomly assigned to two
groups. The researcher is interested in
evaluation of the effort of reward over
retention of a verbal task.
• Continues…

8. The hypothesis is that reward enhances the retention
score. One group is given reward (X) while learning a
task and another group receives no such reward while
learning a task. Subsequently, both are given the test
of retention. A simple test of ANOVA would help the
researcher in making a decision on whether to accept
or reject the hypotheses.

9. Pre-test, Post-test Control Group Design
• These type of design have been described as
true experimental designs because they always
include the processes of randomization. As
with the static group comparison which is
made is not between a controlled intervention
and nothing at all, but between two different
types of experiences of two different groups
Continues…

10. the one being planned and designed and
the other being what would have
happened anyway, The pre-test/post-test
control group design is shown in
Randomly picked
experimental
group R
Pre-test
O1
Special
Treatment X
Post-test
O2
Randomly picked
control group R
Pre-test
O3
No special
treatment
Post-test
O4

11. Pre-test, post-test control group-
design
• It shows an experimental design to discover
whether a new method of history teaching
improve pupils’ knowledge and understanding.
• It is obvious from that the deign has two
groups’ reviews. One group receives the
treatment (X) and another group receives no
such treatment.
• Continues…

12. By use of the control group this design
controls some sources of internal
invalidity like history, maturation and
statistical regression.
Group A Tested for
historical
knowledge
Received-new
method
Post-test
Group B Tested for
historical
knowledge
Received-old
method
Post-test
Continues…

13. Exhibit 8.7 Pre-test, post-test control group
design illustration
• Having conducted the experiment, the researcher
collected and tabulated the data as shown in Table 8.1
Pre-test
(average group
scores)
Post-test
(average group
scores)
Improvement
Group A
Group B
48%
48%
68%
53%
Continues…

14. There are a number of problems, even if
the conclusion was drawn that the new
seemed to have increased pupils
knowledge and understanding. In order to
be absolutely sure, it would be important
to check that the two groups were
similarly matched and that the conditions
in which they were taught were the same.
Continues…

15. A number of variables may have had an
impact on how the old and new methods
are received; for example, there may be a
gender bias in one group compared with
the order, If there is, is it a significant
factor in the success or otherwise of the
teaching programme? For the comparison
between the scores obtained by the two
groups to be valid, those other factors
which may affect the amount and quality
Continues…

16. of learning have to be controlled, Only then can the
researcher be certain that one method of history
teaching (Method A) is a better method than the other
(Method B). If less time is spent in Method B, or
Method B is always taught in the last period in the
afternoon when the children are less attentive, or in
poorly ventilated and crowded conditions, then he or
she cannot be certain that Group B has performed less
well, than Group A only because of the type of
teaching method being employed
Continues…

17. . A partial solution to this problem is afforded by the
process of random sampling. If two groups can be
picked randomly (the one to act as the experimental
group and the other to act as a control group), then it
is possible to be fairly certain that internal
characteristics of the two group will be similar. The
effect of randomly choosing experimental and control
groups is that members of each group have an equal
chance of being influenced by factors other than the
intervention.
Continues…

18. The Solomon Four-Group Design
• The Solomon four Group design developed by
Solomon (1949) is really a combination of the two
equivalents – groups design namely, the post-test only
design and pre-test, post-test only design represents
the first direct attempt to control the threats of the
external validity.
• Continues…

19. A two way analysis of variance is used to compare the
four post-test scores; analysis of covariance to
compare gains on O2 and O4 since this design
provides for two simultaneous experiments, the
advantages of replication being incorporated.
The effect of the intervention is replicated in four
different ways. Experimentalists can compare pre-
tested and post-tested
Continues…

20. scores with a single group (Group A) that has
received the intervention. This will allow them to
make a valid comparison but it may have weal
internal and external validity. The second
comparison they can make is between a group
which has not been subject to the intervention
Continues…
R
Randomized group A
O1 X
Pre-test Receives intervention
O2
Post-test
R
Randomized group B
O3
Pre-test No intervention
O4
Post-test
R
Randomized group C Receives intervention
O5
Post-test
R
Randomized group D No intervention
O6
Post-test

21. Exhibit 8.8: Four-group design
• (Group B) and one that has (Group A), though this
does allow them to filter out the effects of testing-
both its potential unreliability and its potential
receptivity. They can therefore determine the
differential between scores achieved by Group A and
group B, when one has been subjected to the
experimental intervention and the other has not.
Continues…

22. The third comparison they can make is between
Group C and Group D, where the former has only
been tested after the intervention, while the latter has
received no intervention but has been post-tested.
This allows them to make a judgment about the
effects of the intervention when test-sensitizing
effects have been eliminated. The children could not
have behaved in unnatural ways because they were
being tested, though there may still be a problem with
sensitizing effects of the intervention itself.
Continues…

23. This has the effect of increasing the ability of
researchers to generalize to other settings. Finally,
they can compare Group C and Group B, where the
former has been but not pre-tested, while the latter
has been pre-tested and post-tested even though it has
been subject to any intervention again, this allow then
to be confident that the experimental mechanism
which they have set in place is not responsible for the
effects they observed and therefore allows them to be
more confident about generalizing to other settings.