RESEARCH METHODS who classifies facts of any kind whatever, who sees

their mutual relation and describes their sequences, is

applying the Scientific Method and is a man of science
Lecture 3 – Research Design - The ideal of science is to achieve a systematic
interrelation of facts.
Research Methods
“this ideal by experimentation, observation, logical
- those methods/techniques that are used for
arguments from accepted postulates and a combination
conduction of research
of these three in varying proportions.”
- methods the researchers adopt to carry out their
- done through experimentation and survey
research
investigations which constitute the integral parts of
- blueprint for the collection, measurement and scientific method.
analysis of data
- identification or clarification of a problem
Research Methodology
- developing a hypothesis (testable theory)
- way to systematically solve the research problem inductively from observations

- science of studying how research is done - practical or theoretical testing of the hypothesis
scientifically
- rejecting or refining it in the light of the results
- not just know “how” to analyze but they also need to
Nature of Scientific Method
know which/what/why.

Three Groups of Research Methods

1. methods which are concerned with the
collection/ acquisition of data; these methods
will be used where the data already available
are not sufficient to arrive at the required
solution;
2. mathematical/statistical techniques which
are used for establishing relationships between
the data and the unknowns;
3. methods which are used to evaluate the Basic Postulates of Scientific Method
accuracy of the results obtained.
- It relies on empirical evidence;
Research Method or Technique
- It utilizes relevant concepts;
- refers to the behavior and instruments used in
- It is committed to only objective considerations;
selecting and constructing research technique
- It presupposes ethical neutrality, i.e., it aims at
nothing but making only adequate and correct
statements about population objects;
- It results into probabilistic predictions;
- Its methodology is made known to all concerned for
critical scrutiny are for use in testing the conclusions
through replication;
- It aims at formulating most general axioms or what
can be termed as scientific theories.
SCIENTIFIC METHOD (Karl Pearson) Assumptions that Underlie Scientific Method
“The scientific method is one and same in the branches Order - the universe is an ordered system that can
(of science) and that method is the method of all be investigated.
logically trained minds ... the unity of all sciences
consists alone in its methods, not its material; the man
External reality - we all share the same reality that Conceptual Framework (Creation of the
does not depend on our existence. author)
Reliability - we can rely on our senses and - process that involves mapping out or visualizing
reasoning to produce facts. these theoretical threads to form some diagrammatic
representation of inter-relatedness
Parsimony - the simpler the explanation the
better. - describes and explains the concepts to be used in
the study, their relationships with each other, and how
Generality - the ‘rules’ of reality discovered through they are to be measured
research
Main Steps to Create Framework
Research Process
1. Identifying the relevant concepts.
 formulating the research problem;
 extensive literature survey; 2. Defining those concepts.
 developing the hypothesis; 3. Operationalizing the concepts.
 preparing the research design;
 determining sample design; 4. Identifying any moderating or intervening variables.
 collecting the data; 5. Identifying the relationship between variables.
 execution of the project;
 analysis of data; Parts of Research Design
 hypothesis testing; Sampling Design - a suitable method for selecting
 generalizations and interpretation, and items to be observed for the study
 preparation of the report or presentation of the
results, i.e., formal write-up of conclusions Observation Design - conditions under which the
reached. observations are to be made

Framework Statistical Design - methods dealing with as how

many items are to be observed
- Basis of the research problem
Operational Design - deals with the techniques
- Serve as inductive approach towards the statement
of the problem Research design must, at least, contain
- Explains the phenomena upon which the thesis (a) a clear statement of the research problem;
investigation hopes to fill the vacuum in the stream of
knowledge. (b) procedures and techniques to be used for
gathering information;
- Organized, logical, and coherent upon which
problems of the study are based (c) the population to be studied; and
- Through the framework, one can immediately (d) methods to be used in processing and
perceive what the study is about analyzing data
- A reader may conclude at the end whether the Features of a Good Design
theoretical framework has been
- the design which minimizes bias maximizes the
 Fully supported reliability of the data collected
 Partially supported
 Negated by empirical data - the design which gives the smallest experimental
error
Theoretical Framework (Actual theory)
- A design which yields maximal information
- Thought of as a map or travel plan provides an opportunity for considering many different
aspects of a problem
- Important to consider relevant theory underpinning
the knowledge base of the phenomena under study
Features of a Good Design BASIC PRINCIPLES OF EXPERIMENTAL
DESIGNS
(i) the means of obtaining information;
(ii) the availability and skills of the researcher and his
Principle of Replication - experiment should be
repeated more than once.
staff, if any;
(iii) the objective of the problem to be studied; Principle of Randomization - provides
protection, when we conduct an experiment, against
(iv) the nature of the problem to be studied; and the effect of extraneous factors by randomization.
(v) the availability of time and money for the research Principle of Local Control - Under it the
work extraneous factor, the known source of variability, is
made to vary deliberately over as wide a range
Common Terminologies used in Research
Design Experimental Designs
Variable - quantity or characteristic that has two or - refers to the framework or structure of an
more mutually quantity exclusive values experiment and as such there are several experimental
designs.
Control - is used to establish a cause-and-effect
relationship Informal experimental designs - normally use a
less sophisticated form of analysis
Research Hypothesis - prediction statement that
relates an independent variable to the dependent  Before-and-after without control design
variable.  After-only with control design
 Before-and-after with control design
Treatments - different conditions under which
experimental and control groups usually exposed Formal experimental designs - more control and
use precise statistical procedures for analysis.
Experiment - process of examining the truth of a
statistical hypothesis  Completely randomized design (C.R. Design)
 Randomized block design (R.B. Design)
Categories of Research Design
 Latin square design (L.S. Design)
 Designs involving exploratory research studies  Factorial designs
 Descriptive and diagnostic type research
Conditions for Experimental Research
design studies
 Research studies involving hypothesis –  Conditions for Experimental Research (cause
testing occurs, the effect must follow)
 Covariation (the stronger the cause, the
Comparison of Exploratory and
stronger will be the effect)
Descriptive/Diagnostic
 The use of standard procedures (ISO, ASTM)

Considerations in Experimental Design

 Dependent and independent variables
 Sampling method and sampling bias
 Control experiment
 Repeatability and reproducibility
 Accuracy and precision
 Estimation of uncertainty
 Ethical issues
Quasi-Experimental Research Level of Measurement
- non-randomized intervention studies or non- - a function of the rules under which the numbers
experimental research are assigned

Potential errors: Nominal level of data

(1) specification errors that refer to cases when - into separate categories that can then be compared
relevant dependent variables are not identified nor with each other
included in the analysis; and
- every category is distinctive
(2) self-selection errors that arise because of the
- numerical in name only
nonrandom assignment

Survey Research Ordinal divides the data level of data

- data at situations
- Involves analyzing information obtained from
questionnaires or interviews - puts the data into order regarding a particular
property that they all share
- Information obtained from this research comprises
facts, opinions and behaviors - precise measurement of the property is not required
- Conducted through electronic surveys, mail surveys, - increases the range of statistical techniques
telephone interviews and face-to-face interviews
Interval level of data
- The success of a survey research depends on
sampling, measurement and overall survey design. - setting up inequalities wherein we can also form
differences
Survey Research
- the data must be able to be measured precisely
Types of sampling techniques:
- allows yet more sophisticated statistical analysis
 simple random sampling
 cluster sampling Ratio level
 stratified sampling - setting up inequalities wherein we can also form
 systematic sampling quotient
 multistage sampling
- most complete level of measurement, having a true
 convenience sampling
zero
 purposive sampling
- ratio scale can express values in terms of multiples of
Measurement errors: fractional parts, and the ratios are true ratios
Observation errors result from the way questions PARAMETRIC VS NON-PARAMETRIC
are posed, the characteristics of the survey subjects
SCALES
and the interviewers.
PARAMETRIC NON-
Non-observation errors result from
PARAMETRIC
unrepresentative samples, sampling errors and
data can be measured in data is measured in units
nonresponse.
units which are which are not
MEASUREMENT interchangeable interchangeable
can be subjected to most only to a limited extent
- the process of assigning numbers to objects or kinds to mathematical and
observations; process of mapping aspects of a domain of statistical and statistical treatment.
onto other aspects of a range according to some rule of mathematical processes
correspondence.
Simple Test
Measuring
- one value is different from another, you have a
- devising some form of scale in the range and then nominal scale
transform or map the properties of objects from the
- one value is bigger, better or more of anything than
domain onto this scale
another, you have an ordinal scale
- one value is so many units (degrees, inches) more or TYPES OF VARIABLES
less than another, you have an interval scale
Discrete Variables – one that can take on only a
- one value is so many times as big or bright or tall or finite or potentially countable set of values
heavy as another, you have a ratio scale.
Continuous Variables – one that can take on an
UNITS OF MEASUREMENT infinite set of values between any two levels of the
variables. Result of measurement
Nominal: categorize into boxes, names.
1. Measurement of Weights
Ordinal: prioritize according to relative values, put
into order.
2. Psychological Variables
3. Measurement of Heights
Interval: sort according to measured value.
Independent Variables – stimulus variable; which
Ratio: measure in relation to a zero value. is chosen by the researcher to determine its
relationship to an observed phenomena; called cause,
SOURCES OF ERROR IN MEASUREMENT object and variate of the investigation.
 Respondent Extraneous Variables – independent variables
 Situation that are not related to the purpose of study but may
 Measurer affect the dependent variable.
 Instrument
Dependent Variables – response variable, which
VARIABLES is observed and measured to determine the effect of the
independent variable; sometimes called effect, results,
Characteristics
criterion in the study.
1. It is an observable characteristics of a person
Moderate Variables – secondary or special type
or objects being studied
of independent variable chosen by the researcher to
2. It can assume several values representing a
ascertain if it alters or modifies the relationship
certain category
between the independent and dependent variable.
3. These are raw data or figures gathered by a
researcher for statistical purposes Control Variables – variable controlled by the
4. They are values that may arise from counting researcher in which the effects can be neutralized by
and or from measurement eliminating or removing the variable.
5. They are predicted values of one variable
based on another Intervening Variables – variable which interferes
with the independent and dependent variables, but
Determination its effects can either strengthen or weaken the
1. Look for the causes and effects independent and dependent variable.
2. Analyze the relations of one observation with SAMPLING TECHNIQUES
another observation
3. Describe the characteristics that are not the - method that allows researchers to infer information
same from all situations about a population based on results from a subset
of the population, without having to investigate every
CLASSIFICATION OF VARIABLES individual.
Quantitative Data Qualitative Data IMPORTANCE OF SAMPLING
Data that are obtained Data obtained from TECHNIQUES
from ordinal, interval, or nominal measurement
ratio measurements  Resource Constraints
refers to the numerical refers to the attributes or  Drawing inferences about the population
information gathered characteristics of the
about the sample; samples; CONSIDERATIONS OF SAMPLING
numerical data gathered indicates information TECHNIQUES
such as gender, attitude,
emotional condition.  Consistency
 Diversity
 Transparency
TYPES OF SAMPLING TECHNIQUES
PROBABILITY NON-
PROBABILITY
simple random sampling convenience sampling
cluster sampling quota sampling
stratified sampling judgement/purposive
sampling
systematic sampling snowball sampling