Research Methodology
Research Methodology
Characteristics of Research
Empirical: Research is based on direct experience/observation by the researcher.
Logical: Research is based on valid procedures and principles
Cyclical: It is a cyclical process because it starts with the problem and ends with the
problem.
Analytical: Research utilizes analytical procedures to gather data.
Critical: Research exhibits careful & precise judgement
Methodical: Research is conducted in a methodical manner without any bias
Replicability: The research design and process be repeated to arrive at valid and conclusive
result.
Objectives of Research
To gain familiarity with a phenomenon or to achieve new insights to it.
To portray accurately the characteristics of a particular individual situation or a
group.
To determine the frequency with which something occurs or with which it is
associated with something else.
To test a hypothesis or a causal relationship between variables.
Importance of Research
1. The main importance of research is to produce knowledge
2. Research helps in problem solving
3. Research is important to society as it allows us to discover more and more that
helps make lives easier
4. Research helps in rapid economic development
5. It provides basis for government policies
6. It is useful to students, professionals, philosophers etc.
Research problem to be undertaken for study must be carefully selected. A problem cam
spring in the mind of the researcher like a plant spring from a seed, A research guide can
help a researcher to choose the subject.
Research design is a master plan specifying methods and procedures for collecting and
analyzing data. Research design is a blueprint of action.
Broadly it is composed of elements like: - the purpose of study, the unit of analysis, time by
mention mode of observation, sampling, design, observational tools, data, processing, and
analysis.
1. Define the Problem: The first step involves the proper selection and then carefully
defining the problem. By this researcher will be enabled to know about what he has to
search.
2. Objective of the study: The objective should be very clear in the mind of the
researcher to get needed data accordingly.
3. Nature of the study: The research design should be very much in relation to the nature
of the study, which is to be carried out.
4. Data sources: The various sources of the data or the information should be very clearly
stated by the researcher be it primary or secondary sources.
5. Techniques of data collection: For the collection of the required data, it sometimes
becomes very necessary to use some special techniques.
6. Geographical limit: This step becomes a necessity at this point in time as, with the help
of this step, research linked to the hypothesis applies only to a certain number of social
groups.
7. Basis of selection: Selecting a proper sample, acts as a very important and critical step
and this is done with the help of some mechanics like drawing a random stratified,
deliberate, double cluster or quota sample, etc.
8. Data Analysis and Interpretation: Analysis of data is a process of inspecting,
cleaning, transforming, and modelling data with the goal of highlighting useful
information, suggesting conclusions, and supporting decision-making
9. Data Interpretation: Data interpretation can be defined as "the application of statistical
procedures to analyze specific observed or assumed facts from a particular study".
10. Conclusions and Recommendations: Conclusion means the judgement reached after
consideration. Based on the research findings, the conclusion needs to be drawn and
suitable recommendations should be made to help improve the research problem.
MEASUREMENT LEVELS/SCALES
The term scale of measurement is derived from two keywords in statistics, namely;
measurement and scale.
Measurement Scale
A measurement scale is used to qualify or quantify data variables in statistics. It determines
the kind of techniques to be used for statistical analysis.
1. Nominal Scale
The nominal scale is the least powerful level of measurement. This is applied to qualitative
data where objects are classified into various discrete/distinctive groups or categories
without any ranking or order associated with them.
Examples: Categorizing people according to their religion such as Hindu, Muslim, Sikh,
and Christian. Categorizing people as per blood group, eye colour.
Note: Here we assign different numbers to different objects the numbers do not have any
meaning. Ex: in research we categorise males as 1 and females as 2, this does not mean
males are better than females.
2. Ordinal Scale
This possesses the attribute of magnitude only. The various categories of items can be
compared with each other only in order of rank assigned to these categories.
3. Interval Scale
The interval scale is stronger than the ordinal scale because it possesses not only the
magnitude attribute but also the equal intervals attribute as it measures the values of the
quantitative random variables and identifies not only as to which category is greater or better
but also by "row much".
4. Ratio Scale
The ratio scale is also used for the measurement of quantitative random variables from the
interval scale in that it has a true (absolute) zero point, meaning that the values of each
variable can be zero.
Examples: Physical measurements such as height, weight, distance, etc.
SCALING TECHNIQUES
Scaling is an extension of concept of measurement. The difference between measurement
and scaling is that measurement is the assignment of numbers to objects or respondents
and scaling is the process of placing them on a continuum with respect to the number
scored by them.
A) RANKING SCALE:
Ranking scales offers respondent is to select most preferred item of their choice between
two options. Example a researcher may ask a lady to choose among two variants of
toothpaste.
Option A: Colgate
Option B: Pepsodent
However, there may be more than 2 options, in such case respondents could get confused
leading to ties between options. Problem can be solved by using following techniques.
Paired comparison Scaling: In this method, the objects are paired with each other and
presented to respondents who select the item they prefer. It can be anything such as
products, services etc. Example following table shows paired comparison of personal
choices among seven different fruits.
Constant Sum Scaling: In this technique, respondents are provided with a sum of units,
points, credits, money and asked to distribute/allocate them to objects.
Ex: A respondent may be asked to allocate 100% among various activities in a day like
eating, sleeping, working, hobby etc. in a way that total is 100%.
Q-sort scaling: It is a technique where participants are asked to sort objects or statements
into several categories. Ex: a respondent is presented with 50 statements and asked to sort
them into 11 piles where in extremes are highly agree and highly disagree.
RATING SCALES:
Rating scales are non-comparative in nature. They are also called as monadic or
metric scales. Here respondents estimate one item at a time and these scales are useful for
measuring attitudes, perception etc.
Ex: Weather can be rated in terms of humidity and illness can be rated in terms of severity
etc.
Rating scales are broadly of two types: Graphic and Itemized rating scales.
Itemized rating scale: Unlike graphic rating scale, itemized rating scales have a description
against every category of option. These categories are placed in certain order. The
respondents have to select a category that best describes the object.
They are of the following types:
a. Guttmann: In this scale few statements are arranged in accordance with their
specificity in an ascending manner. The respondents are asked to either agree or
disagree. The idea behind placing statements in increasing specificity is that,
respondents start to agree with statements up to a certain level after that they stop
agreeing. The scales determine the specificity of an attitude.
PARAMETRIC TEST
In statistics, the parametric test is a kind of hypothesis test, which case
generalizations for generating records regarding the meaning of primary/original
population.
Parametric test is one that makes assumptions about parameters (properties) of the
population distribution from which one’s data is drawn. In a parametric test, a sample
statistic is used to estimate population parameter.
POPULATION SAMPLE
NON-PARAMETRIC TEST
The non-parametric test does not require any population distribution, which is
meant by distinct parameters. It is a kind of hypothesis test which is not based on
underlining hypothesis. Here the test is based on differences in median. So, this kind of test
is also called distribution free test. The total variables are determined on nominal or ordinal
level. If the independent variables are non-metric, the non-parametric test is usually
performed.
t – TEST
o The objective of any statistical test is to determine likelihood of a value in a sample
given that null hypothesis is true.
o T-test is typically used in case of small samples and where statistic of population
follows a normal distribution.
o t- Test was originally developed by WS Gossett in 1908 to monitor the stout quality
while working in a brewery.
o t-Test is also known as student’s test.
o It is often used to test if “two samples are statistically different from each other.”
o The t-test, does this by comparing the means of both samples.
ASSUMPTIONS OF t-TEST
1. The population from which a sample is drawn is normal
2. The samples have been drawn at random
3. The population standard deviation is not known
4. Sample size should be small that is less than 30.
PROPERTIES OF t-TEST
1. The t-distribution curve is similar to normal distribution that is a bell-shaped
frequency called which is symmetrical in nature. Degrees of freedom is the only
parameter which determines shape of the curve. The shape of the curve changes
with degrees of freedom.
2. Similar to normal curve, the mean of t-distribution is zero.
3. The value of variance of t-distribution is usually greater than one and as the sample
size increases its shifts towards unity.
4. It can also be useful for large samples.
5. The value of the distribution ranges from -1 to +1
1. Test of significance of mean of a random sample: Here, the test is used to determine
whether sample Man drawn from normal population deviate significantly from the
population, mean, when the variance of population is unknown.
X−μ
t=
Formula: s
√N
3. Paired t-test
Th two samples are said to be to be dependent when, elements in one sample are
related to the other in some significant or meaningful manner. In fact, dependent sample
may consist of pairs of observation made on same person or object. It is helpful to pursue
experiments to find out impact of training on some employees, effectiveness of coaching
on students, effectiveness of medicine and treatment of patient, effect of advertisement on
boosting sales etc.
F-Test
o Another important parametric test is F-test which is used to test significance of
population variance for independent estimates.
o The term variance was first used by statistician R A Fisher.
o This test is particularly useful well when multiple sample cases are involved and the
data has been measured on interval or ratio scale.
o This test is useful in economics, business education, agriculture etc.
o The F test can be used to test the quality of variance of two normal populations,
analyze variance for more than two independent samples as well as analysis of
covariance.
F-test is used to test the quality of variance of two normal populations to find whether two
samples can be regarded as drawn from normal population having same variance.
ASSUMPTIONS OF F-TEST
o Normality: the population is normally distributed
o Homogeneity: the variance within age group should be equal for all groups.
o Randomness: the sample item should have been drawn randomly from the
population.
o Independence of error: this implies that variation of each item around the group
name should be independent for each item value.
F-test Formula:
CHI-SQAURE TEST
o Chi-square test statistic is commonly used for testing relationships between
categorical variables.
o The null hypothesis of chi-square test is that no relationship exists on categorical
variables in population that is they are independent.
o Alternate hypothesis states that there is relationship between the categorical
variables.
RESEARCH REPORT
A report is a detailed description of what has been done and how that been done with
respect to a particular area or topic. A research report is a presentation or research
findings in the form of report. It is a necessary part of the research process. It is the
oral or written presentation of evidence. Research report writing is the culmination of
the research investigation. Reporting is the end product of a research activity. (Indeed,
its practical application will follow.)
2) Table of Content
Table of content enlists various sections of research report along with their
corresponding page numbers. It includes various sections and subsections in a
sequence of their appearance in the report. It allows the readers to directly go to a
certain topic of their interest rather than searching the entire report for it.
5) Executive Summary
Executive summary is that part of the research report where the researcher writes a
brief summary of a complete research. It gives an overview about the research. This
section introduces them with the objectives and contents of the research report briefly,
which helps the decision making.
6) Introduction
The main section of the research report starts with the introduction. This part of the
research report describes the purpose, need, and importance of the research project. It
introduces the reader with the research problem and related questions. It also
highlights the previous related researches. It aware the readers about the problem in
broader perspective as well as the related concepts and terms that make familiar with
the background.
7) Methodology
In this section, a brief description about the complete methodology is given. It is kept
in mind that the researcher should avoid using technical jargons. This part should
neither be too detailed nor too brief. Research methodology includes the description
about following aspects:
i) Research Design: Research design illustrates the type of research design, i.e.,
exploratory, descriptive, and causal, chosen for the research work and reasons behind
the selection.
ii)Sampling: Here, the researcher describes about the nature of population, sample
frame, sampling units, sample size, sampling technique, etc., and reasons behind these
selections, so that the readers get familiar with the sample taken for research. iii) Data
Collection: Here, the researcher needs to mention the details about data collection in
the research work. In this section, the researcher is required to discuss whether he has
opted primary data sources or secondary sources, and the reasons behind the selection.
The researcher also explains all the possible methods with the help of which the data
has been collected, i.e., questionnaire, interview, observation, etc. If the questionnaire
is used in data collection, then a copy of it should be attached in the annexure of
report.
iv) Fieldwork: In this section, the complete procedure of fieldwork is discussed. The
researcher explains many things about the fieldwork, such as, the number of
fieldworkers employed, type of training provided to them, monitoring and controlling
of fieldwork, precautions followed during fieldwork, etc.
v) Data Analysis: In this section of research methodology, the researcher describes
the method adopted for analysing the data, and the suitability of selected method.
The complete description is given about the process applied for analysing the data
with proper calculations. Various charts and graphs can also be used to explain the
data easily. The formulae which are complex should be mentioned in the annexure.
Which formula is being used, when, and why, can also be mentioned as footnotes
along with the calculations.
vi) Findings: In findings, the researcher generalises or interprets the findings from the
data analysis. While making generalisations or interpretations, the researcher needs to
keep in the mind that findings should be presented in a logical manner, so that it can
help in decision-making process. The findings should be made in a way that can
answer the research questions. The findings should be represented effectively with the
help of various charts, graphs, and diagrams.
vii) Limitations of Research Project: Any research project is not completely correct
and may have some limitations. In this section of the report, the researcher highlights
the potential limitations that influence the reliability of the findings of a research.
These limitations can be related to anything such as errors in data collection,
sampling, data analysis, etc.
Researchers should always remember that they should not overstate these limitations,
as these may harm the positive effects of the research work.
9) Appendix
Appendix is the set of all those materials that are supplementary to the main research
work, such as, formulae, computations, questionnaire, statistical tables, etc. The
readers can refer to supplementary materials while reading a particular section for
better understanding