Professional Documents
Culture Documents
1
SCOPE AND OBJECTIVE OF THE COURSE:
This course introduces the student to the scientific method of conducting research and will
provide a practical framework on which a student can base his/her research.
The course also introduces the student to various analytical instruments and equipment
commonly used in research laboratories as well as tools for data analysis. As part of the
course, students will work on
Textbook:
C. George Thomas “Research Methodology and Scientific Writing”, Ane Books Pvt. Ltd.,
2015.
2
Course Plan:
Chapter in the
Lecture No. Learning objectives Topics to be covered
Text Book
Need for academic research
Describe the elements of research and the Introduction to the scientific method
1-2 1
scientific method Research methodology and research methods
3
Course Plan:
4
Evaluation Scheme:
In academic fields: Defining, redefining and solving problems, observing facts and their
interpretation, formulation of hypothesis, and testing of the hypothesis through
experiments, revision of existing theories and laws, and practical application of information
already generated.
Methodology may refer to the rationale and the philosophical assumptions that
underlie a particular study relative to the scientific method.
7
Motives of Doing Research
External motivating factors:
For higher degree.
Easiest or the only way to earn a living/ Pursuit of Prestige.
Publishing papers is important for progressing in an academic career.
Obtaining patents is a good way to become rich and famous.
Peer pressure: my friends are all doing research and so should I.
Looking up to someone or trying to be like someone.
Surroundings is non-appreciative.
Inadequate infrastructure to carry out the work.
Non-availability of books, journals and sophisticated laboratories.
Peer-pressure/ family pressure.
Gap between expectations and effort.
Non-fulfillment of short-term objectives.
Getting stuck on a specific problem and being intellectually
drained out.
Lowering of self-esteem/having self doubt.
9
Logical Reasoning
Varun is older than Suraj.
Chetan is older than Varun.
Therefore Chetan is older than Suraj.
Method of Induction:
Inductive reasoning is the ability to reach general conclusion based on perceived
patterns observed in specific events or any experience. Inductive logic is often used
in everyday life. Specific → General (Bottoms-Up approach)
Method of Deduction:
Deductive reasoning involves a general rule or principle that leads to a specific
conclusion. General → Specific (Top-Down approach)
In such cases, the major premise is based on theory, rule, law, or general
understanding.
10
• Scientifically acceptable statements:
Conjecture, Hypothesis, Theory, Law, Fact, Axiom and Theorem, Model
1. Make observations
2. Develop a hypothesis
3. Prediction of results
4. Design of experiments to falsify the hypothesis
5. Conduct experiments and collect data
6. Evaluation and Conclusion
7. Accept, modify, or reject the hypothesis
11
• Research Methodology and Research Methods
Research Method: What we do in lab; methods and procedures we follow to accomplish
a research goal.
Research Methodology deals with general approaches; science and philosophy behind
research methods.
• Research Process
12
Approaches to Research:
• Intuition in Research
Be creative, open minded but always prepare, work hard, have a strategy and follow it.
• Unconventional ways to intuition: Serendipity, unexpected results, newcomers
Types of Research/Approaches
Basic vs. Applied Research
Retrospective or prospective
14
Descriptive and Analytical Research
Descriptive research includes surveys and fact-finding enquiries of different
kinds.
The major purpose of descriptive research is description of the state of affairs
as it exists at present.
Observation and description.
Ex post facto research.
In natural sciences: describe nature, anatomy, astronomy, taxonomic studies.
Social sciences: Census, surveys, analysis.
Knowledge base for generation of a hypothesis.
15
Quantitative and Qualitative Research
Quantitative research is based on the measurement of quantity or amount.
It is applicable to phenomena that can be expressed in terms of quantity.
Collection of numbers and analysis: Statistical data analysis.
Deduction approach.
Relationship between dependent and independent variables.
What, where and when of decision making.
The results are often presented in tables and graphs.
Physical sciences, economics, social sciences and biology.
Example: If one wishes to investigate why certain data are random then it is a qualitative
research. If the aim is to study how random the data is, what is the mean, variance and
distribution function then it becomes quantitative.
16
Conceptual and Empirical Research
Conceptual research is that related to some abstract ideas or theory.
It is generally used by philosophers and thinkers to develop new concepts or to
reinterpret existing ones.
Split down a theory to get better understanding of deeper issues.
Used in combination with other methods of research.
17
Experimental and Non-Experimental Research
Quantitative approach can be further sub-classified into experimental and Non-
Experimental research.
Introducing cause and studying the effect: Experimental approach
Independent variables are observed, introduced and manipulated to observe their effect
on other variables.
Experiments reveal ultimate truth.
19
Experimental Research
Experiments generate empirical knowledge.
“It doesn't matter how beautiful your theory is, it doesn't matter how
smart you are. If it doesn't agree with experiment, it's wrong.”
― Richard P. Feynman 20
Cause Effect Relationship
Select a cause variable, vary cause variable and measure the effect,
eliminating any other factors (hold other variables constant).
Falsification if hypothesis through experimentation.
Some Terminologies:
• Experimental plot/experimental unit: Experimental
area.
• Treatments: Objects under comparison, different
conditions under which experimental groups are put.
• Factors: Independent variables.
• Level: Intensity of a factor.
• Blocks: Categories within a treatment group.
21
Variability in experiments: Statistics
23
Problem: Is density of water different from density of solid
ice? How and Why?
24
Control Experiments
Standard Treatment or No treatment – Control Experiment or
Control group.
25
Errors in Experiments
Difference between truth and measurement.
Experimental Errors
Reproducibility
Possible errors:
27
Principles of Experimentation to reduce errors
28
Replication
To provide an estimate of experimental error.
Reduce the effect of uncontrolled variation (i.e., increase precision).
Quantify uncertainty. A related point: An estimate is of no value without some
statement of the uncertainty in the estimate.
• The number of replications (sample size) is the number of experimental units that
receive each treatment.
• The sample size should be small enough that negligible treatment differences are
not declared statistically significant and large enough that meaningful treatment
differences are declared statistically significant.
• Degrees of Freedom
• Repeated measurements on the same experimental unit may or may not constitute
true replications; treating dependent observations as if they were independent is
one of the most common statistical errors found in the scientific literature.
29
Randomization
Randomization Experimental subjects (“units”).
Avoid bias.
Normal Distribution.
Fisher’s Solution:
Replicate and
randomize to spread
variation evenly
At Rothamsted Agricultural Experiment
Station, Fisher recognized problems with some
among treatments.
32
of the agricultural experiments.
Basic Experimental Design
Fisher’s Solution
– Old Problematic Design: One large field receiving high nitrogen
(N), one large field receiving low nitrogen (N).
Fisher invents
• Basic principles of experimental design
– Replication, Randomization and Blocking
• Control of variation by randomization
• Analysis of variance
33
EXTRA
Fisher: Randomization, Replication & Blocking SLIDE
• No replication (or pseudoreplication) (pre-Fisher):
Field with
Field with
Low N
High N
For example, the probability that she would correctly identify all 4
cups is 1 in 70 .
Suppose the lady correctly identifies all 4 cups.
Conclusion: Either she has no ability, and has chosen the correct 4 cups purely by
chance, or she has the discriminatory ability she claims.
Since choosing correctly is highly unlikely in the first case (one chance in seventy),
we decide for the second.
Note: if she got 3 correct and 1 wrong, this would be evidence for her ability, but not
persuasive evidence since the chance of getting 3 or more correct is 17 70 = 0.2429.
Note: typically, a result is considered statistically significant if the probability of its occurrence
is less than 0.05, that is, less than 1 out of 20. 35
Dan Sloughter (Furman University)
Basic Experimental Design
Different Classifications
Based on applicability of randomization and manipulation of
treatments
• True Experiments
• Quasi Experiments
• ex post facto experiments
37
The Completely Randomized Design
• Each experimental unit is randomly assigned to one of two
treatments.
• Each unit has the same probability of getting any one
treatment
New
Treatment A to Standard
Experimental Treatment B to
Group Control Group
Unit 1 Unit 1
Unit 2 Unit 2
…
Unit n Unit n
38
The Completely Randomized Design
• All experimental units are considered the same and no division or grouping
among them exist.
All the variability among the experimental units goes into experimented
error.
CRD is used when the experimental material is homogeneous.
Select units out of n1 units randomly and apply treatment 1 to n1 to these units.
(Note: This is how the randomization principle is utilized is CRD.)
Select n2 units out of (n-n1) units randomly and apply treatment 2 to these units.
Continue with this procedure until all the treatments have been utilized.
Generally equal number of treatments are allocated to all the experimental units
unless no practical limitation dictates or some treatments are more variable or/and of
more interest.
40
The Randomized Block Design
In the R.B.D the principle of local control can be applied along with the other two principles
of experimental designs.
Block 1 … Block m
Individual 1 Individual 1
Treatment
…
1 …
Individual n Individual n
Individual n Individual
+1 n+1
Treatment
…
…
…
2
Individual 2n Individual 2n
41
If large number of treatments are to be compared, then large number of experimental
units are required. This will increase the variation among the responses and CRD may not
be appropriate to use. In such a case when the experimental material is not homogeneous
and there are v treatments to be compared, then it may be possible to
42
The Randomized Block Design
The RBD utilizes the principles of design - randomization, replication and local
control - in the following way:
1. Randomization: - Number the 𝜈 treatments 1,2,…, 𝜈.
- Number the units in each block as 1, 2,..., 𝜈.
- Randomly allocate the 𝜈 treatments to 𝜈 experimental units in
each block.
2. Replication:
Since each treatment is appearing in the each block, so every treatment will appear
in all the blocks. So each treatment can be considered as if replicated the number
of times as the number of blocks. Thus in RBD, the number of blocks and the
number of replications are same.
43
Latin square design
• The treatments in the RBD are randomly assigned to b blocks such that each treatment
must occur in each block rather than assigning them at random over the entire set of
experimental units as in the CRD. There are only two factors – block and treatment effects
– which are taken into account and the total number of experimental units needed for
complete replication are bv where b and v are the numbers of blocks and treatments
respectively.
• If experimental units are in two directions, ie two types (or if there are more extraneous
factors)
Say k units, then the total number of experimental units needed for a complete replication
are bvk. This increases the cost of experimentation and the required number of experimental
units over RBD.
In Latin square design (LSD), the experimental material is divided into rows and
columns, each having the same number of experimental units which is equal to
the number of treatments. The treatments are allocated to the rows and the
columns such that each treatment occurs once and only once in the each row and
in the each column.
44
Latin square design
• the treatments in a L.S. design are so allocated among the plots that no treatment occurs more than once
in any one row or any one column. The two blocking factors may be represented through rows and
columns (one through rows and the other through columns).
• The field is divided into as many blocks as there are varieties of fertilizers and then each
block is again divided into as many parts as there are varieties of fertilizers in such a way
that each of the fertilizer variety is used in each of the block (whether column-wise or
row-wise) only once.
Latin Square Design Example
Suppose different brands of petrol are to be compared with respect to the mileage per liter
achieved in motor cars.
Important factors responsible for the variation in the mileage are
- difference between individual cars.
- difference in the driving habits of drivers.
We have three factors – cars, drivers and petrol brands. Suppose we have
4 types of cars denoted as 1, 2, 3, 4.
- 4 drivers that are represented by as a, b, c, d.
- 4 brands of petrol are indicated by as A, B, C, D.
Or To choose such 16
experiments, we take the help of
Latin square.
47
Latin Square Design
In RBD, the experimental units are divided into homogeneous blocks according to the
blocking factor. Hence it eliminates the difference among blocks from the experimental
error.
In LSD, the experimental units are grouped according to two factors. Hence two
effects (like as two block effects) are removed from the experimental error.
So the error variance can be considerably reduced in LSD.
Limitation
• Although each row and each column represents equally all fertilizer varieties, there may
be considerable difference in the row and column means both up and across the field.
This, in other words, means that in L.S. design we must assume that there is no interaction
between treatments and blocking factors.
• Another limitation of this design is that it requires number of rows, columns and
treatments to be equal. This reduces the utility of this design.
48
Latin Square Design
In designing a LSD of order p,
• Choose one Latin square at random from the set of all possible Latin squares of order p.
• Select a standard latin square from the set of all standard Latin squares with equal
probability.
• For Latin squares of order less than 5, fix first row and then randomize rows and then
randomize columns. In Latin squares of order 5 or more, need not to fix even the first row.
Just randomize all rows and columns.
49
Randomization Procedures
R.A. Fisher on how to randomize an experiment with small
sample size and 5 treatments
• In a single-factor experiment, only one factor is studied. And the levels of the factor are
the treatments.
• When the number of factors involved in the experiment is more than one, it is known as
factorial experiment.
• In factorial experiments, combination of two or more levels of more than one factor is the
treatment.
• When all the possible treatments are studied, we call it a full factorial experiment.
• If number of factors and levels of each factor is large, its cumbersome to conduct
multiple number of experiments. In such situations, fractional factorial design is
adopted.
54