RESEARCH DESIGN
ARIES A. DE VOTA
Subject Teacher
Steps in Scientific Investigation
Identification of the Problem
◉ Identifying the problem guides the researcher or the
investigator in formulating his hypothesis, experimental
design, interpretation of findings, and to his conclusion.
◉ It must be clear, organized, and specific. It should
identify and specifies the variables to be measured and
observed.
◉ The goals or objectives of the research are set by the
statement of the problem.
Formulating a Hypothesis
◉ Formulating a hypothesis should be relevant to the problem.
It serves as a tentative solution to the problem.
◉ It guides the investigator or researcher to make his
experimental design or work plan.
◉ It can be tested through actual experimentation.
◉ A searcher or an investigator may form more than one
hypothesis from a piece of single information since there are
many possible answers or solutions to a problem.
◉ A hypothesis must be testable using observations or
experimentations done by the investigator or the researcher.
Review of Related Literature
◉ Reviewing relevant literature helps in formulating
creative procedures for solving the research
problem. Reading through various publications
and surfing the Internet can be done to become
aware of previous studies related to the research.
◉ This may also avoid duplicating other people’s
studies unless there is a need to validate the
findings with more advanced detection and
measuring techniques.
Preparation of Research Design
Designing research is necessary to specify the resources
needed for the study and indicates how the resources can
be utilized to answer the problem. A research proposal is a
more comprehensive form of the research work plan. It
contains the research design and the timetable for the actual
investigation and breakdown of expenses for the actual
experimentation.
Actual Experimentation
It is the application of the experimental design or
research work plan. An experiment is an activity
that starts with observations or data which, upon
processing and analysis, test the hypothesis
that results to the solution of the research
problem or generate further studies.
Data Analysis and Interpretation
Use tables and graphs in your gathered data before
they are analyzed using statistical methods. Through
research objectives, the statistical conclusions can be
explained. It is a simpler presentation of the meaning of
the statistical conclusion.
Formulation of a Conclusion
The significance of an experiment’s result is
summarizes by a conclusion. A conclusion is the ending of
an experiment or a research.
But in many cases a research also suggests future
questions left unanswered in the course of the investigation
that may lead to new researches. The reader can evaluate
the value of the complete research by knowing its
objectives and the conclusion.
What is a Research Design?
Research Design
It refers to the complete order of steps that you need to
be followed in your research or experiment.
REPLICATION
Your experiment should be repeated several times for you to
find an estimate of differences among observations on the group of
subjects or experimental units (the main material being used in the
experiment) treated alike.
It allows you to assess the significance of the observed
differences. Only through several repetitions of the experiment will
make a statistical test of significance possible.
RANDOMIZATION
◉ The second principle is randomization which refers to the assignment
of the experimental subjects to the treatments (denotes any procedure
done on the experimental units or subjects) by chance.
◉ It is done for you to create equivalent groups before your experiment.
It tends to average out differences among your groups.
◉ It refers to an unbiased method of assignment of experimental
units/subjects.
◉ It also assures a valid or unbiased estimate of population parameters
and the validity of the statistical test of significance in your research.
LOCAL CONTROL
◉ It is the process of reducing the experimental errors in your
research. It provides relatively different experimental areas into a
similar unit. It also increases the efficiency of your experimental
design.
◉ Two or more treatments should be compared. Your subject can be
blocked according to certain characteristics such as age, weight,
height, grade level, gender, fertility, pH gradient, and so on.
Key Features of a Local Control Group
3. Comparison of Outcomes
The primary purpose of the local control group is to
serve as a baseline against which the effects of the
treatment can be measured. If the treatment group shows
a significant improvement compared to the local control
group, it provides evidence that the intervention had a
positive effect.
Control of Extraneous Variables
It is necessary to make sure that no outside or
extraneous variables will affect your experimental
subjects. Make sure that identical conditions must
be provided for the control and experimental
groups. These two groups should differ only in
terms of the absence (control) or presence of the
given treatment in your research.
Types of Research Design
Completely Randomized Design (CRD)
◉ the objects or subjects are assigned to groups
completely in a random manner, it is applicable when
the set of the experimental unit is the same or
homogeneous for the factors that may affect the results.
In the figure below, there are a total of nine experimental
units, all of the same type. Any experimental units may
be assigned for each of the three treatments (T1, T2, T3).
Randomized Complete Block Design (RCBD).
◉ In this type, the subjects are divided into homogeneous groups called blocks. It
is used when the subjects have different characteristics that may affect the
results of the experiment.
◉ Blocking is an experimental design style that is used to avoid bias. Within each
block, the subjects or experimental units are assigned to a treatment randomly. It
gives more accurate results than a completely randomized design. There are no
limitations as to the number of treatments or number of blocks in the
experiments.
◉ The RCBD is most applicable to use if the subjects can be categorized according
to certain characteristics like gender, age, weight, and height which can affect
the comparison of treatments.
The given figure shows how the experimental units are divided into blocks.