EXPERIMENTAL RESEARCH MANIPULATION

INTRODUCTION • Manipulation refers to conscious

control of the independent variable by
• Experimental is most scientifically the researcher through treatment or
sophisticated research method. intervention(s) to observe its effect on
• It is defined as ‘observation under the dependent variable.
controlled conditions. • In other words, it is a conscious act by
• Experimental research design is the researcher, where he or she varies
concerned with examination of the effect the independent variable & observes
of independent variable on the dependent the effect that manipulation has on
variable, where the independent variable the dependent variable of interest.
is manipulated through treatment or • For example, a researcher is conducting
intervention(s), & the effect of those a study on efficacy of chlorhexidine
interventions is observed on the mouthwash on the prevention of
dependent variable. ventilator-associated pneumonia (VAP)
among patients admitted in CUs.
Experimental research • In this example, chlorhexidine
• An inquiry on cause-and effect mouthwash is the independent
relationships variable, which is manipulated by the
• The researcher manipulates researcher, & is used as an
independent variables (e.g., type of intervention for the experimental
treatment, teaching method, group, while the control group is kept
communication strategy) and measures deprived of it to observe its effect on
dependent variables (anxiety level, the incidence of VAP.
English comprehension, satisfaction)
• The independent variable is CONTROL
CONTROLLED or set by the researcher.
• The dependent variable is MEASURED • Control refers to use of control group &
by the researcher. controlling the effects of extraneous
variables on the dependent variable in
Concept… which researcher is interested.
True experimental research designs are those • The subject in the control & experimental
where researchers have complete control over
groups are similar in number &
the extraneous variables & can predict
characteristics, but the subject in the
confidently that the observed effect on the
dependable variable is only due to the control group receive no experimental
manipulation of the independent variable. treatment or any intervention at all.

• The experimental group receives the

Essential Characteristics… planned treatment or intervention & a
✓ A true experimental research design comparison is made with the control group
must essentially consist of the following to observe the effect of this treatment or
three characteristics: intervention.
• Manipulation
• Control RANDOMIZATION
• Randomization
• Validity
• Randomization means that every
subject has an equal chance of being
assigned to experimental or of study
subjects on a random basis.
• Chances of systemic bias is eliminated.
• Randomization is used in true
experimental research design to minimize
the threat of internal validity of the study
& to eliminate the effect of extraneous
variables on dependent variables.
• On average the characteristics of the
subject in experimental & control groups
are similar, thus influence of extraneous
variables on dependent variable is
eliminated by dispersing the variability of
the subject characteristics equally in both
the groups.

METHODS OF RANDOMIZATION:

❑ Random assignment of subject may done

with simple flip of a coin for each subject;
if coin lands on its ‘head’, subjects are
assigned to first group & with ‘tail’ subjects
are assigned to control group.
❑ Another possible method is to write the Techniques of Research Control
names of the subjects on slips of paper
& put the slips into a bowl & then draw 1. Control of External Factors
lots. The first designated numbers of a. Control of Environment – has to
subjects are placed in one group, & rest be researcher-structured
are assigned under another group. b. Control of Time – constancy of
data collection at the same time
❑ Thirdly a random table may be used to of the day
facilitate the randomization process. In
this method, blind-folded subjects choose 2. Control of Intrinsic Factors
a number from a table of number Homogeneity – only subjects with similar
horizontally (row) or vertically (columns), characteristics are included
till a requisite number is reached for both e.g. A study on effects of age on
experimental & control groups. physical fitness. Participation is
Computer-assisted random sequences limited only to those within
also may be used for the random specified age group
assignment of the subjects
VALIDITY

The researcher is interested in controlling or

altogether removing extraneous variables that
may create spurious or inaccurate results.
Hence, ensuring validity helps to objectively test
hypothesis.
Threats to experimental research TESTING (ENDOGENOUS CHANGE)

Internal validity A common research design is to give a

Internal invalidity asks the question, “Are the group a pre-test, a treatment, and then a post-
measurements I make on my dependent (i.e., test. If the same test is used both times, the
the variable I measure) variable influenced only group may show an improvement simply
by the treatment (IV), or are there other because of their experience with the test. This
influences which change it?” is especially true when the treatment period is
short and the tests are given within a short
1. History time.
2. Maturation (endogenous change)
3. Testing (endogenous change) INSTRUMENTATION
4. Instrumentation
5. Statistical regression (endogenous change) If you use different tests for pre- and
6. Differential selection/ Selection Bias post-measurements, then the change in pre-
7. Experimental mortality and post-scores may be due to differences
8. Selection-Maturation Interaction of between the tests rather than the treatment.
Subjects The best remedy is to use randomization and a
9. The John Henry Effect/ Compensatory post-test only design.
Rivalry
STATISTICAL REGRESSION (ENDOGENOUS
HISTORY CHANGE)

History refers to events other than the Statistical regression refers to the
treatment that occur during the course of an tendency of extreme scores, whether low or
experiment which may influence the post- high, to move toward the average on a second
treatment measure of treatment effect. If the testing. Subjects who score very high or very
explosion of the nuclear reactor in Chernobyl, low on one test will probably scoreless high or
Ukraine had occurred in the middle of a six- low when they take the test again.
month treatment to help people reduce their That is, they regress toward the mean.
“anxiety of nuclear power,” Do not study groups formed from extreme
Must occur during the experiment. scores. Study the full range of scores.
Controlled by control group
DIFFERENTIAL SELECTION/ SELECTION BIAS
MATURATION (ENDOGENOUS CHANGE)
If we select groups for “treatment” and
Subjects change over the course of an “control” differently, then the results may be
experiment. These changes can be physical, due to the differences between groups before
mental, emotional, or spiritual. Perspective can treatment.
change. The natural process of human growth
can result in changes in post-test scores quite Randomization solves this problem by
apart from the treatment. statistically equating groups.
Question: How would a “control group” control
this source of internal invalidity EXPERIMENTAL MORTALITY
 Experimental mortality, also called an experimental treatment, they tend to work
“attrition,” refers to the loss of subjects from harder.
the experiment. If there is a systematic bias in When this occurs, differences between control
the subjects who drop out, then posttest scores and treatment groups is decreased, minimizing
will be are biased. For example, if subjects drop the perceived treatment effect.
out because they are aware that they’re not
improving as they should, then the post-test
scores of all those who complete the treatment External validity
will be positively biased. Your results will Degree to which results are influenced by
appear more favorable than they really are. external factors or population and settings.
How does use of a control group solve the
problem of attrition?
Threats to external validity
SELECTION-MATURATION INTERACTION OF Hawthorne Effect
SUBJECTS Halo effect
Multiple Treatment Effect

Interaction means the mixing or combining of

separate elements. If you draw a group of
subjects from one school to serve as the HAWTHORNE EFFECT
treatment group, and a second group from a It occurs when study participants
different schools to serve as a control, you respond in a particular manner, or there is
could well find -- beyond the simple problem of obvious change of behavior because they are
selection differences (“Are the two groups aware that they are being observed.
equivalent?”) -- a mixing of selection and
maturation factors to compound the HALO EFFECT
extraneous influence on your measurements. Tendency of the researcher to rate the
For example, if the two schools differ in the subject high or low because of the impression
average age of their members, they may well he has on the latter
respond to the treatment differently due to
inherent maturational factors. Remedy to minimize threats
Double Blind Method – neither the subject nor
Randomly selecting all subjects from a defined the person administering the treatment or
population solves this problem. placebo knows what is being administered
Double Observer Method – two observers
record the subjects performance
THE JOHN HENRY EFFECT/ COMPENSATORY
RIVALRY MULTIPLE TREATMENT EFFECT

John Henry, the legendary “steel drivin’ man,” Normally we find a single treatment in
set himself to prove he could drive railroad an experiment. If, however, an experiment
spikes faster and better than the newly exposes subjects to, say, three treatments (A,
invented steam-powered machine driver. He B, and C) and test scores show that treatment
exerted himself so much in trying to outdo the C produced the best results, one cannot
"experimental" condition that he died of a declare treatment C the best. It may have been
ruptured heart. If subjects in a control group the combination of the treatments that led to
find out they are in competition with those in the results.
Treatment C, given
alone, may produce different results. • For example, to study the effect of an
educational intervention related to
Different types of experimental research can be urinary incontinence on the
conducted depending on the nature of subjects subsequent help-seeking behavior of
and the instruments, and the way data are older adults.
collected and analyzed.
- Will there be a control group?
- How many subjects will there be?
- Will the subjects be randomly selected?
- Will each group be pretested?
- How will the obtained data be
analyzed?
-
What factors may affect the internal validity?
-
What factors may affect the external validity

TYPES OF TRUE EXPERIEMNTAL DESIGN

POST-TEST-ONLY CONTROL DESIGN: PRETEST-POST-TEST-ONLY DESIGN

• Composed of two randomly assigned • In this research designs, subjects are
group, i.e. experimental & control, but randomly assigned to either the
neither of which is pretested before experimental pr the control group.
the implementation of treatment on • The effect of the dependent variable on
the experimental group. both the groups is seen before the
• In addition, while treatment is treatment (pretest).
implement on the experimental group • Later, the treatment is carried out on
only, post-test observation is carried experimental group only, & after-
out on both the group to assess the treatment observation of dependent
effect of manipulation. variable is made on both the groups to
• This design can be helpful in situations examine the effect of the manipulation
where it is not possible to pretest the of independent variable on dependent
subjects. variable.
 • For example, such a design could be • In this, experimental group 2 was
used for ‘an experimental study the observed at one occasion, & that score
effectiveness of the use of guava leaves should be similar to average scores of
for external flushing to post partum those in experimental & control groups.
patients to prevent infections • To estimate the amount of change in
experimental & control group 2, the
average test scores of experimental &
control groups 1 are used as baseline
• The solomon four-group design is
believed to be most prestigious
experimental research design, because
it minimizes the threat to internal &
external validity.

SOLOMON FOUR-GROUP DESIGN

• There are two experimental groups

(experimental group 1 & experimental
group 2) & two control groups (control
group 1 & control group 2).
• Initially, the investigator randomly
assigns subjects to the
• four groups.
• Out of the four groups, only
experimental group 1 & control group1
receives the pretest, followed by the
FACTORIAL DESIGN
treatment to the experimental group 1
& experimental group 2.
• In factorial design, researcher manipulates
• Finally, all the four groups receive post-
two or more independent variables
test, where the effects of the
simultaneously to observe their effects on
dependant variables of the study are
the dependent variables.
observed & comparison is made of the
• This design is useful when there are more
four groups to assess the effect of
than two independent variables, called
independent variable (experimental
factors to be tested.
treatment) on the dependant variable.
• For example, a researcher wants to • In this example, to ensure the
observe the effect of two different homogeneity among the subjects
protocols of mouth care on under treatment, researcher
prevention of VAP when performed randomly places the subjects in
at different frequencies in a day. homogeneous groups (blocks) like
• This design also facilitates the testing of patients with primary hypertension,
several hypothesis at a diabetic patients with hypertension,
single time. & renal patients with hypertension .
• Typical factorial design incorporates
2X2 or 2X3 factorial, but it can be in
any combination.
• The first number (α) refers to the
independent variables or the type of
experimental treatments, & the second
number (β) refers to the level or
frequency of the treatment.

CROSSOVER/ COUNTERBALANCED DESIGN

• In this design, subjects are exposed

to more than one treatment,
where subjects are randomly
assigned to different orders of
treatment.
RANDOMIZED BLOCK DESIGN • It is also known as ‘repeat measures
• Control of inherent differences design’.
between experimental subjects & • This design is more efficient in
differences in experimental establishing the highest possible
conditions is one of the difficult similarity among subjects exposed to
problems faced by researcher in different conditions, where groups
biological sciences. compared obviously have equal
• When there are a large number of distribution of characteristics.
experimental comparison groups, the • Though crossover design is considered as
randomized block design is used to bring an extremely powerful research design,
homogeneity among selected different sometimes it is not effective because
groups. when subjects are exposed to two
• This is simple method to reduce the different conditions, their responses of
variability among the treatment groups the second condition may be influenced
by a more homogeneous combination by their experience in the first condition.
of the subjects through randomized
block design. • For example, when we are comparing the
• For example, a researcher wants to effectiveness of the chlorhexidine mouth
examine the effects of three different care protocol on group I & saline mouth
antihypertensive drugs on patients with care protocol on the subjects of group II.
hypertension.
• Later, the treatment is swapped, where
group I receives the saline mouth care • This design is useful when the experimenter
& group II receives chlorhexidine. In wants to measure the effects of a
treatment over a long period of time.
• The experimenter would continue to
administer the treatment & measure the
effects a number of times during the course
of the experiment.
• Generally it is a single-subject research, in
such studies, subjects serve as their which the researcher carries out an
own control. experiment on an individual or on a small
number of individuals, by alternating
between administering & then withdrawing
the treatment to determine the
The term quasi- (pronounced kwahz-eye) means effectiveness of the intervention.
almost, near, partial, pseudo, or somewhat.

“almost true experiments,”

• A researcher might assess pain levels of a

group of patients with low-back pain. After
3 weeks of pain assessment, subjects are
taught special exercises to reduce that pain.
During the next 3 weeks, pain levels would
QUASI – EXPERIMENTAL RESEARCH DESIGN
again be measured.
• Quasi-experimental research design
• Measuring a child’s school performance on
involves the manipulation of
a weekly basis, & then introducing a new
independent variable to observe to
teaching technique. Then again measuring
effect on dependent variable, but it
lacks at least one of the two on a weekly basis.
characteristics of the true experimental
design; randomization or a control NONRANDOMIZED CONTROL GROUP DESIGN
group. • AKA‘nonequivalent control group
• Quasi-experimental designs are design’.
generally used to establish the
• identical to the pretest-posttest control
causality (effect of independent
variable on dependent variable) in group design, except there is no
situations where researchers are not random assignment of subjects in
able to randomly assign the subjects to experimental & control groups.
groups or for various reasons no • In this design, experimental & control
control group is available for an groups are selected without
experimental study. randomization, & dependent variables
are observed in experimental as well as
control groups before the intervention.
 • Later, the experimental group receives • A pretest observation of the dependent
treatment & after that posttest variables is made before implementation of
observation of dependant variables is the treatment to the selected group, the
carried out for both the groups to treatment is administered, & finally a
assess the effect of treatment on posttest observation of dependent variables
experiment group. is carried out to assess the effect of
treatment on the group.

• Some researcher also argue this design as

sub type of quasi-experimental research
design. However in absence of both
randomization & control group.

Static-group comparison design

• At least two groups are involved. After

one group receives the treatment, all
PRE – EXPERIMENTAL RESEARCH DESIGN groups are post tested.
• This design has better control over most
• This research design is considered very of the variables
weak, because the researcher has very
little control over the experiment
TYPES OF PRE-EXPERIMENTAL RESEARCH
DESIGN:
o One-shot case design
o One-group pretest-posttest design
o Static-group comparison design:

One-shot case studies:

• In this research design, a single
experimental group is exposed to a
treatment & observations are made
after the implementation of that
treatment.
• There is no random assignment of
subjects to the experimental group &
no control group at all.

One-group pretest-posttest design:

• It is the simplest type of pre-experimental

design, where only the experimental group
is selected as the study subjects.