Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 17

EXPERIMENTAL

DESIGNS

By: MIGUEL P. GOMEZ JR.


Research design

It is a master plan specifying the methods


and procedures for collecting and
analyzing the needed information in a
research study.
Are of four kinds: historical design (focused
in the past what was); descriptive design
(focused in the present condition what
is); experimental design (future what will
be); and case study design (past, present
and future).
Experimental research design

It is a problem-solving approach that the study is


described in the future on what will be when certain
variables are carefully controlled or manipulated.
This design is most useful in the natural sciences such

as Botany, Zoology, Biology, Phycology, Ichthyology,


Chemistry, Physics, Biochemistry, Molecular Biology,
Microbiology, Biotechnology and many others.
the researcher selects participants and divides them into
two or more groups having similar characteristics and,
then, applies the treatment(s) to the groups and
measures the effects upon the groups.
Types of Experimental
Design
1. Single-group design. This design involves a
single treatment with two or more levels.
Illustration:
A researcher wishes to study the effect of
different levels of nitrogen on the yield of
peanuts. In this experimental study, nitrogen as
a single treatment used in different levels as
fertilizer in planting peanut can be 60-30-30
(NPK); 90-30-30 (NPK); 120-30-30 (NPK) and so
on. N stands for nitrogen, P- Phosphorus and K
Potassium.
2. Two-group design. Two comparable groups are
employed as experimental and control groups.
Illustration:
Suppose a researcher wishes to study the
growth rate of groupers (lapu lapu) cultured in
fish cage with and without supplemental feeds.
Grouper cultured in fish cage with supplemental
feed constitutes the experimental group and the
other one cultured in a fish cage without
supplemental feed constitutes the control group.
The t-test is used to determine the significant
difference in the mean weight of grouper cultured
in fish cage with and without supplemental feeds.
3. Two-pair group design. This design is an
elaboration of the two group design wherein there
are two experimental groups and two control groups.
Illustration:
Suppose an investigator wishes to study the
acceptability, nutritive values, and economics of
canned milkfish (Chanos chanos) in salmon and
Spanish styles with and without Sargassum (brown
seaweed)are the two control groups and canned
milkfish in salmon and Spanish styles with Sargassum
as sea vegetables are the two experimental groups.
The ANOVA (Analysis of Variance) is used to
determine the significant difference on the
acceptability, nutritive values and economics of
canned milkfish.
4. Parallel-group design. This type of experimental design
consists of three or more groups wherein one group is
control group with two or more experimental groups. The
control group or parallel group serves as control or basis for
comparison of the experimental groups which is
manipulated or changed, but the experimental groups vary.
Illustration 1:
For instance, a researcher wishes to determine the
acceptability of canned short-bodied mackerel in salmon
style with and without seaweeds as sea vegetables. The
control group is canned mackerel in salmon style without
seaweed and has two experimental groups, namely,
canned mackerel with Sargassum and canned mackerel
with Halymenia. All things are held constant, except the
experimental groups that have an addition of seaweeds or
sea vegetables.
ANOVA is used.
Illustration 2:
Suppose the researcher wishes to conduct a study on
the effectiveness of culturing commercial catfish using,
soybean meal, meat and bone meal, blood meal, and fish
meal as protein supplements. Fish meal is the control
group or parallel group and the experimental groups are
meat and bone meal, blood meal, and soybean as protein
supplements. Mean and ANOVA: two-factor or F-test: two-
factor are the statistical tools used to determine the mean
weight of catfish and significant difference on the mean
weight. If significant difference exists on the mean weight
of catfish, this means protein supplements really differ on
the weight of catfish because one protein supplement is
most effective. If insignificant difference exists, thus the
weight of catfish using the four protein supplements are
almost the same.
5. Pre-test and Post-test group design. This
design involves the experimental group
and the control group which are carefully
selected through randomization
procedures. Both groups are given pretest
at the beginning of the semester and
posttest at the end of the semester. But
the control group is isolated from all
experimental influences.
6. Three-group design. In this design, there are three groups
of independent variables.

Illustration:
Suppose the researcher wishes to conduct a study on the
effect of using direct seeding, transplanting, and using stem
cuttings in growing of Moringa Oleifera at the municipal
grounds of Barotac Viejo. These three methods of culturing
Moringa Oleifera are independent variables and the
dependent variable is the weight of Moringa Oleifera. Mean
and ANOVA, two-factor or F-test: two-factor are the statistical
tools used. Mean is used to determine the mean weight of
Moringa per method. ANOVA or F-test is used to test the
significant difference on the weight of Moringa of the three
methods. If significant difference exists, this means the three
methods of growing Moringa Oleifera really differ with each
other because one method is most effective. If insignificant
difference exists, hence, the weight of Moringa Oleifera using
the three methods are almost the same.
7. Counter-balanced or Latin square
design. This design is also called rotation
design. It involves an exchange of two or
more treatments taken by the subjects
during the experiment. The arrangement
employed in the design is Latin square in
which each variable is a form of square
occurring once in each row or column. This
is called quasi-experimental design.
Illustration:
Suppose an investigator wishes to determine the
effectiveness of growing squash using four different
organic fertilizers to four different nurseries. Due to
different ecological parameters such as light, humidity,
temperature, and oxygen of each nursery, each is
required to have one organic fertilizer. The first
treatment combination consists of OF, N, and S 1 where
SF stands for Organic Fertilizer; N, Nursery; and S,
Squash or scientifically called Cucurbita. The second
combination is OF, N, S and so on. It will be noted that
S or grouper is the diagonal of the square and
appears in each position from first supplemental feed to
the fourth organic fertilizer. F-test or analysis of variance
(ANOVA) is used to determine the significant difference
on the effectiveness of growing squash using four
different organic fertilizer to four different nurseries.
8. Complete Randomized Design. This is a design in which a
group of test plants or animals is studied only once but
subsequent (kasunod) or successive (sunod-sunod)
treatment is applied to determine the cause of change.
There is no control in this design but the subjects will
undergo randomization process.
Illustration:
An investigator wishes to determine if there is a
significant difference in the treatment of pellets as
supplemental feed upon the yield of prawn (sugpo) cultured
in the fishpond. He uses 300 species of prawn fry and three
compartments in the pond. The 300 pieces of prawn fry are
placed at random. Of the 300 prawn fry, 100 pieces are
placed in each compartment should be observed carefully
and ecological parameters such as pH, salinity, oxygen,
etc., should betaken into consideration. Different levels or
proportion of pellets are applied as supplemental feed in
each compartment.
9. Randomized Complete Block Design (RCBD or
RCB). The RCB is the standard design for
agricultural experiments. This experimental design
uses a group of test plants and animals as subjects
of the study which are studied once but
subsequent treatments applied are replicated to
determine the cause of change. There is control in
this design and the subjects will undergo
randomization process.
Illustration:
Suppose the researcher wishes to determine the
effectiveness of cultivating milkfish in the fishpond
using fish meal, bread meal, and ipil-ipil leaves as
supplementary feeds. There are four groups in this
experimental study treated with different feeds.
These
are as follows: (1) First treatment, control group,
lab-lab only as natural food for milkfish; (2) Second
treatment, 1st experimental group, lab-lab and fish
meal; (3) Third treatment, 2nd experimental group,
lab-lab and bread meal; (4) Fourth treatment, 3 rd
experimental group, lab-lab and ipil-ipil leaves.
Each group is replicated three times. In other
words, there are 3 compartments for the control
group or first treatment (T1); 3 compartments for
the 2nd experimental group or third treatment (T3);
and 3 for the 3rd experimental group or fourth
treatment (T4). A total of 12 compartments in all
are under study.
10. Correlational design. This design is used to
determine the relationship of two dependent
variables (X and Y) on how they are manipulated by
the independent variable.
Illustration:
For instance, the researcher wishes to determine
the weight and length relationship of grouper
cultured in a concrete tank using fish meal as
supplemental feed. Fish meal is the independent
variable and weight (X) and length (Y) of grouper
are the dependent variables. In other words, the
weight and length of fish are dependent on the
feed used.
The correlation coefficient (r) is the statistical tool
used to determine the weight (X) and length (Y)
relationship of grouper.
Independent Variable Dependent
Variable
Supplemental
Weight (X)
Feed (Fish
Length (Y)
Meal)

Figure 10.1. Independent and Dependent


Variables of Correlational Design.

You might also like