Sunteți pe pagina 1din 89

1

Presidency College
Bharatpur, Chitwan

Unit I: Introduction to Research and Project Work


1. Meaning of Research
How does the researcher proceed? Few social scientists have been interested in that problem
although every elementary textbook is full of advice regarding what the researcher should do:
bring a theory out of his hat, formulate hypothesis in advance, test reliability and validity of his
instruments, draw a representative sample, find something useful for his society report his
findings so that other researchers can replicate investigation.

The literal meaning of research is investigation or to search again. The first look may not be
sufficient and prone to error. The word research refers to the systematic investigation towards
increasing the sum of knowledge. The word is derives from the French word researche which
means to search or to seek again. Therefore research is a process of verifying and testing the old
facts or theories (deductive approaches) and discovering new facts (Inductive approach). So,
there are various meaning as:
A careful investigation or enquiry specially through search for new facts in any branch
of knowledge
Systematized effort to gain new knowledge
A research is systematic, controlled, and empirical and critically investigation of
hypothetical propositions (statements of relationship between two or more variables)
about the presumed relations among natural phenomenon. Fred N Kerlinger
Research is a systematic and organized effort to investigate a specific problem that
needs a solution (Sekaran, 1992).
As a whole research is the application of scientific method to the study of a problem. The
scientific method involves;
a. Careful logical analysis of the problem, and formulating hypothesis.
b. Formulating research design
c. Observation or data collection
d. Generalizations or drawing conclusions

Therefore, the process of investigation involves a series of well known activities gathering
information, recording, analyzing and interpreting the results with the objective of answering the
solution to the problem, is called research. The explanation of research has two crucial aspects.
They are:
1. Firstly, it is sufficiently broad to include all types of investigations requiring solution to the
problem.
2. Secondly, it explicitly recognizes the systematic nature of the research process in which
data are gathered, recorded, analyzed and interpreted in an orderly manner.
We can define research as an organized, systematic, data- based, critical, scientific inquiry or
investigation into a specific problem, undertaken with the objective of finding answers or
solutions to it (Sekaran, 1992).

For instance, AIDS broke out in the United States; doctors did not know at all what disease it
was, what cause it or how to cure it. They have only known as it seemed to attack the immune
system of victim. For the purpose of research, medical researchers set out to investigate the
causes of the disease and to seek a cure. At recent, a lot is known about this disease because
researchers have spent considerable resources in gathering information through observing
2

patients and conducting experiments. So, research means to investigate, repeatedly search or
quest in an Endeavour or resource to discover answers to problems through reliance on
empirical evidences. Research activities are designed to discover knowledge to aid in answering
specific questions or issues.

In general, the following activities are involved in the research process:

Define the problem and identify the variables causing it or related with it;
Get insight into the existing stock of knowledge on the issue or problem under reference and
contribute to the furtherance of knowledge (reviewing);
Construct models or formulate research questions educated conjectures or hypothesis for
testing;
Objectively collect and analyze the facts, evidences or information to identify the
interrelationship between or among variables or multiple effects of factors; and
Build theories to understand and generalize the phenomena to throw insight into the problem
under investigation to analyze, predict and control the situation

1.2 Type of research

Researches conducted for - solving -existing problem and generate a new knowledge to a
particular area or theory building.

1.2.1 Fundamental Research

Fundamental or basic research is conducted for the purpose of increasing our knowledge
of certain problems that commonly occur in organizational setting; and how to solve
them.
Adding to our knowledge i.e. fundamental and generalizable.
Pure or fundamental research.
Not apply the findings to solve an immediate problem at hand but rather than to
understand more about certain phenomena and problems that occur in several
organizations, and how they can be solved.
Theory building or formation.

For instance, On-the-job training has great impact on the productivity of workers.

1.2.2 Applied Research

AR conducted to a specific problem, which requires a solution.


Practical application of knowledge, systematically acquired and validated and meet
solution
Answer practical and useful questions about policies, programmes, projects, procedures,
or organization.
Such research, being of practical value, may relate to current activity or immediate
practical solutions.
Findings for immediate problem facing a society or business executives.
Hired researchers and consultants to study a problem of concern to them in order to find
solutions that can be implemented to rectify the problem situation and immediate
applications
3

It is also called decisional research.


E.g. Curative measures for AIDS.

For instance, How to improve the productivity of Cigarette Factory? Both improving the
existing brands or new brand development may be the study areas.
Factory does have the research, taking into accounts from its capabilities, know-how,
resources and so on.

Both fundamental and applied research may not be mutually exclusive; and can be
placed in a continuum.
Some research areas have both applied and fundamental research flavour and scientific
method provides key advances in both fundamental and applied research.

Table 1: Characteristics of applied and fundamental research


Fundamental Research Applied research
Purpose: Purpose:
Expand knowledge processes of Improve understanding of particular
business and management business or management problem
Results in universal principles relating to Results in solution to problem
the process and its relationship to New knowledge limited to problem
outcomes Findings of practical relevance and
Finding of significance and value to value to manager
society in general
Context: Context:
Undertaken by people based in Undertaken by people based in a
universities variety of settings including
Choice of topic and objectives organization
determined by the researcher Objectives negotiated with originator
Flexible time scales Tight time scales

1.3 Scientific Research Process

What is science?

Science as a systematized knowledge derived through observation, experimentation or any other


method or procedure in order to determine the nature of phenomena being studied or the causes
behind specific events.

In science, an understanding is made through logico-empirical process. The logico-empirical


process means reasoning and direct experience or observation. A scientific theory is dealing with
the logical aspect of science; and research is dealing with observation aspect of science.

The scientific process includes the methodology of understanding things. There are two
approaches of logico- empirical process.
a. Deductive Approach
b. Inductive Approach
1.3.1 Deductive Approach (General to particular)
4

In deductive reasoning, one starts from some general theories or laws and applies it to particular
instances. In this method, conclusions are deduced from some fundamental assumptions or
axioms established through other methods by logical process of reasoning. One starts with
theory as a means for many concepts, specifies indicators of measurement, makes observations
and tests relationships. The goal is to define or prove support the theory already known. In other
words, this approach involves understanding of thing from general to particular.
Hypothesis Observation Hypothesis testing Conclusion
Deductive process
Sociologist describes the theory as a scientific study between the causes and effects. He
tries to answer the question of what, when and how.
Sociologist constructs theories that have deductive structure in which conclusion from
assumptions.
Checking the validity of conclusion against the real world data.
When facts are inconsistent with the theoretical conclusion, refines theories or abandon
them for new theories.
Continuous interplay (interaction) of new or refine theories and their empirical
verification.
For instance, parents with children spent more over without kids. How? And Why?
1.3.2 Inductive Approach (Particular to general)
In this approach, one starts with observation driven by one or more research questions. In other
words, generalization is made from particular or one intends to develop the theory from general
to particular. This method is based on prior examinations of facts i.e. one starts with observed
data and develops a generalization which explain the relationship between the objects observed.
Observation Findings a pattern tentative conclusion Theory
Theory

Empirical generalization Hypothesis

Observations

INDUCTIVE PROCESS

Scientific research process

There are eight steps in scientific method. They are:

1. Sensing or realizing problems: It is the first step in scientific inquiry in observing the
situation and sensing the problem. New problem are emerging in the environment. We are
knowingly and unknowingly sensing these developments occurring in the environment. At this
stage, we may not know exactly what is happening but we can definitely sense that things are
not going on as smoothly as they should be.
5

Setting or
Observation Realizing Library search
problem

Problem
Identification
Theoretical
Framework

Deduction Hypothesis Formulation


Reasoning

Research Design

Data Collection

Analysis of Data

Refinement of Theory (Basic Research)


Implementation (Applied Research)

Figure 1: The Scientific Research Process

2. Problem identification: Once we increase our level of awareness


Characteristics of Scientific Research
1. Purposiveness
2. Testability
3. Replicability
4. Objectivity
5. Rigor
6. Generalizability
1.4 Management Research Methods:

Management of research is concerned with the systematic and objective collection and
evaluation of information about specific aspect of management problems in order to assist
managers make effective decisions.
6

The objective is to reduce the risk and uncertainty when strategy is in planned and
operation.
The role of sound information is to guide business decisions, set strategies and monitor
implementation to feedback on whether it has been successful or unsuccessful.

Managers need to ask the following questions to help to provide what research to carry out
to collect the required information and solve business problem. They are:

What information will help make decision- making?


What are we going to do with this information?
How should we collect this information?
What are we going to measure?
How should we analyze the results?
How much should we spend on collecting the information?

Research identifies the benefits of research knowledge to mangers as follows:

Identify and solve small problems in the work setting.


Know how to discriminate good from bad research.
Appreciate and constantly remember the multiple influences and multiple effects of factors
impinging on a situation.
Take calculated risks in decision- making, knowing full well the probabilities attached to
the possible outcomes.
Prevent possible vested interests from operating in a situation.
Relate to hired researchers and consultants more effectively.

Stages in Management Process and Types of Information Needed

Stages of Management Process Major Information Needed


1. Analysis
Performance against plans
Environmental, competitive and eternal
information to identify problems and
opportunities
2. Planning Evaluation of 1 and 2 in order to make a prediction or
estimate of alternative courses.
3. Execution Communicate details of plans and control standards.
4. Control Performance against plans

Type of Management Research

1.4.1 Policy Research


1.4.2 Managerial Research
1.4.3 Action Research
1.4.4 Evaluation Research

1.4.1 Policy Research


7

The development of foundation of information to be used as the basis for making plans and
decisions that will impact policy with in the context of political, social, and economical realities
of time and setting.

The policy research is composed of three basic elements:


This research studies how policy formulation occurs with a view to understanding and
improving the process.
The research is designed to analyze situations at the strategic level and to formulate overall
policy proposals.
This research systematically evaluates the priorities to be accorded to conflicting and
complementary alternatives (Formulating major policy proposals and establishing their
priorities).

1.4.2 Managerial Research

Related to the specific problem of limited scope for which management has need of
additional information on which base decision.
For instance, market potential for new product, best approach of implementation of a new
MIS system.
MR is seeking of solution to a given problem and implements the solution.
From the present analysis, MR is taking effective action for future.

1.4.2 Action Research


Involves continuous gather and analyzing of research data during the normal and on- going
operations of an organization.
Designed to identify effective way of dealing with problems in the real world.
Concern the execution of specific management programme, feeds simultaneous results to
organization; and helps in changing the mode of functioning.
AR is task- oriented form of study designed to provide continuous feedback regarding the
performance of management activity and to improve that performance from investigation.

1.4.3 Evaluation Research


The process of determining the value or worth of something is called evaluation.
It is oriented toward formal and objective measurement of extent which a given action,
activity and programme objective.
Evaluation Research related to policy research in which policies, objectives, strategies and
programmes are examined.
It evaluates the successful and analyzes the underlying causes of failure.

Methodology for Management Research


Objective/purpose Data requirement and analysis Research output
Policy Research
To formulate major policy Macro level data about: Policy options
proposals a) Environmental forces Identification of policy
To establish their b) Overall organizational situation priorities for the
priorities c) Competitive standing of the organization
To identify their organization
implication Longitudinal data
Time series data
a) Predictions (Econometric
8

model)
b) Projections (Parameters)
Managerial Research
To study the on-going Specific and detailed data about Identification of the
operation or projects the operations or projects problem situation
To help in improving Data collection through MIS Identification of
managerial Collection of qualitative data decision option
effectiveness Exploring the situation for in- Precise recommendation
To help in decision depth understanding. for actions to be taken
making a) Survey research methods
b) Observational methods
c) Experimental methods for
testing of alternatives
Action Research
To feed information into Continuous of gathering and To recommend the
the organization to analysis of data actions to be taken
improve the functioning Problem specific data: with regard to:
To improve the a) Opinion survey a) Organizational
managerial actions b) Observation structure
To suggest future c) Satisfaction levels b) Rules
courses of action d) Morale survey c) Reward and
e) Frustration index punishment system
d) Performance
evaluation criteria
e) Work environment.
Evaluation Research
To measure Micro level data Identification of
achievements against a) Programme utilization level programme strengths
the objectives b) Comparative level and weaknesses
To identify gaps and Internal surveys Recommendation for
problem areas programme
To suggest improvement
improvements required.

Attribute of good Management Research

Purpose clearly defined


Research process and work plan detailed
Research design thoroughly planned
High ethical standards applied
Limitations frankly revealed
Findings presented unambiguously
Conclusions justified
Researchers experience reflected.

1.5 Project Work

1.5.1 Meaning of Project Work


The Project is defined as an organized, systematic, and data based investigation into a specific
situation undertaken with the objective of gathering information that enables the students to gain
9

familiarity with the situation or reality and generate more knowledge about the phenomenon
under investigation.
1.5.2 Objectives of Project Work

Specific objectives are:


Expose students to business or social reality by providing them the opportunity to get first-
hand information and actual knowledge about the working of an organization, a system or
a situation.
Promote student- centred learning by encouraging them to take the initiate to become self-
directed learners and thinkers.
Provide opportunities to students to work on those issues or problems which are of
particular interest to them.
Develop interpersonal and communication skills by encouraging students to interact with
the practitioners.
Develop data- processing and report writing skills to explore, analyze and describe a
business situation or a social phenomenon

Processes in Project Work

Activities involved in a Project Work:


Selecting of relevant topic
Locating the sources of information
Extracting the relevant information from these sources
Identifying the various dimensions of the problem or issue
Organizing and analyzing the data effectively
Drawing conclusions
Writing report

Major activities involved in project work


Defining the Topic and research questions
The research topic should be introduce clearly raised the research questions which were not
answered before. Some potential problems in this process may be:
Inexperience resulting in difficulty deciding on an appropriate and feasible topic.
Over- ambitious scale (e.g. under- estimation of time, resources).
Re-inventing the wheel resulting from ignorance of previous work done.

Project Work design


This involves selecting a methodology for obtaining the relevant and their analysis to draw
inferences.
Both raw and secondary data (reviewing)

Data Collection: Some potential problems

Under- estimation or workload


Poor planning and coordination of interviews and field trips
Making contacts and appointments
Delays caused by waiting time
Determining scope, type and relevance of data to be collected
Unpredictability of data (lack of support)

Data analysis and report preparation


10

Some potential problems:


Misinterpretation (bias)
Being bogged down by details: inability to the woods for he trees
Insufficient care in documentation (quotations, bibliographic/ references)
Lack of discipline in structuring report framework before starting to write.

1.6 Method of Project Work


1.6.1 Exploratory or descriptive studies
1.6.2 Small- scale survey
1.6.3 Case studies, and
1.6.4 Feasibility studies

1.6.1 Exploratory or Descriptive Studies

Clear view of the subject under investigation


It is done when we do not know much about the situation
Comprehend the nature of problem and very few studies have been conducted to the
related phenomena
Approaches
Examining the existing literature
Questioning the knowledgeable individuals
Examining a few selected case

For example, demographic information of employees, a profile of suppliers, complaints


registered by the customers

Under descriptive studies


The demographic data of employees in term of their age, sex, education, family income,
etc.
The profile of suppliers: quality of supplies, rejections, regularity, demand for advance
payments, and so on.
The consumer complaints: frequency, nature of comments, causes, etc.
Descriptive studies present the data to meaningful form help to:

Understand the features and aspects in a situation under study,


Offer the ideas for further investigations and research, and
Help in identifying problems and make certain simple decisions.
For instances
a. The trend of employment during the last twelve months in an organization: Gender
composition, educational level, no. of absent days, ethnic groups, etc.
b. A profile of bank borrowers of outstanding of loans for six months and above: describe the
defaulter's age, earnings, employment, occupation and the like. You can recommend loan
payment issues for management decision to types of individuals for loan extending, etc.
c. Marketing Areas: pricing, sales, distribution method of company, etc.

Pricing methods, costing methods, pricing of related commodities, or products, price hikes
and sales
A profile of sales, sales by region, etc.
Types of distribution channels used.
Advertising and sales promotion expenses
11

Gathered information, describe and identify the company and help recommend a strategy for its
future action.

1.6.2 Small- Scale Survey

A survey study is the systematic gathering of information from respondents for the purpose
of understanding and predicting some aspects of the behavior of the population of interest.
Survey studies are usually more extensive and involve the careful identification of the
population, selection of the sample for the study and collection of comparable data to
make qualified generalizations.
For instance, survey study of the business describes location of business, employees, products,
prices, and markets, other competitive business, customers, etc.

Methods of survey

1. Telephone interviews- collection of information from respondents via telephone.


2. Mail surveys- collection of information from respondents via mail or similar other
techniques.
3. Personal interviews- collection of information in a face-to-face situation.
Home interviews- personal interviews in the respondents or office.
Intercept interviews- personal interview in a central location like shopping certres, haat
bazaar, etc.

Surveys can be done within organization and in the field.

1. Surveys conducted within organizations are as following types:


Surveys of employee attitudes
Surveys of the nature of jobs
Surveys of record system
Surveys of salary scales
Surveys of visitors to an organization
2. Field level surveys are as following types:
Surveys of consumer attitudes
Surveys of shop-keepers in a haat bazaar
Surveys of dealers and retailers
Surveys of patients visiting a medical clinic
Surveys of tourists visiting a place.

1.6.3 Case Studies

A case study is an intensive investigation and description of the study unit.


Appropriate method of project work
Both words and numbers description
Selected unit
Study different research methods
Collected related aspects

1. A case study may be done of a specific subject like records management system, sale
unit personnel department or organization.
12

For instances,
1.Records management system of an organization
Inputs procurement unit
Processing unit
Output supplying
Manpower
Control of records
Records inventory
Computer used
Data collected and presented in terms of efficiencies, effectiveness and future improvement in
the organization.

2. Section or department as a unit of analysis:

Account Department of an organization to analyze the accounting system:


Structure of unit
Work procedures
Accounting manuals
Revenue and expenditure policies
Work- load of staff
Format of the books of accounts
Audit report and so on.

Analysis of above facts, the case study highlights the strengths and weakness of
accounting system of the organization and recommend for improvements.

3. An organization as an unit of analysis

Organization goals
Policies
Structure
Manpower
Work system

These all areas are carefully studied to analyze the situation and identify the
managerial issues.
Both field cases based on real life and field data; and secondary data which is library based are
two types of cases. These two methods have their own features, relevance and methodology.

1.6.4 Feasibility Studies

Undertaken to assess the potential of a new scheme, idea and provide base for
investment decision.
Feasibility study define and analyze the critical elements such as:
operational aspects
Technical aspects
Marketing
Commercial base
Cost- benefit analyze together with alternatives scheme or ideas
Feasibility studies covers:
13

From simple, informal observation to complex, formal analysis


Complexity and formality- directly related to nature of project or scheme.
Feasibility is simple or complex, informal or formal, but careful planning is must.
In project work, student should decide scope of the study, time- frame, information
needs and methodology
For instance, Feasibility study within organization
1. Feasibility study of installing computer- based management information system in
organization
2. Feasibility study of opening a branch of a business in a location

Feasibility study outside the organization

1. An assessment of the feasibility of opening a stationary shop in a locality


2. Feasibility study starting a small- scale biscuit factory in a location
3. Feasibility study starting a gazal restaurant in certain location of a town
Following aspects should be studied and assess as a method of analysis:
1. Environmental factors
2. Market factors
3. Investment requirements
4. Technical aspects or issues
5. Cost- benefit estimates or analysis
14

Organizing and writing the Project Work report

1. INTRODUCTION
1.1 Background Information
1.2 Problem of the study
1.3 Objective of the study
1.4 Importance of the study
1.5 Limitation of the study
1.6 Organization of the study
2. REVIEW OF LITERATURE
2.1 Conceptual/Theoretical Framework
2.2 Review of related study
3. METHODOLOGY

3.1 Research design


3.2 Population and sample size
3.3 Sampling procedure
3.4 Data collection technique
3.5 Statistical procedure
3.6 Data analysis procedure
4. RESULTS AND DISCUSSION
4.1 Data presentation and analysis
4.2 Major findings of the study
5. SUMMARY, CONCLUSION AND RECOMMENDATIONS
5.1 Summary
5.2 Conclusion
5.3 Recommendation
References
Appendix
15

UNIT- II: PREPARATION AND SUBMISSION OF RESEARCH


PROPOSAL

Guidelines for developing a Project Concept Notes (PCNs)


Or Proposal Writing and Submission
Title:

The title of the research and development project must be brief and clear enough. It must be
specific and as concise as possible. The theme of the proposed study (problem and solution)
must be approximately reflected in the title, which must be informative enough to alert to the
reader and give him idea that the researcher proposes to do research and not otherwise.

Project Leader:
Name of Institution: Name and Address
Collaborators: Name and Institutions
Duration of Project:
Starting Date:
Completion Date:
Total Budget of Project:

1. Introduction
1.1 Background Information
Information includes a description of the importance of the researchable constraint that the
project is seeking to address and a very brief summary of any significant research or
development already carried out. The proposal should awareness of the problems in the context
of research and development already conducted or implemented. Particularly, an explanation of
the origin of the proposal and how it fits in the donors programmes or focus areas.
1.2 Statement of problems (or rationale of the study)
To convince the funding agency that ones proposal is worthwhile for undertaking, it would have
some rationale. It should state why the study is being proposed. It has been noted that social
scientists are crisis- oriented. They justify their proposal by citing the gravity of social problem
that has focused their attention. It is supposed to convince the people of the worth on ones
proposal. It is also important for the researcher himself to convince himself of the soundness of
his proposal to make him in the use of his time.
1.3 Scope and limitation of the study
The propose project should reflect the scope of findings by solving the prevailing problems of
the areas. It should be clear the coverage of the study areas; and what are the possible constraints
of studying limited areas either time or financial problems.
1.4 Objectives
The aim of the study or what is to expect of the study should be stated. The objectives such as
to solve the social problems of the country or sites or to attain self- sufficiency in rice or milk
or meat are presumptions. They should be narrowed down to attainable objectives under
reasonable conditions. The title of the proposal is often a broad statement of the objectives. They
should be made consistent.
1.5 Hypothesis
1.6 Almost every step in the development of scientific thought has been made by the
anticipation of nature, i.e. by the formulation of hypotheses which through verifiable,
often had very little foundation to start with. The first step in undertaking a piece of
scientific research should be the formulation of the problem as in definite or specific
terms as possible. Such a statement constitutes a trial or working hypotheses to be
tested and confirmed, modified or even abandoned as the investigation proceeds.
16

The trial or working hypothesis marks the beginning of a piece of research, theories, laws; and
systems as explanations in the form of generalizations come as final or end products. Where do
hypotheses come from? They are not received from without through sense of perception, but
made by the mind. Logicians generally recognize four criteria as the adequate measure of good
hypotheses.
The hypothesis should take into consideration all of the relevant facts and should
contradict none.
It should be plausible and in general, should not contradict any of the laws of nature.
The hypothesis should of such a character that it is amendable to deductive application
and testing i.e. it should be capable of disproof or verification.
It should be simple as possible for, from the beginning, science has demanded not only
accuracy and precision but simplicity.
Which of the hypothetical proposition is to be tested? Several hypotheses rather than one as a
rule are formulated in the course of the investigation. The least probable are rejected and the
most one confirmed. But null hypothesis (Ho) or alternative hypothesis (Ha) can be use equally.
2. Review of Literature
To be fully prepared for the study of any problem, it is supposed that the proponent knows well
that others have said or written about the problem; that he has sufficiently mastered the auxiliary
aid to research such as statistical tools; that he has sufficient experience in that research are to
enable him to do the job. In case where problem arise because the institution does not have up to
date library acquisitions, consulting colleagues who are knowledgeable on the problems helps.
The task becomes easier if we remember that other people one way or another could have
known something about the problem and often even know more than the researcher at the outset.
The review of literature shows that the proponent has researched or is knowledgeable in the area
being proposed. It also indicates findings on which the proposal is building on. Hence, a solid
foundation of background knowledge is imperative so that the researcher does not have to invent
the wheel all over again.
3. Outputs
The research results or products should be appropriate to the project purpose and include
identified promotion pathways to target institutions, agencies or beneficiaries. Target institutions
are those formal or informal institutions, which will take up the products of research and engage
in the process of transferring knowledge/ technology/ methodology to the beneficiaries. e.g.
FNCCI, INGOs, NGOs, etc. The beneficiaries are individuals and customers living in the
particular areas.
4. Research Methodology
How the study will be conducted to attain the desire objectives to be able to test the hypotheses
should occupy most of the researchers time. Do not dismiss the procedure by a few senseless
statements like sampling will be done data will be collected from secondary sources, data
will be analyzed and interpreted. Procedure must be discussed clearly and in sufficient detail to
be understood by prospective reviewers. Reviewers, seeing a well written, comprehensive, and
specific procedure, that the investigator is fully equipped and thoroughly prepared for his
project.
4.1 A complete procedure should consist of the following parts;

Theoretical framework

Sampling techniques and sample size.

Method of data collection

Method of data analysis
4.2 The basic considerations in determining the sample size are:
The degree of homogeneity or heterogeneity in the population
17

Degree of accuracy required


Resource at the disposal of the investigator, e.g. time, fund and manpower
4.3 Research activities
Research studies, surveys, experiments, etc. designed and implemented to achieve output of
project. There should be an activity of group of activities associated with the each output of the
project, the activities defining the action strategy for accomplishing each output. This section
also includes any facilities or expertise already available to the investigator or collaborator,
which will be utilized in the implementation of the project. In addition, the location of the
specific components of the research to be carried out and any special resources required to
implement of the project should be included.
5. Beneficiaries
It is necessary to indicate the main beneficiaries of the research. The beneficiaries are those who
gain social, economic, environmental advantage, etc. from the technology, methodology or
knowledge transfer activities of the target institutions. They may be identified for instance, the
households, the village community or the global community.
6. Budgetary Requirements
6.1 Equipments
The required all inputs must be mentioned and budgeted with respect to prevailing market
prices. The purchased of heavy equipment or using laboratory facilities of other institution and
the charges of lab facilities should included in the budget. Apart from this, these include
investment for construction of buildings, roads, bridges, irrigation facilities, and purchase of
land. Proposals of this nature require preparation of detailed development plans and are
generally not submitted as part of the research project.
6.2 Personnel requirement
The name of the existing personnel who will work on the project, their positions, salaries and
total man days required for the project must be indicated. Beside from salaries, the break down
of travel request, the following can be used as guidelines. They are place of visit, frequencies of
visit, number of people, tentative schedule, purpose of visit, mode of transport, travel allowance
per diem and daily allowances (TA/DA) and transportation cost.
6.3 Financial summary
Items Year 1 Year 2 Year 3 Total
1. Staff salary
2. Operational cost (TA/DA)
3. Consumable/ Photocopies
4. Equipment and Supplies
5. Training and publication
6. Overheads (1-5)
7. Contingencies (1-5)
Total
7. Risks and Assumptions
It includes those factors, which might contribute to the project failing to achieve its objectives.
Important assumptions are external conditions or a factor over which the project chooses not to
exert control or do not have control, but on which the accomplishment of the objectives depend.
8. Curriculum Vitae
The purpose of including the curriculum vitae is to determine the research capability of the
research leader and his accompanies. This will enable to research to convince the reviewer that
he can do.
9. References
All the reference materials reviewed should be listed.
18

Questions for Researcher or Development Project

1. Does the PCNs title fully reflect the proposed research project?
2. Is the justification fro the project clearly stated and convincing?
a. Will the project contribute significantly to either identifying or resolving the
farmers problems?
b. Does it fit into the World Trade Organization, Agriculture Perspective Plan, or
Tenth Development Plan?
c. Does the information quoted (literature or statistics) justify the proposed project?
d. Does the proposal consider?
i. Socio-economic aspects?
ii. Environmental effects?
iii. Gender issues?
iv. Sufficiently address the disadvantage groups
3. Is the proposed project technically and logically consistent?
Purposes Outputs Activities
a. Is the proposed of research project clearly stated?
b. Are the outputs logically framed to achieve the project purpose?
- Results of trials
c. Are the research activities technically framed to produce the stated outputs?
- Technical approaches
4. Are the beneficiaries clearly identified?
a. Does the proposed project specify the immediate end- users?
b. Does the proposal specify the ultimate beneficiaries?
5. Is the proposed budget rationally presented?
a. Does the budget show any obvious irregularities?
19

Guidelines for developing a Project Concept Notes (PCNs)

Title:
Does the PCNs title fully reflect the proposed research project?
Brief and clear enough, specific and as concise as possible.
Theme of the proposed study must be approximately reflected in the title (problem and
solution)
Must be informative enough to alert to the reader and give him idea that the researcher
proposes to do research and not otherwise.

Basic for front page

Project Leader: Name of Institution


Collaborators: Name and Institutions Duration of Project:
Starting Date: Completion Date:
Total Budget of Project:

1. INTRODUCTION
1.1 Background Information
Information includes a description of the importance of the researchable constraint that
the project is seeking to address and a very brief summary of any significant research or
development already carried out.
Proposal should awareness of the problems in the context of research and development
already conducted or implemented.
Particularly, an explanation of the origin of the proposal and how it fits in the donors
programmes or focus areas.
1.2 Statement of problems/Justification of the study
Is the justification fro the project clearly stated and convincing?
a. Will the project contribute significantly to either identifying or resolving the
farmers problems?
b. Does it fit into the APP or Tenth Development Plan?
c. Does the information quoted (literature or statistics) justify the proposed project?
d. Does the proposal consider?
i. Socio-economic aspects? ii. Environmental effects?
iii. Gender issues? iv. Disadvantage groups
To convince the funding agency that ones proposal is worthwhile for undertaking, it
would have some rationale.
Should state why the study is being proposed.
Justify their proposal by citing the gravity of social problem that has focused their
attention. To convince the people of the worth on ones proposal.
It is also important for the researcher himself to convince himself of the soundness of his
proposal to make him in the use of his time.

1.3 Scope and limitation of the study


Propose project should reflect the scope of findings by solving the prevailing problems
of the areas.
Should be clear the coverage of the study areas; and what are the possible constraints of
studying limited areas either time or financial problems.
20

1.4 Objectives
Aim of the study or what is to expect of the study should be stated.
Objective such as to solve the social problems of the country or sites or to attain self-
sufficiency in rice or milk or meat are presumptions.
Narrowed down to attainable objectives under reasonable conditions.
Title of the proposal is often a broad statement of the objectives- should be made
consistent.

1.5 Hypothesis
Hypothesis should take into consideration all relevant facts & avoid contradictions
Should be plausible & general, should not contradict any of the laws of nature.
Hypothesis should of such a character that it is amendable to deductive application and
testing i.e. capable of disproof or verification.
Should be simple as possible for, from the beginning, science has demanded not only
accuracy and precision but simplicity.

However try to answer the following issues:


Which of the hypothetical proposition is to be tested?
Several hypotheses rather than one in the course of the investigation.
Least probable are rejected and the most one confirmed.
Null hypothesis (Ho) or alternative hypothesis (Ha) can be use equally.

2. REVIEW OF LITERATURE
To be fully prepared for the study of any problems:
Proponent knows well that others have said or written about the problem;
Sufficient/auxiliary aid to research such as statistical tools; that he has sufficient
experience in that researchers are to enable him to do the job.
Institution does not have up to date library acquisitions, consulting colleagues who are
knowledgeable on the problems helps.
Other people one way or another could have known something about the problem and
often even know more than the researcher at the outset.(farmers)
Proponent has researched or is knowledgeable in the area being proposed.
Indicates findings on which the proposal is building on- a solid foundation of
background knowledge is imperative so that the researcher does not have to invent the
wheel all over again.

3. OUTPUTS/RESULTS/OUTCOMES/ FINDINGS
Is the proposed project technically and logically consistent?
Purposes Outputs Activities
a. Is the proposed of research project clearly stated?
b. Are the outputs logically framed to achieve the project purpose?
c. Are the research activities technically framed to produce the stated outputs?
Technical approaches
Research results or products should be appropriate to the project purpose
Include identified promotion pathways to target institutions, agencies or beneficiaries.
Target institutions are those formal or informal institutions, which will take up the
products of research and engage in the process of transferring knowledge/ technology/
methodology to the beneficiaries.
21

e.g. Department of Agriculture, Department of Livestock Services, INGOs, NGOs, etc.


Beneficiaries are individuals and farming households living in the particular areas.
4. RESEARCH METHODOLOGY
How the study will be conducted to attain the desire objectives to be able to test the
hypotheses should occupy most of the researchers time.
Do not dismiss the procedure by a few senseless statements like sampling will be done
data will be collected from secondary sources, data will be analyzed and interpreted.
Procedure must be discussed clearly and in sufficient detail to be understood by
prospective reviewers. Reviewers, seeing a well written, comprehensive, and specific
procedure, that the investigator is fully equipped and thoroughly prepared for his project.
4.1 A complete procedure should consist of the following parts;

Theoretical framework

Sampling techniques and sample size.

Method of data collection- trial or questionnaire

Method of data analysis- SPSS, R2
4.2 The basic considerations in determining the sample size are:
The degree of homogeneity or heterogeneity in the population
Degree of accuracy required
Resource at the disposal of the investigator, e.g. time, fund and manpower
4.3 Research activities
Research studies, surveys, experiments, etc. designed and implemented to achieve output
of project.
An activity of group of activities associated with the each output of the project, Activities
defining the action strategy for accomplishing each output.
Includes any facilities or expertise already available to the investigator or collaborator
roles
In addition, the location of the specific components of the research to be carried out and
any special resources required to implement of the project should be included.
5. BENEFICIARIES
Are the beneficiaries clearly identified?
a. Does the proposed project specify the immediate end- users?
b. Does the proposal specify the ultimate beneficiaries?
Indicate main beneficiaries- beneficiaries are those who gain social, economic,
environmental advantage, etc. from the technology, methodology or knowledge transfer
activities of the target institutions.
They may be identified for instance, the households, the village community or the global
community.
6. BUDGETARY REQUIREMENTS
Is the proposed budget rationally presented?
a. Does the budget show any obvious irregularities?
6.1 Equipments
Required all inputs must be mentioned and budget with market prices.(rate)
The purchased of heavy equipment or using laboratory facilities of other institution and
the charges of lab facilities should included in the budget.
Apart from this, these include investment for construction of buildings, roads, bridges,
irrigation facilities, and purchase of land.
Proposals of this nature require preparation of detailed development plans and are
generally not submitted as part of the research project.
22

6.2 Personnel requirement


Name of the existing personnel who will work on the project, their positions, salaries and
total man days.
Beside salaries, break down travel request and following guidelines.
Place of visit, frequencies of visit, number of people, tentative schedule, purpose of visit,
mode of transport, travel allowance per diem and daily allowances (TA/DA) and
transportation cost.

6.3 Financial summary


Items Year 1 Year 2 Year 3 Total
Staff salary (with inflation)
Operational cost (TA/DA)
Consumable/ Photocopies
Equipment and Supplies
Training and publication
Overheads (1-5)
Contingencies(1-5)
Total
7. Risks and Assumptions
Includes those factors which might contribute to the project failing to achieve its
objectives.
Important assumptions are external conditions or a factor over which the project chooses
not to exert control or do not have control (Killer assumptions)
On which the accomplishment of the objectives depend.
8. Curriculum Vitae
Curriculum vitae are to determine the research capability of the research leader and his
accompanies.
Will enable to research to convince the reviewer that he can do.
9. References
All the reference materials reviewed should be listed.
23

UNIT- III RESEARCH DESIGN

3.1 What is research Design?


What is the researcher want to know?
What has to be dealt with in order to obtain the required information?
3.2 Elements of a Research Design
The basic elements of a research design are:
i. the problem
ii. the methodology
iii. data gathering
iv. data analysis
v. report writing

Conceptualizing the Research Design


The research design has two purposes. The first purpose is to answer the research question or
test research hypothesis. The research hypothesis is introduced as tentative theory is to be
proved or disproved in the research study.

The second purpose of research design is to control variance. For instance, at the Faculty of
Management students, it was decided that there could be a high degree of merit in using the case
method for teaching management. It becomes necessary to test this hypothesis, i.e. that there is
significant difference in the achievement scores between students who were taught by the case
method and students who were taught by the lecture method.
The experiment was set as:

Before After
Test Results Test Results
(first day) (Last day)
Students with L1 L2 Control Group
lecture method
Students with C1 C2 Experiment Group
case method

Figure: An example of experimental research design


How variance is controlled by the research design.
1. Maximization of experimental variance: This means in the above example that it is
necessary to create as much difference between L 2 and C2 as possible. One must try to
maximize the difference of the dependent variables for the two classes creating as much
difference as possible. This is also done by making sure that the class using the case
method is really taught in a very different style.
2. Control of extraneous variables: It must be shown that the results of the experiment are
due entirely to the independent variable and not due to other factors or variables. The
higher results are obtained by the effect of teaching methods and not other factors like
intelligence, motivation, age etc. If one believes that the age is the important variable in
the results, the researcher should try to randomly distribute the age in both of the classes.
If it is found the higher result by higher age student, it is due to extraneous variable.
3. Minimize the error variance: It is difficult to control three variances. But, in the
students research he or she will have little control over some errors. If questionnaire or
some observational methods are used for controlling data, then errors can be minimized
by understanding and pre- testing the data collecting instruments. Probably the best one
24

can do in controlling the errors is to be familiar with his or her data collecting methods
before using them.

These three aspects of controlling variances have been put into one principle called the
Maxicon Principle and are stated as: Maximize the systematic variance understudy; control
extraneous systematic variance; and minimize error variance.
3.4 Types of Research Design
1. Historical
2. Descriptive
3. Developmental
4. Case Study
5. Correlated
6. Causal- Comparative, and
7. Action Research Design
1. Historical
Past phenomenon
Process of collecting, evaluating, verifying, synthesising past evidences systematically
and objectively to reach a conclusion.
Past events to present.
Characteristics
1. Historical Research (HR) depends upon data observed by other rather than investigator.
Good data result from painstaking work, which analyses the authenticity, accuracy, and
significance of source material.
2. Contrary to popular notions, historical research must be rigorous, systematic and
exhaustive; much claiming to be historical is an undisciplined collection of
inappropriate, unreliable, or bias information.
3. HR depends up on two kinds of data such as primary sources where the author was direct
observer of the recorded event and secondary sources where the author is reporting the
observations. Of the two, primary sources carry the authority of firsthand evidence and
priority in data collection.
4. Two basic criticisms weigh the value of data: The external criticism, which asks, Is the
document or relic (historic event) authentic? and internal criticism asks, If authentic,
are the data accurate and relevant? The critical evaluation of the data is what makes true
HR so rigorous in many ways, more demanding than experimental methods.
5. HR is similar to review of literature, which precedes other forms of research; the HR is
more exhaustive, and seeking out the information from larger array of sources. It track
down the more information, which were old and unpublished, and material not cited in
the references.
Steps:
1. Define the problem and asks this question yourself: Is the HR approach best
suited for these problems? Are pertinent data available? Will the findings be
educationally significant?
2. State the objectives and if possible, the hypotheses that will give direction and focus
to the research.
3. Collect the data: Both the primary and secondary should be collected separately. An
important HR is note taking. Note cards or small file cards (size- 3x5 or 4x6) are
used for taking information with one card for one topic for convenient and
rearranging.
4. Evaluate the data, applying both internal and external criticism.
25

5. Report the findings: including a statement of problem, review of literature,


important assumptions, basic hypotheses, methods, findings, interpretation,
conclusions and references.
2. Descriptive Research (DR)
Descriptive Research is process of accumulating facts does not explaining relationship, test
hypotheses, make predictions or make meaning and implication of a study. It is an extension of
an exploratory research.
Characteristics of DR
1. Literal meaning of describing situation or events.
The accumulation of a database that is solely descriptive and it does not seek or explain
relationship, test hypotheses, predictions and implication.
Research aimed it as more powerful purposes may incorporate DR.
It includes all forms of research except historical and experimental for broader context. It
also used survey method for widen descriptive method.
2. Purpose of Descriptive Studies:
To collect detail factual information that describe existing phenomenon
To identify the problems or justify current conditions and practices
To make comparisons and evaluations
To determine what others are doing in same conditions and benefits from their
experience in making future plan and decisions.
Steps:
1. Define objectives in clear and specific term. What facts and characteristics are to be
uncovered?
2. Design the approach:
How the data will be collected?
How will be the subject is selected to ensure they represent the population to be
described?
What instruments or techniques available or to be developed?
Will the data collection methods need to be field- tested and will data gathers need to be
trained?
Collect the data
Report the results

3. Developmental Research
The Developmental Research focus on variables, their rate of change, directions, sequences and
other inter- related factors over a period of time. Particularly, developmental research is done for
the purpose of future trends.
Methods of Developmental Research
a. Longitudinal growth study
b. Cross- sectional growth study
c. Trend study
d. A Cohort Study
e. A panel Study
a. Longitudinal growth study
This study measures the nature and rate of change in a sample at different stages of
development. Data were collect two or more different point of time from same groups of
individuals. This method is not a cross- sectional or one short studies because data are
collected two different points of time. So, it is the study carried longitudinal across the
period of time.
26

b. Cross- sectional growth study


This measures the rate of change by drawing sample from a cross- section of the society for
comparing and describing the groups. Data are gathered at once over a period of time in
order answer the research question. It is also called as one- shot studies.
c. Trend study
This study is designed to establish patterns of change in the past in order to predict future
patterns or conditions. Data are collected at intervals spread over a period of time and
called trend study. These studies are used to obtain and analyse social, economic, and
political data to identify trends and to predict what takes place in the future. Regression
analysis is frequently used for trend studies.
d. A Cohort Study
It is the study of specific group such as those born in the same period. A sample of sub-
group is studied at different points of time. To form cohort studies, data are compiled for
the same population over time. Such studies are rare due to difficulty of maintaining
contact with the members of the cohort from year to year.
e. A panel Study
It is the best method of research for change. A panel is group of individuals who are
agreed to provide information to researcher over a period of time. For this purpose of
study, we take same people and study their attitudes towards a particular phenomenon
over time. For instance single parenthood, we should take a group of people and interview
them at periodic intervals on the same subject over a number of years.
Panel study
Longitudinal study

Developmental studies Cohort Study

Cross- sectional study Sample survey

Figure: Form of developmental study

Characteristics of Developmental Research


1. Developmental Research focuses on the study of variables and their development over a
period of times or years and asks, What are the patterns of growth, their rates, their
direction, their sequences and the interrelated factors affecting these characteristics?
2. The sampling problem in the longitudinal method is complicated by the limited number of
subjects it can follow over the years; any selected factor affecting the attrition (slow
destruction) biases the longitudinal study. It can be avoided by sampling stable population
but it introduces unknown biases to such population.
3. Cross- sectional studies usually use more subjects, but describe fewer growth factors than
longitudinal studies. The longitudinal studies is the only direct method of studying human
development, the cross- sectional is less expensive and faster since the actual passage of time
is eliminated by sampling different subjects across age ranges such as children.
4. Trend studies are vulnerable unpredictable factors that modify or invalidate trends based on
the past. In general, long- range prediction is an educated o experience guess while short
range prediction is more reliable and valid.
Steps:
1. Define the problems or state objectives
27

2. Review of literature for baseline information and comparing the research methodologies
for data collection
3. Design the approach
4. Collect the data
5. Evaluate the data and report the results

4. Case Study
In case study, an investigator makes and intensive of a social unit i.e. a person, a family, a social
group, an institution, a community, etc. The investigator collects present status, past experiences
and environmental forces that contributes to the individuality and behaviour of the unit. After
analyzing the sequence and inter- relationship of these facts, the researcher conducts a
comprehensive study of the social unit as it functions in society. These studies make useful
contribution to research as:
A case study is more expensive because of it exploratory in nature.
A generalization drawn from a single case cannot apply to all cases in a give population.
There are some elements of subjectivity. The researcher tries to control the personal
biases and standards to influence his or her interpretation.
Characteristics of case study
Case studies are in- depth studies of a given social unit resulting in a complete and well
organized picture of that unit. It cover the entire life cycle of a given segment and it
concentrates to a specific factors or events to a given elements.
As compared to survey study which tends to examine a small number of variables across
a large sample of unit, a case study tends to examine a small number of units across a
large number of variables and conditions.
Strengths:
It develops background information for planning major investigation in the social
sciences. Case studies are intensive; bring important variables, processes and interactions
that deserve extensive attention for study. It pioneers the new ground and often the
sources of fruitful hypotheses for further study.
Case studies prove useful examples to illustrate generalised statistical findings.
Weakness:
Case studies are limited in representativeness and do generalized findings until the
appropriate follow- up research is accomplished, focusing on specific hypotheses and
using proper sampling methods.
Case studies are vulnerable to subjective biases. The selective judgement rule in or out or
high or low value of significance, subjective interpretation is influencing the outcomes.
Steps:
State the objectives. What is the unit of study, characteristics, relationships, and
processes will direct the investigation?
Design the approach. Unit selection, sources of data and data collection should decide.
Collect the data.
Organize the information to form the coherent, well- integrated reconstruction of the unit
of study.
Report the results and discuss their significance
5. Correlational Research
The correlational research is used to ascertain the extent to which two variables are related, in
correlational relationship, change in one variable accompany changes in another or exist the
relashionship between them, such variables are said to be covary.
Use to obtain description of phenomenon
28

Two variables are related


Changes in one variable accompany changes to another but actual influences is difficult
to identify
These variable are called covary or co- variables.
Three types of correlation
Positive correlation, eg. Benefits of workers and productivity
Negative correlation, inverse relation, eg. Absence rate of employees result in decrease in
production
No correlation
Correlation range
+ Perfect positive
- Perfect negative
0.0 No correlation
Characteristics
1. Appropriate where variables are complex and/or do not lend to experimental and
controlled manipulation.
2. Permits the measurement of several variables and their relationship simultaneously and
in realistic setting.
3. Get at the degree of relationship rather than the all- or- nothing question posed by
experimental design: Is an effect present or absent?
4. Among its limitations are the following:
It only identifies what goes with what- it does not necessarily identify cause- and effect
relationships.
It is less rigorous than the experimental approach because it exercises less control over
the independent variables.
It is prone to identify spurious relational patterns or elements, which have little or no
reliability or validity.
The relational patterns are often arbitrary and ambiguous.
It encourages a shot-gun approach to research, indiscriminately throwing in data from
miscellaneous sources and defying any meaningful useful interpretation.
Steps:
1. Define the problem
2. Review the literature.
3. Design the approach:
Identify the relevant variables.
Select the appropriate subjects.
Select or develop appropriate measuring instruments.
Select the correlational approach that fits the problem.
4. Collect the data.
5. Analyze and interpret the results.
6. Causal- comparative research
Studies that establish causal relationship between variables and also called explanatory
studies.
Research investigates the possible causes affecting the results.
Explaining the relationship between two variables.
Known as ex post facto research.
Kerlinger states:
Ex post facto research is that research in which the independent variable or variables
have already occurred and in which the researcher starts with the observation of a dependent
variable or variables. He then studies the independent variables in retrospect for their possible
relations to, and effect on, the dependent variable or variables.
29

Three possible relationships between two variables


Symmetrical Relationship: Two variables fluctuate together.
Reciprocal Relationship: Such relationship exists when two variables mutually influence
or reinforce each other. This could occur if the reading of an advertisement leads to the
use of a brand of product.
Asymmetrical Relationship: This relation postulates that changes in one variable
(Independent Variable) are responsible for change in another variable (dependent
variable).
Characteristics:
Causal- comparative research is ex post facto in nature, which means the data are collected
after the research has been already occurred. The researcher takes one or more effects and
examines the data going back time, seeking out causes, relationships, and their meanings.
Strengths:
1. The causal- comparative method is appropriate where the more powerful experimental
method is not possible:
When it is not always possible to select, control and manipulate the factors necessary to
study cause- and effect relations directly.
When the control of all variations except a single independent variable may be highly
unrealistic and artificial, preventing the normal interaction with other influential
variables.
When laboratory controls for many research purposes would be impractical, costly or
ethically questionable.
2. It yields useful information concerning the nature of phenomena: what goes with what,
under what conditions, in what sequences and patterns, and the like.
3. Improvements in techniques, statistical methods and design with partial control features.
Weaknesses:
1. The main weakness of any ex post facto design is the lack of control over independent
variables. To reach sound conclusion, the researcher considers the possible reasons or
rival hypotheses for the results.
2. The relevant causative factor is actually included among the many factors under study.
3. No single factor is causing the outcome but the combination and interaction of factors go
together under certain conditions to yield a given outcome.
4. No multiple causes: One cause in one instance and another cause in another instance.
5. Ambiguous cause and effect (May reverse).
6. The additional factor may not discover.
Steps:
1. Define the problem.
2. Survey the literature.
3. State the hypotheses
4. List the assumptions upon which the hypotheses and procedures will be based.
5. Design the approach:
Select the appropriate subjects and sources materials.
Select the techniques for collecting data.
Classify the data categorically which are unambiguous, appropriate to show significant
relationship.
6. Validate the data gathering techniques.
7. Describe, analyze, and interpret the findings in clear and precise terms.
7. Action Research
Working definition of action research:
If yours is a situation of action research
30

People reflect and improve (or develop) their work and their own situations
By tightly interlinking their reflection and action
And also making their experience public not only to other participants but also to other
persons interested in and concerned about the work and the situation
And yours is situation in which there is increasingly
Data gathering by participant themselves in relation to their own questions
Participants (in problem posing and in answering questions) in decision- making
Power- sharing and the relative suspension of hierarchical ways of working towards
industrial democracy
Collaboration among members of the group as a critical community
Self- reflection, self- evaluation and self- management by autonomous and responsible
persons and groups
Learning progressively (and publicly) by doing and making mistakes in a self-
reflective spiral of planning, acting, observing, reflecting, replanning, etc.
Reflection which supports the ideas of the (self-) reflective practitioners
Then
Yours is a situation in which ACTION RESEARCH is occurring
The working definition is jointly authored by the participants at the International Symposium on
Action Research, Brisbone, 1989. We would suggest that this working definition was acceptable
to the full range of participants because it is:
Not too threatening to existing understandings and practices
Not too vague so that everything was included, but
Rather rich in examples which might support the development of shared meanings
Open enough so that further elaboration and development seemed possible
Allowing for an ex post facto incorporation of projects into the discussion, and
Above all, shared with respect to the process of it formulation for a specific content.
31
32

UNIT IV: MEASUREMENT AND SCALING

4.1 Types of variables: Dependent, Independent and Intervening Variables


4.2 Scales of Measurement: Nominal, Ordinal, Interval and Ratio Scale
4.3 The construction of attitudes and social scales (Likert- Type, Thurstone and Guttaman
Scaling)
4.4 Reliability and validity of measurements.

Variable
Variable are characteristics of persons, things, groups, objects, etc.
Symbol to which numerals or values assigned. E.g age, sex, jobs, attitudes, etc.
A variable is defined as anything that can take on differing or varying vales.
Examples, age, sex, productivity
4.1 Types of variables
Dependent variable
Independent variable
Intervening variable
Moderating variable
1. Dependent variable
Values depend upon other variables
Investigators purpose is to study, analyze and predict the variability in the dependent
variable.
What result in dependent, if certain changes appear in other related variables?
Measuring the variability is the interest of investigator
Variable used to describe or measure the problem under study is called dependent
variables.
Examples,
Production manager- employees productivity which vary low, medium and high
Marketing manager- consumer complaints or dissatisfaction of products, variance in
consumer complaints is dependent variables.
Personnel officer- training will increase the production of workers, production-
dependent variable and training- Independent variable
2. Independent variable
If it is not influenced by other variable under study but influence to dependent variable.
Any change independent variable (+ve or ve), change in dependent variable (increase or
decrease).
Basis for prediction- dependent variable is being predicted
Independent variables are used to describe or measure the influence the problem.
How change in independent variable affects to the value of dependent variable.
Examples,
Marketing manager is interested to increase the sales by advertising: Advertising
(independent variable) leads to volume of sales (dependent variable)
Annual saving of employees is the function annual earnings

Agricultural output is function of inputs (fertilizers and irrigation


3. Intervening variable
There should relationship between dependent variable and independent variable.
A third variable influences the nature and degree of relationship between independent
variable and dependent variable such variable is called intervening variable.
Give the true picture of cause and effect.
33

Example

Challenging jobs Motivation Job Performance

Independent variable Intervening variable Dependent variable

Figure: Showing the relationship between dependent and independent variable with the presence of
intervening variable
4. Moderating variable
Strong contingent effect on the dependent- independent variable relationship.
Believed as second independent variable to significant effect to originally expected
relationship.
Third or moderating variable modify the above relationship.
Example,

Training Productivity

Independent variable Dependent variable

Workers age over 50 vs below 50


years

Moderating variable

Figure: Showing the moderating variable

Wage Job
satisfaction

Dependent variable
Independent variable

Extra income

Moderating variable

Figure: Showing the moderating relationship


4.2 Scales of Measurement
1. Nominal
2. Ordinal
3. Interval
4. Ratio Scale
34

1. Nominal Scale
To assign certain subjects, group or categories
Is the simplest and lowest form recording the data, and
Provide the very basic information of personnel data.
Examples:
Gender: Male, female
Religion: Hindu, Buddhist, etc
Occupation: Teacher, student, agriculture, business
Nationality: Nepali, Indian, etc
Department: Sales, account, production
2. Ordinal Scale
A scale is ordinal when object can be assigned in order on some characteristics but they
cannot be assigned values that represent the degree of difference on that characteristic.
Ranking in order likes drinks- coca- cola, pepsi, fanta, etc.
Social status of jobs:
Jobs Rank
Doctor
Engineer .
Professor
Lawyer ..
3. Interval Scale
Scale assumes that the data have equal intervals
Example: Behavioural research- measurement of attitudes and certain psychological
characteristics such as intelligence and learning

Possible answer Interval scale


Strongly agree 1
Agree 2
I dont agree 3
Disagree 4
Strongly disagree 5
4. Ratio Scale
To indicate ratio
Ratio indicates not only the magnitude of the differences but also the proportion of the
differences.
Responses could be in range from 0 to any figure
Example: 10 years old boy is the twice of the 5 years old boy, other ratio scales are age, income,
working hours, etc.
What is your annual income?
How many children do you have?
How many workers are the members of the union?
Changing the scale format

Ordinal Interval Ratio


What is your income? What is your income? What is your income?
Very low Less than Rs. 1000
Low 1000 to 2000
Average 2001 to 3000 My annual income is Rs.
High 3001 to 4000
Very High More than 4000
35

4.3 The Construction of attitudes and ratio scales

Feelings, perceptions and behaviours towards other things, people, places, etc.
Feelings and perceptions are called attitudes and measured by attitude scales.
There are procedures of developing measures.
Specify the
construct
Generate sample
of items
Collect
Data
Purify
measure
Figure: Procedure fro developing measure Assess
validity

Scales of attitude measurement


1. Arbitrary scales
2. Likert- Type Scales (Summated Rating Scale)
3. Thurstone Scales (Equal- Appearing Interval Scales)
4. Guttman Scales (Cumulative Scales)
1. Arbitrary scales
Ad hoc basis or arbitrary
Researchers subjective selection of items
Examples: How do you assess the reputation of this college? (Excellent facilities,
learning (Bad,---------, Good)
2. Likert- Type Scales (Summated Rating Scale)
Self- report technique for attitude measurement in which the subjects are asked their
degree of agreement and disagreement with each of statements.
The summing score is the total attitude score, is the Likert scale. Example,
Most people can be trusted.
Strongly agree5_ Agree4_ Undecided3_ Disagree2_ Strongly disagree1
Any body select 1 to 5
Take other similar items- to find out general attitude of indication
Add the total score- the top 25% - the most favourable attitude and bottom 25%- the least
favourable attitude toward the topic being studied;

Guide for the construction of the Likert Scale


Nature:
This is summated scale consisting of a series of to which the subject responds.
Respondent indicate agreement or disagreement- in intensity
Likert technique produces an ordinal scale, which requires non- parametric statistics for
analysis.
Utility
This scale is reliable when it roughly ordering of peoples particular attitude or attitude
complex.
The score includes a measure of intensity as expressed on each statement.
Construction:
The investigator collect a large number of items considered relevant to the attitude
being investigated and either clearly favourable or unfavourable.
36

These items are administered to a group of representative whom are you asking the
question.
The responses to the various items are scored in such a way that a response indicative of
the most favourable attitude is giventhe highest score.
Each individuals total score is computed by adding his or her item score.
The responses are analyzed which items differentiate most clearly between the highest
and the lowest qualities of total score.
The items which differentiate best (at least 6) are used to form a scale.
3. Thurstone Scale (Equal- Appearing interval Scale)
Scaling technique requires item to be selected by a panel judges.
The items are evaluated on:
Relevance to the topic area
Potential for ambiguity
Level of attitude they represent
Developed by LL Thurstone and called equal appearing interval scale
Indicating agreement or disagreement
All the items on the objective test are equal value.
Actually some items are given higher value than other items.
Example,
My job is like a hobby to me. (10.1)
I am satisfied with my job for the time being. (7.2)
I am often bored with my job. (3.2)
Most of the time I have to force myself to got to work. (1.1)
A guide for the construction and use of Thurstone of Equal appearing Interval Scale
1. Nature:
Items of scale are determined by ranking operation done by judges.
Responses which best describe how he or she feels
2. Utility
Scale measurement is interval
The distance between any two numbers of scale is of known- size.
Parametric and non- parametric statistics may be applied.
3. Construction
Several hundred statements conceived to be related to the attitude being
investigated.
A large number of judges (50- 300) independently classify statement to eleven
groups ranging from most favorable to neutral to least favorable
The scale value of a statement is computed as the median position to which it is
assigned by the group of judges.
Statements which have too broad a spread are discarded as ambiguous or
irrelevant.
The scale is formed by selecting items which are evenly spread along the scale
from one extreme to the other.
4. Guttman Scales (Cumulative Scales)
The cumulative scales like the other scales consist of set of items which are answered.
The Guttman scale tend to be cumulative, that is if a person can answer the first item he
or she probably will be able to answer the second or visa versa.
Example: Four children were given three arithmetic items:
a. 28/7 = ?
b. 8 x 4 = ?
c. 12 + 9 = ?
37

a b c Total score
First child 1 1 1 3
Second child 0 1 1 2
Third child 0 0 1 1
Fourth child 0 0 0 0

A guide for the construction and use of Guttman Cumulative Scale


1. Nature:
It attempts to unidimensionality of scale.
Items of scale are determined by ranking operation done by judges.
Responses which best describe how he or she feels
2. Utility
Scale measurement is interval
The distance between any two numbers of scales is of known- size.
Parametric and non- parametric statistics may be applied.
3. Construction
Several hundred statements conceived to be related to the attitude being
investigated.
A large number of judges (50- 300) independently classify statement to eleven
groups ranging from most favourable to neutral to least favourable
4.4 Reliability and validity of measurements.
Validity:
Data are considered to be valid when they measure what they are supposed to measure.
How? Validity generally results from careful planning of questionnaire.
Cautious wording
Preliminary testing of items to detect misunderstanding
Reliability:
Data are considered to be reliable when they give assurance that they are reasonably
close to the truth.
Reliability results from taking larger sample of respondents. Suppose, if 10 respondents,
results might vary, if 90 respondents added, results reached to stable and not change.
Reliability would then be assured.
Again survey more people, would give same result. Whereas, validity on the other hand
results from careful planning of questionnaire

4.4.1 VALIDITY
Validity is concern with systematic error. A valid measurement reflects only the
characteristics of interest and random error.
Types of validity:
1. Content validity
2. Construct Validity
3. Criterion- related validity
1. Content validity
It involves assessing the representativeness or sampling adequacy fro measuring the
measurement errors.
It is with multi- item measures of measuring instrument for assessing above
representativeness and adequacy.
The two most commonly used methods of content validation involve the use of logical
and personal judgments of groups of expert.
It is commonly used in applied research
38

2. Construct validity
It involves understanding the meaning of the obtained measurements.
It is concerned with knowing more than just that a measuring instrument works or
explains the variance of measurement errors. This validity is assessed through:
Convergent validity- two score from two different instruments are highly
correlated.
Discriminate validity- variables predicted to be uncorrelated
3. Criterion- related validity
It involves inferring an individuals score or standing on the measurement.
Concurrent validity- individuals present standing with respect to some other
variables.
Predictive validity- current measurement to future prediction of an individual.

4.4.2 RELIABILITY
Reliability means the consistency between measurements in the series.
It indicates the precision of measurement scores or how accurately such scores produced
with repeated measurements. But, same result is affected by some variable or random
error.
1. Test- retest reliability- repetition of same measure
2. Alternative form reliability- giving another form of measure
3. Split- half reliability- dividing in to two equivalent parts.
1. Test- retests reliability
Repeating measurement using the same instrument under as nearly equivalent conditions
as possible.
Compare the results- greater the difference, the lower the reliability.
2. Alternative form reliability
Applying two equivalent forms of measuring instruments to the same objects.
As in above test-retest reliability the results of two instruments are compared item- by-
item and the degree of similarity is determined
Basis is same as test- retest approach
3. Split- half reliability-
It is modification of alternative form approach.
It is obtained by calculating the correlations between two halves of an instrument.
The usual approach to split- half reliability involves dividing the total no. of items into
two groups (Odd and even numbers.
Computing correlation coefficient of two halves.
It is consistency is the reliability, which is ranging from o to 1.
If, less than 0.6 is unsatisfactory result
Relationship between reliability and validity
1. For a set to predict a particular criterion, predictive validity is more important than
reliability. When predictive validity is satisfactory, low reliability is not a serious
problem.
2. Given two tests measuring the same thing, the test with the higher reliability will also
have the higher validity coefficient.
3. Rule for the limit reliability places on validity. The correlation between the test and
independent criterion can never be higher than the squire root of the product of reliability
test.
Sources of measurement problems
1. Respondent- associated Errors
Respondent may not understand question.
Respondent may not remember the necessary information and rely on recall basis.
39

Reluctant to give information.


Problem of environment or conditions:
Problem of expression
Question ask to wrong person- electric shaver
Instrument- associated errors
1. Excessive and lengthy question.
2. Ambiguous instructions
3. Poor choice of words
4. Limited space, confusion or biasing.
5. Lack of sequencing
40

UNIT V: SAMPLING PROCESS AND DATA COLLECTION

5.1 Sampling and its significance in research


Sampling consists of obtaining information from only a part of a large group or population; and
it indicates about the whole population. The objective of sampling is thus to secure a sample
which will represent the population and reproduce the important characteristics of the population
under the study as closely as possible. The value calculated from a defined population, such
mean (), standard deviation (), standard error of mean (s.e.) is called a parameter. It is a
constant value because it covers all the members of the population. A value calculated from a

sample is called statistic such as mean ( X ). Standard deviation (s) and proportion (P).

The data is the basic units in statistical analysis and inference; is either collected by
experimentation or by sampling methods. Method of collection of statistical data by complete
enumeration of the population is census. If the data collected by a certain group or part of
population is called sampling enquiry. The principal advantages of sampling as compared to
complete enumeration of the population are:
Reduced cost
Save time and speed up
Greater scope and improved accuracy
Some terminology used in sampling
a. Universe/Population:
It is the set of object under study. In a census survey, all the universe or population is
studied while in a sample survey an appropriate number of units called samples is
selected and studied; the generalization is made for the universe or population from
which the samples are drawn.
b. Finite population:
The number of items or the units under the study is known.
c. Infinite population:
The number of units of the items is unknown.
d. Element:
Each and every unit of population or universe is called element. An element constitutes
one case for analysis.
e. Sampling unit:
The smallest unit of population to be sampled is called sampling unit and on which
observations can be made.
f. Sample and sample size:
An element or sampling unit from which information is collected is called a sample. A
sample should be optimum, effective, representative, reliable and flexible. The term
sample size refers to the number of items to be selected from the universe to constitute a
sample. This is the number of respondents or units in the population included in a sample
for studying the population.
g. Sampling Frame or source list
A list of all the units of population from which a sample is selected is called sampling
frame.
h. Parameter
A coefficient or value for the population that corresponds to particular statistic from a
sample is called parameter. A parameter is characteristic of population. For instance,
mean, standard deviation, etc.
i. Statistic
It is characteristics of a sample and is hence computed from the actual data.
41

j. Respondent
A sampling unit from which information is collected is called respondent.
k. Non- respondent
Those respondents who were included in the sample but failed to respond because they
refused, could not reach, or some other responses.
l. Reliability:
It refers to the freedom from random error. It is also termed as repeatability, the ability to
get the same data values from several measurements from a similar manner.
m. Validity:
A survey is valid to the degree that is measures what and only what it is supposed to
measure. To be valid it must be affected by numerous factors that systematically push
or pull the results in the one particular direction.
n. Sampling Error:
The degree to which the results from the sample deviates from those that would be
obtained from the entire population because of random error in the selection of
respondent and the corresponding reduction in the reliability.

The Sampling Process

Step 1 Define the population

Specify the sampling frame


Step 2

Step 3 Specify the sampling unit

Step 4 Selection of sampling method

Step 5 Determine the sample size

Step 6 Specify the sampling plan

Step 7 Select the sample


42

5.2 Types of sampling or Sampling Techniques

Sampling

Probability Non- probability

Simple Stratified Quota Snowbal Convenience


random random l
Purposive Self-
Systematic Cluster selection

Multistage
Extreme Homogenous Typical
case case
Heterogeneous Critical case

Figure: Sampling techniques


5.2.1 Probability sampling
Probability sampling is the most commonly associated with survey- based research where you
need to take inferences from your sample about a population to answer your questions to meet
your objectives. The process of probability sampling can be divided into four stages:
1. Identify a suitable sampling frame based on your research questions or objectives;
2. Decide on a suitable sample size;
3. Select the most appropriate sampling technique and select the sample;
4. Check that the sample is representative of the population.

If the population is less than 50 take all the entire population for the study

1. Identifying a suitable sampling frame


The sampling frame for any probability sample is a complete list of all the cases in the
population from which your sample will be drawn. The completeness of your sampling frame is
very important. An incomplete or inaccurate list means that some cases will have been excluded
and so it will be impossible for every case in the population to have a chance of selection.
Consequently sample may not be representative of the total population.
Checklist for selecting a sample frame
Are cases listed in the sampling frame relevant to your research topic, for example are
they current?
Does the sampling frame exclude all the cases, in other words is it complete?
Does the sampling frame exclude irrelevant cases, in other words is it precise?
(For purchased lists) can you established precisely how the sample will be selected?
2. Deciding on a suitable sample size
Generalizations about the populations from the data collected using any probability samples are
based on probability. The larger your samples size the lower the likely error in generalizing to
the population. Probability sampling is therefore a compromise between the accuracy of your
43

findings and the amount of time and money you invest in collecting, checking and analyzing the
data. The size of sample based on:
The confidence you need to have in your data, that is the level of certainty that the
characteristics of the data collected will represent the characteristics of the total
population;
The margin of error that you can tolerate, that is the accuracy you require for any
estimates from your sample;
Statistical techniques have minimum threshold of the cases for each cell for instance chi
square and to a lesser extent;
The size of the total population from which your sample is being done.
Calculating the minimum sample size
Based on
How confident you need to be that the estimate is accurate (the level of confidence in the
estimate);
How accurate the estimate need to be (the margin of error that can be tolerated);
The proportion of responses you expect to have some particular attribute.

The formula is:


2
z
n p % xq% x
e%

where, n is the minimum sample size required


p% proportion belonging the specified category
q% is the proportion not belonging to the specified category
z is the z value corresponding to the level of confidence required(Table below)
e% is the margin of error required
Where your population is less than 10000 a smaller sample size can be used without affecting
the accuracy. This is called the adjusted minimum sample size. It is calculate from the following
formula:
n
n'
n
1
N
where, n is the adjusted minimum sample size
n is the minimum sample size (as calculate above)
N is the total population

Worked example:
Given,
Total population = 4000
95 % certain and z value = 1.96
The margin of error that can be tolerated = 5%
You still need to estimate the proportion of responses who receive a visit from their home care
assistant at least once a week. From your pilot survey you discover that 12 out of 30 clients
receive a visit at least once a week, in other words that 40 percent belong to specified category
and 60 percent do not.

These figures can then be substituted into the formula:


2
1.96
n 40 x 60 x
5
44

2400x ( 0.392) 2
2400x 0.154
369.6

Your minimum sample size is therefore 370 returns.


As the total population of home care clients is 4000 the adjusted minimum sample size can now
be calculated:
n
n'
n
1
N
369.6

369.6
1
4000
369.6

1 0.092
369.6

1.092
338.46
Because of the small total population you only need a minimum sample size of 339. However
this assumes a response rate of 100 percent.

The importance of high response rate


There are problems of responses. They are:
Refusal of respond;
Ineligibility to respond;
Non- contact
Total number of responses
Response rate =
Total number in sample - (Ineligible + unreachable)

The actual sample size:


Given,
Adjusted minimum sample size is 439
Response rate would be 30 percent
Then,
439 x100 43900
na 1463
30 30

3. Selecting the most appropriate sampling technique and the sample


There are five main techniques can be used to select the probability sample. They are:
3.1 Simple random
3.2 Systematic
3.3 Stratified
3.4 Cluster
3.5 Multi-stage

3.1 Simple random


In random sampling all the items of population have equal chance of being selected in the
sample whereas in the non- random sampling personal knowledge and opinion are used to
45

identify those items from population that are to be included in the sample. Random sampling
ensures the law of statistical regularity, which states that if on an average a sample chosen is a
random on then the sample will have the same composition and characteristics as the universe.
This is the reason why random sampling is considered the best technique of selecting a
representative sample. Since all groups of the population are proportionately represented in the
sample, random sampling is also called representative or proportionate sampling.

Simple random sampling is the technique of drawing a sample in such a way that each unit of
the population has an equal and independent chance of being included in the sample. The
selection is not affected by the bias of the investigator and it depends completely on the element
of chance.

Simple random sampling with replacement and simple random sampling without replacement
are the two types simple random sampling. If a unit selected in a draw is replaced in the
population before making the next draw, then it is known as simple random sampling with
replacement. But in the other hand if the unit selected is not replaced in the population, then it is
known as simple random sampling without replacement. For SRSWR the probability of
selection of any unit in the universe with N objects always remains 1/N whereas for SRSWOR,
the number of items in the universe goes on decline.

Selection of Simple Random Sample


Lottery method
Random numbers method

Lottery Method
This is the simplest method of selecting a random sample from a finite sample. To select r units
as a random sample from a finite population of N units we write 1, 2, 3, , up to n in the
faces of n cards. Then these N cards are thoroughly shuffled and r cards are drawn on by one.
The r unit corresponding to numbers on the cards drawn constitutes a random sample.

Random Number Method


The other method of obtaining a random sample is by the use of random number Tables.
Statisticians like Tippet, Yates and Fisher have prepared Tables of random numbers. Particularly,
Tippets random table is used for this purpose. This random table is used in the case when, if the
population is not widely spread geographically and the population is more or less homogenous.
Steps:
Identify the N units in the population with the numbers from 1to N.
Select at random any page of number table and pick up the required numbers N in any
row, column or diagonal at random.

3.2 Systematic
Systematic sampling involves you selecting the sample at regular intervals from the sampling
frame. To do this you:
1. Number each of the cases in your sampling frame with a unique number. The first case is
numbered 0, the second 1 and so on;
2. Select the first case using a random number;
3. Calculate the sampling fraction;
4. Select the subsequent cases systematically using the sampling fraction to determine the
frequently of selection.
46

Actual sample size


Sampling fraction =
Total population
Example:
Given,
Total patients population = 1500
Acceptable and accuracy sample size = 300
How to decide to select them by using systematic sampling?
1. Calculate the sampling fraction,
300 1
it means you can select the every fifth patient from the sampling frame.
1500 5
2. You decide to select one- digit random number between 0 and 4.
3. If the selected number is 2, then you would select the following patients numbers:
2 7 12 17 22 27 32 37
and so on until 300 patients had been selected.
3.3 Stratified random sampling
Stratified random sampling is modification of random sampling in which you divide the
population into two or more relevant and significant strata based on one or a numbers of
attributes. In effect your sampling frame is divided into a number of subsets. A random sample
(simple or systematic) is then drawn from each of the strata. Consequently stratified sampling
shares many of the advantages and disadvantages of simple random and systematic sampling.
Steps:
1. choose the stratification variable (s);
2. divide the sampling frame into the discrete starta;
3. number each of the cases within each stratum with a unique number as discussed earlier;
4. select your sample using either simple random or systematic sampling as discussed
earlier.
Example:
Female stratum Male stratum
Number Customer Selected Number Customer Selected
000 Mrs. B Baker 000 Mr. L Baker
001 Mrs. P Davis 001 Mr. S Davis
002 Mrs. P Lewis 002 Mr. P Lewis
003 Mrs. M Saunders 003 Mr. M Saunders
004 Mrs. K Smith 004 Mr. K Smith
005 Mrs. A Thornhill 005 Mr. A Thornhill
006 Mrs.D Wollons 006 Mr.D Wollons
007 Mrs. P Wordden 007 Mr. P Wordden
3.4 Cluster
This method is used when the total area under study is big and study population is spread well
over the area thus making difficult to prepare sampling frame. In this method, the area is first
divided into smaller unit called clusters (based on the purpose of the study), and these cluster are
selected for the purpose of study by using either random sampling techniques or systematic
sampling technique. Some researchers say that it is an extension of multi- stage sampling. All
the units in that cluster are used as sample. Another purpose of it is to reduce the cost of
traveling. As compared with the SRS and Systematic Sampling, this technique is relatively less
representative. If the technique is used for geographical areas, the technique is called Area
Sampling.
Stages:
1. Chose the cluster grouping for sampling frame.
47

2. Number each of the clusters with a unique number. Like 000, 001, 002,
3. Select sample using- simple random sampling (either).
3.5 Multi-stage sampling
In this technique the items are selected at different stages at random and are used when the
universe is very large:
Stage 1: District
Stage 2: VDCs
Stage 3: Wards
Stage 4: Households (HHs)
Stage 5: Individuals (Family members)
This technique helps to reduce the traveling cost and cost of interviews. Moreover, this
technique does not require a complete sampling frame of final stage units at the beginning
sometimes also called as the combination of SRS and Stratified Sampling.

5.2.2 Non- Probability Sampling


5.2.1.1 Quota sampling
Quota sampling is entirely non- random sampling and is normally used for interview surveys.
Quota sampling is therefore a type of stratified sample in which selection cases within strata is
entirely non- random. This technique has relatively less bias than of other non- probability
sampling. Certain quota is fixed in each group under study and then samples are selected. If
random procedure is applied in this technique becomes the stratified sampling. To select a quota
sample you:
1. divide the population into specific groups;
2. calculate a quota for each group based on relevant and available data;
3. give each interviewer an assignment, which states the number of cases in each quota
from which they must be collect data;
4. combine the data collected by interviewers to provide the full sample.
Example:
A market research survey requires you to interview a sample of people representing those age 20
64 who are in employment. No sampling frame is available. You wish to disaggregate your
findings into groups dependent on respondents age and type of employment. Previous research
suggests that gender

5.2.1.2 Convenience sampling


Convenience sampling refers to samples selected not by judgment or probability techniques but
because the elements in the fraction of the population can be reached conveniently. There is no
attempt made to have a representative sample. Selection of sampling units is totally based on the
convenience of the researcher. When both time and money are limited, this sampling is widely
used.

For instance: If you have to conduct man-on-the-street interviews, you probability use
convenience sampling. You need to stand up in a corner of a street and interview the desired
number of passers by.

Commonly used convenience samples are friends, family members, associations and passers- by.
Although convenient sample is not very scientific it is perfectly valid in exploratory research or
in the pre- test phase of a study where there is a need to get only an approximation of the actual
value. This method is quick, convenient and less expensive.
5.2.1.3 Judgment or purposive sampling
Judgment samples are selected from the population through researchers intuition or on some
other subjective basis. The selection of sample is deliberate and purposive and it is not random.
48

Sampling is done with sound judgment or expertise and appropriate strategy; one can carefully
choose the elements to be included in the sample. The selected elements are representative in
such a way that the errors of judgment in the selection will cancel each other out. Thus, sample
representativeness is highly dependent upon the good judgement of the researcher. Probably the
most valid usage of judgmental sampling is to obtain expert opinion.

For instance, if you want to study the issue of corruption I nNepalese society, you select a
sample of 20 of the senior professors of Tribhuwan University to give their opinion on the
subject. Certainly the judgment of these professors is much superior to a convenience sample
that might arbitrarily be limited to the opinion of twenty people in your neighbourhood.

In judgment sampling, subjects are selected on the basis of their experience in the subject
investigated. As in the case of convenience samples, judgment samples can be very misleading if
they are interpreted as accurate reflections of universe characteristics.
5.3 Sampling Error and Non- sampling Error
A sampling error is the error, which is made in selecting samples that are not representative of
the population. This error is the difference between the sample value and true value of the
population to be exactly representative of the population; some degree of sampling error will be
present whenever we select a sample.

Increasing the sample size can reduce sampling error. The error can be completely eliminated by
increasing the sample to include every item in the population. Because the sampling error is the
result of chance, it is subject to the laws of probability; and because of this, a sampling error will
be normally distributed and we can define its range at a given confidence level.

Non-sampling error, is every thing else (beside the sampling error) that can inject inaccuracies
and bias into the results of a study. Nonsampling error can enter the data in an insidious way,
these error includes to:
Inaccurate reporting by the respondent (biased guess, inaccurate memory, poor recall,
etc.
Actual lying by the respondents.
Poor sampling design, for instance the inability to locate proper respondents due to poor
instruction, poor maps, nonexistent address and so on.
Misinterpretation of question due to ambiguous wording.
Respondents terminating their participation in the data gathering because it is felt to be
too long, too tedious, or too personal.
Failure of the interviewers to follow instructions, which leads to their leading
respondents, giving nonverbal clues, and recording errors.
Coding and/or editing errors.
Some broad guidelines for minimizing nonsampling errors in surveys:
Keep the sample surveys as easy to execute as possible.
Use the smallest sample consistent with study objectives.
Restrict the questionnaire to data essential to the main issue.
Pre- testing the questionnaire. Regularly undergo the interview yourself to determine
your ability to answer and find how much fatigue is involved in answering the
questionnaire.
Make an effort to minimize participants fatigue.
Rotate key questions to discover when respondent fatigue begins.
Establish procedures for keeping both respondent and interviewer involved in the study.
49

Dont ask question to respondents they really cannot answer.


Dont ask the interviewer to do the impossible. Such request encourage sloppy work and
cheating.
5.4 Determination of Sample Size in Estimating Population Mean and Proportion
Parameter Statistics Error

If denotes the population mean and
x denotes the sample mean (estimator) then the

relationship can be expressed as:



x error or x error
By properly estimating the error, it is possible to provide an interval estimate instate of point
estimate of the population mean. The error is usually specified by standard deviation of the
statistic, which is known as standard error. While estimating the population mean , an interval
estimate is the form:

x 1.96
n
Where is the mean, is the standard deviation of the population and n is the sample size. The
factor 1.96 follows from the theory of normal distribution.

There is a 95% chance that the true population mean lies in the interval

x 1.96 , x 1.96 called the 95% confidence interval of .
n n
In many practical situations it is difficult to determine the population standard deviation () and
the population mean (). Under these circumstances, a sample standard deviation (s), the point
estimator of the population standard deviation is used to construct an interval estimate with the
following formula:
s
x 1.96
n
s s
So that the 95% confidence interval of will be x 1.96 , 1.96
x
n n
Example ( known)

The department of agricultures future wheat gain production, based on estimate from a sample.
A sample of 100 plots produced a mean of 35 bushels/acre. The department agriculture assumes
a population standard deviation of 4.5 bushels. Calculate a 95% confidence interval and interpret
the findings.
The 95% confidence interval of the population mean () are,
4.5 4. 5
x 1.96 = 35 1.96 = 35 1.96 = 35 0.882 (95% confidence interval)
n 100 10
= 34.118 and 35.882 bushels/acre

The probability is 0.95 that the mean bushel per acre of wheat is in the interval of 34.118 and
35.882. In this case, about 5 out of 100 confidence interval would not contain the population
mean () per acre.

Example ( unknown)
Environmental scientist estimates the mean biomass from a sample of forest is 5.6kg/sq. m. from
50 plots, with a standard deviation of 1.2 kg/sq.m. Construct the 95% confidence interval. What
are the 95% confidence limits? Interpret the findings.
50

Then,
95% confidence interval are
s
x
n
1.2 1.2
= 5.6 1.96 = 5.6 1.96 = 5.6 0.33 (95% confidence interval)
50 7.07
= 5.27 and 5.93 kg/sq. m, which is confidence limits. The probability is 0.95 tha biomass
mean is in the interval of 5.27 and 5.93 kg/sq.m.

Approach in Determining Sample Size


Approach 1:
The approach based on precision rate and confidence level i.e researcher will have to specify the
precision that he wants in respect of his estimates concerning the population parameters.

Example
Mean of universe within 3 of the true mean with 95% confidence. It means that the precision
specified by the researcher is 3 i.e. if the sample mean is Rs. 100 the true value of mean will
be no less than Rs. 97 and no more than Rs. 103. Simply it is enough to say that, acceptable
error, e is equal to 3.

Approach 2
Through the approach of Bayesian statistics t weigh the cost of additional information against
the expected value of additional information.

Sample Size when estimating a Mean

The confidence interval for the universe (for ) is given by


x z
n

Where, x is the sample mean, z is the standard variate at a given confidence level. This is 1.96
for 95% confidence level, n is the size of sample and is the standard deviation of the
population which is estimated from the past experience or on the basis of a trail sample.

Suppose, we have = 4.8. If the difference between and x i.e. the allowable error is 3 of
the sample mean with 95% confidence, then we can express the acceptable error e as
4.8
ez 3 1.96 n 10
n n
In general way, if we want to estimate in the population with standard deviation with an
error no greater than e by calculating a confidence interval with confidence corresponding to z,
the necessary sample size, n is computed as
z 2 2
n 2
e
Where z is standard normal variate at a given confidence level, is the standard deviation of
population and e is acceptable error. This formula is applicable when the population is infinite.
The sample size so determined is only a approximate value. If the sample fration n/N is
negligible (say 0.5) it can be taken as satisfactory sample size. Otherwise, the required sample
size will be
51

n
n*
n where N is the size of population
1
N
Example
Suppose a sample of farmers is to be selected by simple random sampling to estimate the cost of
cultivation of paddy per hectare. The precision is 50 rupees with 95% confidence. From a past
survey, it is known that s = 250. How large must the sample be if the population size ia 1000?
Find the sample size, if the population size is 10000?

Then,
Error e= 50, s = 250 and the reliability coefficient i.e. 95% confidence value, z = 1.96
(approximately 2)
z 2 2 2 2 250 2
n 2 100
e 50 2
The population size is, N = 1000
Therefore,
n 100
0.10
N 1000
Since the sampling fraction is considerable we have to recalculate n using thee formula,
n 100 100
n* 91
n 100 1 0 . 10
1 1
N 1000

If, N = 10000
n 100
0.01
N 10000

Determining the sample size


Thumb rules of determining the sample size are:
Sample sizes larger than 30 and less than 500 are appropriate for most research.
When samples are to broken into sub- samples (male/female, juniors/seniors), a
minimum sample size of 30 fro each category is necessary.
In multivariate research (multiple regression analysis), the sample should be several
times (preferably 10 times or more) as large as the number of variables in the study.
For simple experimental research with tight experimental controls (matched pairs, etc),
successful research is possible with samples as small as 10 to 20 in size.
Some principles that influence the sample size:
The greater the dispersion or variance within the population, the larger the sample must
be to provide estimation precision.
The greater the desired precision of the estimates, the larger the sample must be.
The narrower the internal range, the larger the sample must be.
The higher the confidence level in the estimate, the larger the sample must be.
The greater the number of sub- groups of interest within a sample, the greater the sample
size must be, as each sub- group must meet minimum sample size requirements.
If calculate sample size exceeds 5 percent of the population, sample size may be reduced
without sacrificing precision.
52

5.5 Primary and Secondary Data


5.5.1 Secondary data
Secondary sources refer to those for already gathered by others. The sources of secondary data
can be divided into two groups: internal and external. The internal secondary data are found
within the company such as sales information, accounting data and internally gathered research
reports. External secondary data are collected from sources outside the company likes books,
periodicals, published reports, data services and computer data banks.

Documents and Records:


Written Document
Organization records
notice and publications
Websites and internets
Committee Records
Books, Journals, Newspaper, research reports
Non- Written Records
CD- ROMs, Films,
Tapped interviews,
TV and radio recording
Picture and drawings

Secondarydata
Secondary Data

Surveys:
Census
Population, Industrial, Employment,
Landholding, agricultural
Regular and occasional surveys
Price Index, Family income and spending
Labor markets, imports and exports,
Organizational surveys
International indexes,
Attitudes surveys, Occasional surveys

5.5.2 Primary data


Primary data are original data gathered by the researcher from the research project at hand. Thus
these data collected from meeting the specific objectives of the study. Primary data can be
collected through interviews, questionnaires, observations or experiments.
Sources and methods of Primary data
1. Interviewing: Collection of data by asking the interest area of research and recording.
2. Questionnaire
A questionnaire is the formal list of questions designed to gather responses from the respondents
on a given topic. It involves several steps, including writing question items, organizing the
question items on a questionnaire, administering the questionnaire, and so on.

Developing Questionnaire
1. Introduction
Questionnaire is the list questions for collection of data from respondents.
Questions are the foundations questionnaire- framing is important.
53

Questionnaire should translate the research questions or objectives. Data are for testing
the hypothesis between interest variables.
2. General issues
Formulate the precise questions.
Order of questions and sections
List the categories of questions.
Insert the data processing requirements and instruments.
3. Level of measurements
Considered the best level of measurements such as nominal, ordinal, interval, n ratio, etc.
4. Mode of administration of instruments
Face to face.
Telephone
Mail questionnaire
E-mail.
The choice of the mode depends up on the nature topic of the study and sensitivity,
characteristics of target groups (age, education level, health), expected response rate and
resources availability (costs, fund).
5. Problems in asking questions
Length.
Complexity.
Sensitiveness
Hypothetical.
Threatening.
Leading
6. Designing questionnaire

A. Conceptualization and operationalisation


Determine the concepts to be measured
List the concepts- the specific hypothesis and explanatory techniques such as focus
group discussion for concepts of interest
Identify and list the dimensions of concept
Operationalisation- Specification of empirical observations that can be indicators of the
attributes contained with the given concept. For example, Reading news paper: Visit to
bookshop, Visit to library, Frequency of visit, buying himself
B. Selecting the questions
Content of questions
Behaviour /events- that respondents have done or experienced.
- What newspaper?
- When: How long.? When did you.?
- How often: How many times; how frequently; etc
Attitude/opinion of the respondents
- Affection- like or dislike,
- Cognitive- What do you know about ..? What do know about it.?
- Action- Willingness, Are you willing to.?
- Importance- How important?
Very, important, important andnot important.
- Satisfaction- How much satisfied?
Very satisfied, generally satisfied, dissatisfied,
- Agreement- strongly agree, agree, neither agree or disagree, .
54

Respondents demographics
- Age, gender, marital status, employment status, religion, ethnicity,..
Format- closed and open ended

7. The Layout of Questionnaire


Sequencing of questions.
Start with simple and non-threatening questions.
Put difficult questions at the end.
Put related content together.
Provide proper introduction and transitions.
Distinguish questions from instructions.
Put screeners and skip pattern as per necessary.
Use transitional phrases when switching from one topic to another to switch
respondents set of mind.
Draw the attentions like, Next, I would like to learn more about.
Now, I would like to ask you about..
Sequencing within the topic area and gradually move from general to specific
questions.
8. Procedure for drafting and testing questions
Decide the concepts to be measures, supporting the variables, indicators and the
responses categories.
Draft or copy of new question and response categories
Pre-test new questions
Rewrite the questions
Check the validity of questions
Put the questions in order
Pilot test the questionnaire
9. Goals of writing questions
- Interviewers can questions consistently.
- Question wording- Easy cognitive study and respondents can understand the questions
easily.
- Respondents will interpret the question in the same way as the researchers do.
10. Question wording
- Question should be in word.
- Clear about the objective.
- Be simple and clear.
- Define the terms before presenting.
- Face to face interview- follow choice question stem.
- Provide the response options at the end of questions.
- Avoid unnecessary calculations such as total of last year, time spend for shopping, etc.
- Avoid two answers question
- Avoid leading question.
- Provide possible clues, etc.
- Use chart when necessary like calendar, checklist.
- Use flash cards when ranking, etc.
11. Types of questions
Open - ended questions (free answer/responses):

Advantages
- Good enough when little is known about the topic?
55

- Preserve respondents wording, thoughts and emphasis.


- Better for sensitive topic.
- Less influence on respondents.
- Better for respondents insights and opinion.

Disadvantages
- Demanding for respondents.
- Answer not sufficiently focused.
- Rarely exact enough.
- Costly coding.
- Open to interviewer bias.
Closed- ended questions (Fixed answers):
- Questions provide a predetermined set of response categories from which respondents must
choose one or more.
How do you rate the internal security implemented by the previous government of Nepal?
Would you rate it poor, below average, average, above average and excellent?
1. Poor
2. Below average
3. Average
4. Above average
5. Excellent
Advantages:
- Easy to answer to respondent.
- Responses can be standardized.
- Easy for coding.
- Same time and money for data processing.
- No bias to interviewer.
Disadvantages:
- Some relevant categories might be missing.
Farmers categories
Cat A = Food sufficient for 12 months
Cat B = Food sufficient for 9 to < 12 months
Cat C = Food sufficient for 6 to < 9 months
Cat D = Food sufficient for 3 to < 6 months
Cat E = Less than 3 months and landless
- Bias due to order of responses.
- Imposing to respondents.
- Affect the responses.
- Suitable for only well defined cases only.
Contingency questions:
- At the last, it is better to ask about the respondents view or expectations, overall
problems of the issues, which might be different from your objectives. It might be
courage to respondent.
Introductory paragraph (Opening scripts):
- Introduction each other.
- Where you from- institution.
- Objectives of the study.
- Use of the information in future.
- Level of confidentiality.
- Possible time taken; and opportunities of asking the questions.
56

UNIT VI: TESTING OF STATISTICAL HYPOTHESIS

6.1 Statistical Hypothesis


Another major topic inferential statistics is testing hypotheses made about population
parameters. Test of hypotheses is also called the test of significance, on the basis of sample
results, enables us to decide if the deviation between the observed sample statistics is significant
or might be attributed to the chance or the fluctuation of sampling.

A statistical rule of testing validity of hypothesis, which tells us whether the hypothesis is
correct and is sustained or whether it is false and is to rejected, under some specified
conditions. In other words, it aims to develop an understanding of the concept like to decide
whether:
1. a newly invented vaccine is more effective than the one in usual practice for curing a
disease
2. the mean body temperature of healthy adults is 98.600 F

Fundamental of Hypothesis Testing


Null and alternative hypothesis
To test a hypothesis we must first state the null hypothesis. The concept of null hypothesis is
very important and significant in hypothesis testing. It refers to beginning hypothesis that the
researcher wishes to disprove, because disproving it provides a stronger conclusion for the
researcher. Null hypothesis is commonly denoted by as H0. The alternative hypothesis denoted
as H1, on the other hand, it refers to the complement of the null hypothesis that is used to verify
the null hypothesis.

For the null hypothesis, H0 : = 0 about the population mean (), three basic forms of
alternative hypothesis are:
H1 : 0
H1 : < 0
H1 : > 0
It means that whenever the null hypothesis is not true, one of the alternatives must be true.
Although the researcher may hope that the evidence favors H 1 hypothesis tests are setup to test
the null hypothesis, H0 . The alternative hypothesis will be accepted only if strong evidence is
found against H0 .

Two Types of Errors


The main objective of sampling theory is to draw a valid inference about the population
parameters on the basis of the sample results. In practice when testing a null hypothesis, we
arrive at a conclusion of rejecting it or failing to reject it. In carrying out such decision process,
for any test we come to encounter with the four possible mutually disjoint and exhaustive
decisions given in the following Table.
True state Decisions and error types
We decide to accept H0. We decide to reject H0.
H0 is true Correct decision Type- 1 error
(Rejecting a true null hypothesis)
H0 is not true/ false Type- 2 error
(Failing to reject a false null Correct decision
hypothesis)
Type I (Alpha) Error: When we mistakenly reject the null hypothesis when indeed the
null hypothesis is true, then the type of wrong decision is known as a Type I or Alpha
error. Since such an error is particularly embarrassing to the researcher (saying
57

something exists when it does not), he or she tends to avoid risking this error and
chooses a conservative level of significance, which minimizes it.
Type II (Beta) Error: When we mistakenly accept our null hypothesis when in fact it is
false, then we commit another error known as the Type II or Beta error. In other words,
to conclude falsely that a difference does not exist in the data when in fact it does.
The probability of rejecting the null hypothesis when it is true is called the level of significance
and is denoted by the symbol (alpha). It is maximum size of type- 1 error, which we are
prepared to risk and its value is typically predetermined. Though we can choose any level that
the specific research problem warrants, very common choices of are: = 0.05 and = 0.01.
These are understood as 5% and 1% level of significance, respectively. If we adopt 5% level of
significance, it implies that in 5 samples out of 100 we are likely to reject a correct H 0. In other
words, we are 95% confident that our decision to reject H0 is correct, =0.01 is used for high
precision and = 0.05 is used for moderate precision.
Steps in hypothesis testing
1. Formulate the null hypothesis in statistical terms.
2. Formulate the alternative hypothesis in statistical terms. Since the form of the alternative
hypothesis determines the rejection, the alternative hypothesis must be carefully stated to
represent the research problem.
3. State the level of significance and sample size n. Since the null hypothesis is usually set
up in such a way that we want strong evidence against it before we reject it, the level of is
usually rather small.
4. Select the appropriate statistic and the rejection region. In all hypothesis testing problems,
common sense will dictate the proper form of the rejection region. Choosing the exact
region requires knowing the distribution of the statistic.
5. Collect the data and calculate the statistic. The appropriate formula is used for the
appropriate statistical test and the value of the test statistic is found.
6. Test the hypothesis. If the calculated statistic falls in the region, reject null hypothesis in
favour of alternative hypothesis, if the calculated statistic falls outside rejection.

Testing the Null Hypothesis


Accepting and Rejecting Null Hypothesis
Testing the null hypothesis results in one of the following two outcomes:
Accepting the null hypothesis as true, in which case it is concluded that and difference in
the results are not statistically significant, therefore, are probably due to sampling error
or chance.
Rejecting the null hypothesis as false, in which case it is concluded that the differences in
the result are statistically significant, therefore are probably due to to some determining
factor or conditions, other than chance.
Accepting the null hypothesis also means that the corresponding research hypothesis is not
supported. Rejecting the null hypothesis also means the corresponding research hypothesis is
supported.
6.2 Level Significance
The mistake of failing to reject the null hypothesis when it is false is type- 2 error. For the same
test of testing the mean body temperature of healthy adults, a type- 2 error would be the mistake
of failing to reject the null hypothesis ( = 98.6), when it is actually false (that is, the mean is
not 98.6). The symbol (Beta) is used to represent the probability of type- 2. It is also called the
beta risk.

The mathematical theory of testing of hypothesis is based on the principle that at fixed , the test
procedure should have the least - risk. So is given the level of significance. A lower such as
0.01 or 0.001, etc. is preferable if we consider type- 1 error more undesirable than type- 2 error.
58

But, if the type-2 error is costly, we may prefer to run higher risk of type- 1 error by increasing
to 0.10 or 0.20, etc.

Test Distribution

Types of Tests

1. One Tail Test:


A statistical test in which the alternative hypothesis specifies population parameter lies entirely
above or below the value specified in the null hypothesis is called a one tail test.
H0 : Rice yield = 2 t/ha
Ha : Rice yield = > 2 t/ha
Or, Ha : Rice yield = < 2 t/ha

2. Two Tail Test


If a hypothesis does not specify that the parameter lies on one particular side of the specified in
the null hypothesis, is called Two Tail Test.

H0 = Rice yield = 2 t/ha


Ha = Rice yield 2 t/ha

Region of Region of
Acceptance rejection rejection

H0 H0

One Tail Test Two Tail Test

Steps in Hypothesis Testing

Hypothesis testing is a process of determining whether the expectation that a hypothesis


represent or found to exist in the real world or determination of the validity of an assumption is
called hypothesis. The steps at hypothesis testing are given below. They are:

1. Setting hypothesis:

A hypothesis is based on research objective and is developed and supports the general
assumption about the particular statement. The foremost job in hypothesis testing is to define a
hypothesis and alternative hypothesis. For instance,
H0: U1 = U2 = 20
U1 = 20
U1 - U2 = 0
It means the sample statistic and the assumption of the specified values are equal or there is no
difference between the specified value and population parameter, or two means are equal.
Ha: U1 U2 OR, U1 > U2 OR, U1 < U2
U1 20 U1 > 20 U1 < 20
U1 - U2 0 U1 - U2 > 0 U1 - U2 <0
59

It means that the sample statistic and the assumption of the specified values are not equal or
more than or less than the population parameter or two means are not equal, etc. Or, there is
difference between the specified and population parameter, or two means are not equal, etc.

2. Fixing Level of Significance


After setting null hypothesis and its alternative, the next step I to fix the level of significance. It
is probability at which the given hypothesis is accepted or rejected. For instance, if the level of
significance is fixed at 5%, it means the researcher is willing to take as much as a 5% risk of
rejecting the null hypothesis when it (H0) happens to be true. It is the maximum value of the
probability of rejecting H0 when it is true and is usually determined in advance.

3. Selecting the appropriate tools or test statistics and computations:


There different test statistics used for hypothesis testing. These tests are t- test, Z test and 2
test, etc. Based upon the type of variables and relationships, appropriate test statistics is selected.
Table 1: Statistical measures of association
Independent Variables
Categorical Continuous
Depe Categorical Chi-squire F- Ratio
nden e.g. Perception : sex Farm size : Income
t Continuous ANOVA (F- Ratio) Regression (F- Ratio
Vari Paired t- test Correlation analysis
ables Yield : Farm size Yield : Inputs
Income : Occupation Income: Land holding
Income : Age

4. Decisions rules and conclusions


There are two regions in a normal distribution curve for testing hypothesis. One is acceptance
region for null hypothesis and another is rejection region. The decision rules depend upon test
statistics value where it falls. If the test statistics value falls in the acceptance region, null
hypothesis is accepted and vice versa at the given level of significance. Conclusion is drawn
based on the acceptance or rejection hypothesis.

Region of Region of
rejection rejection
Acceptance region

Errors in Hypothesis Testing:


If a hypothesis is accepted or rejected, a researcher incurs two types of hypothesis. One is Type I
(Alpha) and another is Type II (Beta) error. If a hypothesis is rejected which is in fact true is a
Type I () error. Whereas, if a hypothesis is accepted which is in fact false is Type II (Beta-)
error.
Table 2: Error in Hypothesis Testing
Hypothesis Decision
Null Hypothesis Accept (H0) Rejected(H0)
True Correct decision Type I Error (- Error)
False Type II Error () Correct decision
60

Type I Error is also called the level of significance. It is usually determined in advance and is
understand as the level of significance of testing the hypothesis. If Type I Error is fixed at 5%. It
means that there are chances in 100 that we will reject null hypothesis when null hypothesis is
true. Type I error can be control by fixing it at a lower level. For instance, if we fix it 1% we will
say that the maximum probability of coming Type I error would be one in 100. However, when
we try to reduce Type I error, the probability of coming Type II error increases thus making trade
off.
6.3 Difference Between Parametric and Non- Parametric Tests
Most statistical tests fall into one of two types such as parametric which are based around
normally distributed data and non- parametric, which is also called distribution free- tests.
Although, parametric tests are commonly used, there are certain circumstances where non-
parametric tests should be used in preference.
Characters of Parametric tests
1. Require interval/ ratio data used on actual observation or measurement
2. Require data to be normally distributed
3. With small sample sizes it is difficult to check that the data are normal
4. Use all the information in the data
5. Slightly more likely to detect a statistical difference or relationship if present (data
are normally distributed)
Characters of Nonparametric tests
1. May be used on nominal or ordinal data as well as interval/ ratio data (can be used on
observations, measurements or rank values)
2. Do not require normally distributed data (transformation of data is not necessary)
3. May be used on relatively small sample sizes since normality is not needed
4. Do not used all information in the data (ranks are used as opposed to the actual data
values)
5. Slightly less likely to detect a statistical difference or relationship if present (if data
are normally distributed)
6.4 Use of z- Distribution in Hypothesis Testing of Population Mean and Population
Proportion in one Sample Case.
When observations from counts belonging to particular categories such as diseased or healthy,
dead or live etc, the data are normally summarized in terms of proportions. We may then be
interested in comparing the proportions of incidence of an attribute in two populations. The null
hypothesis set in such cases is H 0 : P1 P2 and the alternative hypothesis is H 0 : P1 P2 (or
P1>P2 or P1 < P2) where P1 and P2 are proportions representing the two populations. In order to
test our hypothesis, we take two independent samples of large size, say n 1 and n2 from the two
populations and obtain two sample proportions where P1 and P2, respectively. The test statistic is
p1 p2
z
p1q1 p2 q2

n1 n2
Where q1 = 1- p1 and q2 = 1- p2. This statistics follows a standard normal distribution for large n1
and n2.

Example:
Consider an experiment on rooting of stem cutting of Casuriana equisetifolia wherein the effect
of dipping the cutting in solutions of IBA at two different concentrations was observed. Two
batches of 30 cuttings each, were subjected to dipping treatment at concentrations of 50 and 100
ppm of IBA solutions, respectively. Based on the observations on number of cuttings rooted in
each batch of 30 cuttings, the following proportions of rooted cuttings under each concentration
were obtained. At 50 ppm, the proportion of rooted cutting was 0.5 and at 100 ppm, the
61

proportion was 0.37. The question of interest is whether ther observed proportions are indicative
of significant differences in the effect of IBA at the two concentrations.
Given,
p1 = 0.5 and p2 = 0.37
q1 = 0.5 and q2 = 0.63
Then,
0.5 0.37
z 1.024
(0.5)( 0.5) (0.37)(0.63)

30 30
Since the calculated value of z (1.024) is less than the Table value (1.96) at 5% level of
significance, we can conclude that there is no significant difference between the proportion
rooted cuttings under the two concentration level.

Test of Goodness of Fit (or Chi-square test)

In testing the hypothesis, sometimes our objectives may be to test whether a sample has come
from a population with a specified probability distribution. The expected distribution may be one
based on theoretical distribution like normal, binomial or a pattern expected under technical
grounds. For instance, one may be interested in testing whether variable like the height of trees
follows normal distribution. A tree breeder may be interested to know whether the observed
segregation ratios for a character deviate significantly from the Mendalian ratios. In such
situation we want to test the agreement between the observed and theoretical frequencies. Such a
test is called test of goodness of fit.

For applying the goodness of fit test, we used only the actual observed frequencies and not the
percentages or ratios. Further, the observations within the sample should be non- overlapping
and thereby independent. The expected frequency in each category should preferably be more
than 5. The total number of observations should be large, say more than 50.

The null hypothesis in goodness of fit tests is that there is no disagreement between the observed
and theoretical distribution. The test statistic used is,
k
(O Ei
2 i
i i Ei
Where Oi = Observed frequency in the ith class
Ei = Expected frequency in the ith class
k= Number of categories or classes
2
The statistic given above follows as chi- square distribution with k-1 degrees of freedom. In
case of expected frequencies are derived from parameters estimated from the sample, the degree
of freedom is (k p 1) where p is the number of parameters estimated. For instance testing the
normality of a distribution, and 2 would be estimated from the sample by x and s2 and
degree of freedom would be therefore reduce to (k 2 1).

The expected frequencies may be computed based on the probability function of the appropriate
theoretical distribution as relevant to the situation or it may be derived based on the scientific
theory being tested like Mendels law of inheritance. In the absence of a well- defined theory, we
may assume that all the classes are equally frequent in the population. For example, the number
of insects caught in a trap in different times of a day, frequency of sighting an animal in different
habitats, etc. may be expected to be equal initially and subjected to the statistical testing. In such
cases, the expected frequencies are computed as:
62

n
E
k
Where, E = expected frequencies, n = total number of observed frequencies, k = number of
groups.

Example:
The Table below represents the number of insect species collected from an undistributed area at
a wildlife sanctuary in different months. Test whether there are any significant differences
between the numbers of insect species found in different months.

Solution:
The null hypothesis is, The diversity in terms of number of insect species is the same in all
months in the sanctuary. Expected frequencies derived with; total number of observed
frequencies, n = 804 and the number of groups are the twelve month, k = 12.

Computation of 2 using the data on number of species of insects collected in different months
Months O n (O E ) 2
E
k E
January 67 67 0.00
February 115 67 34.39
March 118 67 38.82
April 72 67 0.37
May 67 67 0.00
June 77 67 1.49
July 75 67 0.96
August 63 67 0.24
September 42 67 9.33
October 24 67 27.60
November 32 67 18.28
December 52 67 3.36
Total N = 804 804 134.84

Calculated 2 value is 134.84. From chi- square Table for (12 1) = 11 degree of freedom and
= 0.05, we get the critical value of 2 as 19.7. Therefore, we reject the null hypothesis and
conclude that the occurrence of the number of insect species in different months is not the same.
63

UNIT VII: WRITING THE RESEARCH REPORT

A report is simply a description of things that have already occurred. It is concise, clear
communication of the important findings of the research work. A thesis report conveys the
information to the examiner about the entire activities and findings of the student had
undertaken during the research process. The research report enables a student to demonstrate his
or her understanding and ability too, use techniques in methods of analysis and provides his or
her with the opportunity to significantly develop his or her management and research skills.

7.1 Purpose of Writing a Report


A research report serves the following purposes:
It is a means whereby the data, analysis and conclusions are placed in a organized form.
These information can be used both for academic and application purposes.
The thesis work of a student will be judged mainly by the quality of the report. The
examining committee may not see the effort in the field. Therefore, the student should
show his or her performance, skills and thoughts in his or her report, as there are vital to
its assessment and grading.
The effectiveness of report may be judged by its use. The organizations, professors,
researchers and students are using good reports for different purposes.
7.2 Contents and Styles of Report

1. INTRODUCTION
1.1 Background Information
1.2 Problem of the study
1.3 Objective of the study
1.4 Importance of the study
1.5 Limitation of the study
1.6 Organization of the study
2. REVIEW OF LITERATURE
2.1 Conceptual/Theoretical Framework
2.2 Review of related study
3. METHODOLOGY

3.1 Research design


3.2 Population and sample size
3.3 Sampling procedure
3.4 Data collection technique
3.5 Statistical procedure
3.6 Data analysis procedure
4. RESULTS AND DISCUSSION
4.1 Data presentation and analysis
4.2 Major findings of the study
5. SUMMARY, CONCLUSION AND RECOMMENDATIONS
5.1 Summary
5.2 Conclusion
5.3 Recommendation
References
Appendix
64

TABLE OF CONTENTS

Pages
Acknowledgements i
List of Tables ii
List of Figures ii
1. INTRODUCTION 1
1.1 Background Information 1
1.2 Problem of the study 1
1.3 Objective of the study
1.4 Importance of the study
1.5 Limitation of the study
1.6 Organization of the study

2. REVIEW OF LITERATURE 5
2.1 Conceptual/Theoretical Framework
2.2 Review of related study

3. METHODOLOGY
3.1 Research design
3.2 Population and sample size
3.3 Sampling procedure
3.4 Data collection technique
3.5 Statistical procedure
3.6 Data analysis procedure

4. RESULTS AND DISCUSSION


4.1 Data presentation and analysis
4.2 Major findings of the study

5. SUMMARY, CONCLUSION AND RECOMMENDATIONS


5.1 Summary
5.2 Conclusion
5.3 Recommendation

References
Aappendix

7.3 Types of Report: Descriptive and Analytical Report


7.3.1 Descriptive Report
Descriptive reports are mere description of facts or opinions gathered by the students during
his/her field study. The presentation of facts in an organized way may be of real value in
properly understanding the situation. These reports indicate the nature of problems facing the
organization under study and also indicate the reforms required to overcome the problems.
7.3.2 Analytical report
Analytical reports go one step further than descriptive reports. These reports in addition to
representing the facts and statistics, interpret these information in relation to the problem under
consideration. Focused on a single or limited area of the problem, these reports follow the
process of scientific investigation and reporting. These reports also recommend the actions to be
taken for improvements in the situation.
65

The student may choose one of these types of reports. However, the choice of the types would
depend on the nature of investigation. Some thesis report may include the features of both these
types of reports.
7.4 Presenting Data: Table and Figures in Report
A Table is a presentation of data in column form. The term Figure usually includes graphs, maps,
drawings and charts.

Tables
The presentation of Tables is concerned with the labeling techniques to make the content clear.
The following practices should be used in the Table constructions.
The Table should be numbered consecutively throughout the report. While referring to
the Tables, the student should write Table 12.1 instead of writing the following Table
or Table on the following page.
Each Table should be given a title that is complete of information and date of published.
The title of Table can be written in all capital letters for the emphasis.
Column of the Table should label clearly enough to identify the items. Column heading
should be short.
Rows in the Table can be arranged without any difficulty. In case of need, two lines can
be used for giving row headings.
If the source of information in the Table is to be identified, it should be placed below the
Table in the same format as the footnote. To number the Table footnotes, an asterisk
could also be used.
Sources of data should be cited in order that the source may be referred to if necessary.
The report writer can use abbreviations if necessary. These abbreviations should however
be explained in footnotes below the Table.
Table 12.1 Sectoral distribution of labor forces
Country Agriculture Industry Services
1965 1990 1965 1990 1965 1990
Indonesia 71 23 9 37 22 40
Malaysia 59 20 13 39 21 41
Thailand 82 17 5 32 13 51
Bangladesh 84 37 5 17 11 46
India 73 31 12 31 15 38
Nepal 94 60 2 - 4 -
Pakistan 60 26 18 25 22 49
Source: World Development Report, 1992

Figures
In reports graphic forms like bar charts, pie charts, maps and pictograms are often presented.
These presentation help clarify and understand data at glance. The report writers should ask the
following questions. They are:
To what extent does the graphic presentation contribute to the overall understanding of
the data? If graphic presentations do not assist the reader in understanding the subject,
these should not be used in the report.
Can the data convey the meaning to the reader in a better way using the graphic
presentation? Graphics often facilitate communication which is difficult to explain in
words alone.
Is the graphic easily understood?
66

Is the graphic honest? The hand is often quicker than the eyes; and data can be distorted
rather easily.
In a small report, the Figure should be numbered consecutively throughout the report and
referring is similar to Table.
7.5 Use of Quotations, Abbreviations
7.5.1 Use of Quotations
Some guidelines of quoting other writers ideas are:
Direct quotations would be used the words of the author add force to validate the
argument in the report.
Direct quotations may be used for presenting the major argument where a reference
would not be sufficed.
Direct quotations may be used when the writer wishes to comment upon, refute or
analyze ideas expressed by the author.
Direct quotations may be used when changes might cause misunderstanding of the
original idea or expression.
Direct quotations should be used when citing mathematical and other formulas.
Short Quotation (upto three lines)
Incorporate the quotation into sentence or paragraph, without disrupting the flow of the
text.
Use double quotation marks at the beginning and end of the quotation.
Use the same spacing as the rest of the text.
Example,
Most of the South Asian countries have initiated the far- reaching reforms like opening up their
economies, strengthening the market forces and providing a greater role for the private sector. In
Nepal also steps are being taken to foster the private sector by opening up new areas for
investment removing licensing and other controls on business and simplifying rules and
regulations (Gupta, 1994, p.14).
Long quotations (for four or more lines)
Use no quotation marks at the beginning and end of the quotation.
Use single line spacing.
Indent the quotation five spaces from the left margin.
Example,
Most of the south Asian countries have initiated far- reaching reforms like opening up
their economics, strengthening the market forces and providing a greater role for the
private sector. In Nepal also steps are being taken to foster the private sector by opening
up new areas for investment removing licensing and other controls on business and
simplifying rules and regulations (Gupta, 1994, p.14).
7.5.2 Abbreviations
A number of abbreviations also appear in the research report. These abbreviated words should be
arranged in ascending order under Abbreviations or Acronyms. Examples of some of the
frequently used abbreviation are given below:
ADB Asian Development Bank
bk., bks. book (s)
e.g. for example
ed. edition
eds editors
i.e. that is
IMF International Monitory Fund
N.B. nota bene
67

p., pp. page (s)


viz. namely

7.6 Rules of Typography (font, size, margin, spacing of paragraph etc).


Font: Times New Roman
Font size: 12
Paper size 8.6 x 11 inches and only side of paper
Margins Margins indicate the boundaries of the text. Margins of 1.6 inches on the left and
one inch on the right- hand side of the page are commonly used. The last line of
writing should be one inch above the bottom edge of the page.
Spacing The text of the report should be double- spaced. Indented quotation and footnotes
should be single- spaced. Same style and size of font should be used through out
the report. If you use indent in starting paragraph, there is no spacing between the
two paragraphs. If you are not using indent, second paragraph is starting by
leaving one space or single line.
Page number Page number should come at the top right hand corner of the page, one inch from
the top edge and 1.25 inches from the right edge. The first line of the text should
be two spaces below the page number.
Pagination Pages should be numbered consecutively in Arabic numerals from the first page
of the text to end of manuscript. The pages in the introductory sections (preface,
table of contents etc) should be numbered with small Roman numerals i, ii, iii,..
one inch from the bottom page.
Proof-reading: The manuscript should be read critically, searching for inaccurate statements,
wrong entries, omissions and inconsistencies. After verifying and locating errors
in quotations, footnotes, Tables, Figures, paragraphing sentence structure,
headings, spellings, style, bibliography, mark the copy to provide the typist with
necessary directions for providing a satisfactory transcript and must know the
symbols of proof- reading.
Citation
Referencing is a standard method of acknowledging sources of information and ideas that we
have used in research. Direct quotations, facts as well as ideas and theories from published and
unpublished works must be referenced. There are different methods of documenting (citing and
referencing) sources. The researchers commonly employ the following methods:
The Publications Manual of the American Psychological Association (APA) offers
detailed information regarding citations, quotations, references, and so on.
The Harvard System, was developed in the 1930s and has been updated periodically. It
usually, uses author- date system. This system is also widely used for citation and
references.
The Manual Style of the University of Chicago is also widely used for citations,
quotations and references.
The ACS Style Giude is published by the American Chemical Society. This style guide
explains how organized numbered references.

The documenting method has two parts: the citations and bibliography, whci may also be called
references.

APA Method of Citation


The APA method is widely used style of citing and referencing in the social science area. We
also recommend that this method should be followed.

Guidelines:
68

When quotation and paraphrase are cited, the authors name and year of publication
appear in the body of the report as follows:
For instance Paudyal (2000) stressed that business firm must think globally to remain
competitive.
When a work is authored by two individuals, the surnames of both the authors are cited.
For instance Peppard and Rawland (2001) define organizational culture as shared values and
beliefs, which take the form of rules of behaviour.
When a work has more than two authors, cite all authors the first time the reference
occurs, and subsequently include only the surname of the first author followed by et al.
For instance Shukla, Gupta, Mathema and Singh (2003) found that
Shukla et al. (2003) found that
The authors name and year of publication appear at the end when direct quotation is
made.
For instance Strategic processes are those processes by which the organization plans for and
develops in future (Chitrkar, 1997).
If the same author is cited in the same paragraph, the year of publication is not necessary
to be mentioned.
For instance, Panta (2001) wrote a book entitled assignment and Report Writing.
Pant found that
If no author is given for the work, treat the title as the author.
For instance, The specification writers job is likely to change (Style Manual, 1992).
When a works author is unknown, cite the text the word the Anonymous followed by the
comma and date: (Anonymous, 1982).
When more than one author has to be cited in the text, these should be in the alphabetical
order of the first authors surname and the citation should be separated by semi colons:
For instance, The concept of the job design (Maskey, 1987; Subedi, 1972; Uprety and Dhakal,
1976).
The page numbers, if required to be mentioned, should be cited as follows:
For instance, Pant (2001) argues that a research report can be worthless if interpretation is
faulty, even if valid and reliable data have been collected (p, 63).
Hypothesis testing is called deductive research (Wolf and Pant, 1999, p. 14).
When citing block quotations, do not place quotation marks before and after a block
quotation. Indent the left margin five spaces. Do not indent the right margin.
For instance, According to Berger (1991):
A PET bottle weighs approximately 91 grams. Of this 91 grams, only 63 grams or 69
percentage of the bottle is pure PET. The bottle has three other major components. The
bottle cap, usually made of aluminum, weighs roughly 1 gram or 1 percent weight. The
bottle label, usually made of paper, weighs 5 grams or 6 percent of the bottle weights (pp.
37- 38).
Personal communication through letters, memos, telephone conversation, and the like
should be cited in the text only and not included in the reference list since these are not
retrievable data. In the text, provide the initials as well as the surname of the
communicator together with the date.
For instance, A. Pradhan (Personal Communication, November 15, 1998) feels that
M. Neupane (Personal Interview, April 7, 1998) feels that
T. P. Mishra (Personal letter, Professor, Tribhuvan University, February 5, 1997)
argues that
69

APA References
The references list contains bibliographic information in each source you use. The list is
arranged alphabetically by the surname of the author.

Guidelines:
Present information for all entries in this order: Authors name. Date. Title. Publication
information.
Example,
Harrison, A. (1992). Just- in- time manufacturing in perspective. New York: Prentice Hall.
Peppard, J., and P. Ronald (1992). Business process re- engineering. New Delhi: Prentice Hall of
India.
Use only the initials of the authors first and middle names.
Capitalize only the first word and proper nouns and adjectives in the title.
Use double- space between the entries. The second line of an entry should be single-
spaced.
The secondline of an entry starts three spaces in from the left- hand margin.
Place the date in parenthesis immediately after the name.
Place the entry in alphabetical order.
If there are two or more works by one author, arrange them chronologically, most recent
last.
o Aryal, N. (2001)
o Aryal, N. (2006)
Some common entries for references
1. For book with one author
Kotler, P. (1998). Marketing management: Analysis, planning, implementation and control. New
Delhi, Prentice Hall of India.
Arnorld, J., I. T. Robertson, & C. L. Cooper, (1996). Work psychology:Understanding human
behaviour in the workplace. Delhi, Macmillon India.
2. For an edited book
Pant, P. R., & N. Mamandhar (Eds), (1998). Industrial relations in Nepal: A book of readings.
Kathmandu, FNJ and IRF.
Ross, A. M. (Ed), (1996). Industrial relation and economic development. London, Macmillon.
3. For an eassay in a book
Jyoti, P. (1998). Labor management: A critique. In: P. R. Pant and N. Manandhar (Eds),
Industrial relation in Nepal: A book of readings (pp. 297- 302). Kathmandu, FNJ and
IRF.
4. For an encyclopedia
Sonar. (1984). Encyclopedia Britannica.
5. For a journal article
Pant, P.R. (1993). Human resource development: A cross- country overview of the Asian region.
The Nepalese Management Review, IX (I): 1- 13.
6. For an article in monthly or weekly magazine
Shukla, M. (1985, February). Managing creative talents. Indian Management (pp. 3- 7).
7. For an unpublished materials
Joshi, M. (1992). Fine tea plantation: A case study. (n.p.).
Khetan, M. (1992). Motivation behind acquisition of time saving electrical appliances by
Nepalese households. Unpublished Masters thesis, central department of Management,
Tribhuvan University.
70

8. For a newspaper
The Rising Nepal. (1996, July 15). P.2.
9. For paper presented at seminar
Sharma, B. (1998, June). WTO: Its impact on the education actor. Paper presented st thr seminar
of Association of Business Students, Kirtipur, Nepal.
10. For personal e- mail
Gurung, R. K.(2004, February). Meditation program: A feedback. Available e-mail:
deanmagmt@healthnet.org.nep. [2004, February 7].
11. For journal articles on the Internet
Subba, A. (2003). The language management profession. The Nepal Review [online].
Available: <http:/www.nepre.org.np /6/2/>[2003, September].
12. For home page
Bhandari, C. (1999, March 26). Nepals culture and learning [Home Page of ABC Institute].
[online]. Available: <http:/www. ABC.co.np> [1999, March 28].
Differences between APA and Harvard System
Harvard system APA system Differences
Referencing in the text
(Lewis 2001) (Lewis, 2001) Note punctuation
(Saunders and Williams 2001) (Saunders & Williams, 2001) & not and; Note punctuation
(Williams et al. 1999) (Williams, Saunders & Staughter,For 1st occurance
1999)
(Williams et al. 1999) Williams et al., 1999) For subsequent occurrences; Note
punctuation
Referencing in the references or bibliography
Thornhill A, Lewis P, Thornhill, A., Lewis, P., Note commas and full stops
Millmore M and Saunders Millmore, M. & Saunders, Note of and, &
MNK (2000) Managing Change: A M.N.K. (2000), Managing change: Note use of order, use of colon,
Human Resource Strategy A human resource strategy comma and full stop
Approach, FT Prentice Hall, approach. Harlow: FT
Harlow Prentice Hall

Abstract, Summary, Conclusions and Recommendation


Abstract

Ethnicity and Rural Development:


A Sociology Study of Four Tharus Villages in Chitwan, Nepal

Abstract

The objectives of the study were: 1) to investigate the effect of different situations in rural Nepal
on ethnicity, and 2) the relationship between ethnicity and various structural variables relevant to
process of rural development. It focuses on the Tharus people, an ethnic group indigenous to the
Terai region of south central Nepal.

Two matching pairs of Tharu communities were surveyed; on pair in a more centrally- located
area (consisting of predominantly Tharu village and ethnicity diverse village where Tharus are
in numerically minority) and a similar matching pair in a more remote area. Data were obtained
through a general census of 260 households in the four villages, and direct interviews of sample
of 100 Tharu household heads (25 from each villages) plus non- Tharus from the nearby diverse
village.
71

Supplementary information was obtained from informants, documents and through observation
(i.e. living in the villages while interviewing).

Ethnicity viewed as both structural variable (composition of the village) and a behavioural
variable (manifestations of ethnic identity and loyalty). The latter is measured by a seven points
Guttam- Scale.

The locational factor (relative centrally or remoteness of village) emerged as a more important
explanation of ethnic behavior than the ethnic composition of village per-se. No relationship is
found between ethnicity and adoption of improved farm practices; it appears that ethnicity and
adoption with regards to these Tharus communities are independent phenomenon.

It is concluded that ethnicity contrary to what is often argued, does not stand in the way rural
development not does it hinder social interaction. Indeed, through the maintenance of group
solidarity, the motivations of underprivileged minority groups as the Tharus of Nepal are
strengthened and they are better able to participate in the process of nation- building.

Questions: Identify the following areas from the above abstract of PhD Thesis:
1. Problem of research
2. Objectives
3. Methodology used
4. Dependent and independent variables
5. Findings of research
Above five things must be included in abstract. If you are able to solve above five questions, you
will not have problem of writing abstract.

Summary
The current section is most important in that it relates the data findings a logical, rational fashion
to the problem area and research question within the framework of the theory structure
established in the Introduction. The relevance of the hypothesis to the theory and its contribution
to the solution of the problem investigated are presented.

It is vital that the researcher reviews his or her in introduction section before attempting to write
this section of the study to insure that he or she has a solid grasp of what it was, he or she set out
to investigate, and the theory underlying the study.

This may have a number of subheadings if several questions and/or hypothesis were
investigated.
Conclusions
As a follow- up to the interpretation of the data, conclusion is made as to how well they assist in
attempting to solve the problem, and how well the purpose of the study was accomplished. It is
at this point in the study that the researcher should feel free to express his or her own
conclusions and generalizations.
Recommendations
If decisions of any sort, either theoretical or operational, were contingent upon the outcome of
the study, it is at this point that recommendations by the researcher are made. Justification for
those recommendations presented in the previous section the report.

It is fairly traditional that the researcher indicates that his or her study has accomplished its
purpose has made a contribution, and that further investigation in the area is needed.
72
73

Additional Note for Hypothesis Testing: Z-Test

Decision making about the characteristics of the population on the basis of study of the sample
taken from the population involves the risk of taking wrong decision. For instance, we may want
to decide which of the two brands of the fertilizers is more effective. The probability theory
plays a vital for decision making for this situation. The method of statistics which helps in
arriving the decision is called Test of Hypotheses or Hypotheses Testing or Test of Significance.

The test of hypotheses is a process of testing of significance the parameter of population on the
basis of sample drawn from the population.

General procedure of testing a Hypothesis:

1. Formulating the hypothesis: In statistics, by a hypothesis, we mean a tentative conclusion


logically drawn regarding any parameter of the population,
The average height of students in the class is 160 cm.
A given drug cures 90% of the patient taking it.
The common way of stating a hypothesis is that there is no difference between the
population mean and sample mean. The statistical hypothesis may be divided into two types:
Null Hypothesis and Alternative Hypothesis.

Null hypothesis:
The purpose of possible acceptance is called a null hypothesis and denoted by H0. In above
case,
H0 : = 160 cm.

H0 : = X

Alternate Hypothesis
Any hypothesis which is complementary to the null hypothesis is called an alternative
hypothesis and denoted by H1 . For example, it we want to test the null hypothesis that average
height of the student is 160 cm i.e. H : = 160 cm = 0
then the alternative hypothesis could be
(i) H1: 0 (i.e. 0 or < 0 )
It is called two- tailed alternative hypothesis.
(ii) H1: > 0
It is called right- tailed alternative hypothesis.
(iii) H1: < 0
It is called left- tailed alternative hypothesis.
We formulate a null hypothesis (H0) and alternative hypothesis (H1) which we accept when H0
must be rejected.

2. Computing the test statistic: After formulating the hypothesis the next step is to calucuate an
appropriate test statistic which is based on an appropriate probability distribution. For testing
whether the null hypothesis should be accepted or rejected, we use Z- distribution under
normal curve for large sample (n 30) and t- distribution for sample (n < 30).

3. Defining Type- I Error () and Type- II Error (): A decision to accept ot reject H is made
on the basis of the sample data and there is always a chance of making an error.
Type- I Error: Reject a null hypothesis when it is true.
Type-II Error: Accept a null hypothesis when it false.
74

The four situations which may arise in testing hypothesis are given in the following Table.
Actual Decision
Accept H0 Reject H0
H0 is true No Error Type II Error
Correct Decision Wrong Decision
Probability = 1 - Probability =
H0 is false Type II Error No Error
Wrong Decision Correct decision
Probability = Probability = 1-
4. Fixing the level of significance: The level of significance denoted by , is maximum
probability of making type- I error. The commonly use level of significance are 5% (0.05)
and 1% (0.01). If we use 5%, it implies that in 5 cases out of 100 cases, we likely to reject
H0. In other words, this implies that we are 95% confident that our decision is to reject H0 is
correct. The level of significance should be fixed in advance before applying the test.
5. Finding critical region: This region is the region of acceptance and rejection. When the
computed test statistic lies in this region or out of this region for accept or reject the Null
Hypothesis. This region is generally taking from the Z- distribution for large sample.
6. Deciding the Two- Tailed or One- Tailed Test: The critical region may be represented by a
portion of area under the normal curve in two ways:
(a) Two tails or two side under the curve.
(b) One tail or one side under the curve which is either left tail or right tail.

Two- Tailed Test


Lower Critical Value
Upper Critical Value

Rejection Rejection
region (/2) Acceptance Region
(1 ) region (/2)

- Z Z= 0 + Z

Right- Tailed Test Left- Tailed Test


(Level of significance ) (Level of significance )
Rejection Rejection
region () region ()

Acceptance Region Acceptance Region


(1 ) (1 )

Z= 0 + Z - Z Z= 0
Figure 1. Critical region

The test of hypothesis which is used on critical region represented by both the tails under the
normal is called two- tailed test or one- tailed test (Figure 1). When the sample mean is
significantly different from the population mean a two- tailed test is used. When the population
mean is considered to be at least as large as some specified value used right- tailed test; whereas
75

when the population mean is considered to be at least as small as some specified value the left-
tail test is used.
7. Making decision: If the calculated value of test statistic (T) is less than critical value which
lies in acceptance region and H0 is accepted. On the other hand, calculated value is greater
than critical value, the value lies in region of rejection and H 0 is rejected. In testing the
hypothesis we generally used 5% level of significance.
If T T , accept H0.
If T > T , reject H0.
Summarizing the hypothesis- testing step as follows:
Step 1. State the null (H0 ) and alternate ( H1 ) hypothesis
2. Select and compute the test statistic T.
3. Obtain the significance level and identify the acceptance and rejection region.
4. Identify the type of test one tailed or two tailed tests to be applied.
5. Make the decision by comparing T and critical values.
Test of Hypothesis for Large samples

a. Test of significance of single mean:


The steps in testing the significance of sample mean of large sample (n 30) are as follows:

Step 1. Formulate the null hypothesis (H0) in any way of the following forms.
(i) H0 : = 0 (i.e. population mean has specified mean 0)

(ii) H0 : there is no significant difference between sample mean ( X ) and population
mean ()
(iii) H0 : The sample has been drawn from the given large population with mean 0 and
standard deviation .

X
Z
Step 2. Compute the test statistic,
n
Z- N (0,1) i.e. Z follows the normal distribution with the mean 0 and the standard
deviation 1.
If is not known the its estimate is given by sample standard deviation by

1
S
n 1
( X X )2

where S = is an unbiased estimate of the population Standard Deviation (SD).
Step 3. Select the level of significance,
We usually fix at 5% (0.05) unless otherwise stated.
Decide whether two- tailed test or one- tailed test has to be applied
Step 4. Write down the critical or tabulated value of Z at predetermined level of significance, .
Critical values Level of significance
Z 1% 5%
Two- tail test 2.58 1.96
One- Tailed test: Right + 2.326 + 1.645
Left - 2.326 - 1.645

Note that the critical value of Z for one-tail test at the level of significance is the same
as the critical value of Z for two-tailed test at the level of significance of 2.
76

Step 5. Make the decision by comparing the computed and the critical (or tabulated) values of Z.
If the computed value of /Z/ (only positive values) is less than critical value Z, it lies in
the acceptance region. Hence H0 is accepted at the level of significance and we
conclude that there is significance difference between sample mean and population mean
or the sample has been drawn from the given population.

If the computed value of /Z/ is greater than Z, H0 lies in the rejection region and we
reject H0. It may be concluded that there is a significant difference between the sample
mean and the population mean or sample has not been drawn from parent population.
Example 1:
A sample of 50 pieces of certain types of string was tested. The mean breaking strength turned
out to be 14.5 kg. Test whether the sample is from a batch of strings having a mean breaking
strength of 15.6 kg and standard deviation of 2.2 kg.
Solution:
Given,
Sample size, n = 50

Sample mean X =14.5 Kg
Population mean = 15.6. Kg
Standard deviation = 2.2 Kg
Null Hypothesis, H0 : = 15.6 Kg
Alternative Hypothesis, H1 : 15.6 kg (two tailed test)
Computation of test statistic,

X
Z

n
14.5 15.6 1 .1 1 .1
3.55
= 2. 2 2 . 2 0.31
50 7.142
/Z/ =/ - 3.55/ = 3.55
Tabulated value of Z for two tail test at 0.05 level of significance is 1.96.
Decision: Since the computed value of /Z/ is greater than it tabulated value, H 0 is rejected i.e.
the sample has not been drawn from the normal population with the mean 15.6 kg and standard
deviation 2.2 kg.
Example2. The mean lifetime of a sample of 100 fluorescent light tubes produced by a company
is computed to be 1570 hrs with a standard deviation of 120 hrs. The company claims that the
average life of the tubes produced by it is 1600 hrs. Is the claim acceptable by using a level of
significance of 0.05.
Solution:
Here,
n = 100

X = 1570
S = 120 hrs
= 1600 hrs

H0: = 1600 hrs, i.e. the average life of tubes produced is 1600 hrs.
H1: 1600 hrs i.e. the average life of the tubes is not 1600 hrs (Two- tailed Test)
77


X 1570 1600 30
Z 120 2.5
Computation of test statistic: S
12
n 100
/Z/ = /-2.5/ = 2.5
Tabulate value of Z for two tail test at 0.05 level of significance is 1.96.
Decision: Since the computed value of /Z/ is greater than its tabulated value, H 0 is
rejected i.e. the average life of tubes produced by the company is not 1600 hrs and hence
the companys claim is not acceptable.

Example 3. A sample of 400 students is found to have a mean height of 171.38 cm. can it be
reasonably regarded as a sample from a large population with mean height 171.17 cm and SD
3.30.
Level of significance: 1% 2% 5% 10%
Tabulated value 2.58 2.33 1.96 1.65
Solution:
Given,
n = 400

X = 171.38
= 3.30
= 171.17
Computation of test statistic:

X 171.38 171.17 0.21
Z 3.30 1.27

0.165
n 400
Since the computed value of Z (1.27) is less than its tabulated values at 1%, 2%, 5% and 10%
level of significance, H0 lies in the acceptance region and here H0 is accepted at these levels.

Example 4. In the past a blending process has produced an average of 5 kg of waste material for
every batch, with SD of 5 kg. From a sample of 100 batches an average of 16 kg of waste per
batch is obtained. At 5% level of significance is it reasonable to believe that the average has
increased?
Solution:
Given,
n = 100

X = 16 kg
= 5 kg
= 5 kg
H0 : = 5 kg i.e. the average of waste material has not been increased.
H1 : > 5 kg i.e. the average has been increased. (Right- tailed test)

X 16 5 11 22
Z 5
Computation of test statistic: 0 .5
n 100
Tabulated value of Z for right tailed test at 5% level of significance is 1.645.
Decision: Since the computed value of Z is greater its tabulated value, H 0 is rejected and the
average of waste material has been increased.
Example 5. An insurance agent has claimed that the average age of policyholder who ensure
through him is less than the average for all agents, which is 30.5 years. The mean age obtained
78

from a random sample of 100 policyholders who has insured through him is 28.8 years with
standard deviation of 6.35 years. Test his claim at the 5% level of significance.

Solution:
Given,
n = 100

X = 28.8 yrs
S = 6.35 yrs
= 30.5 yrs

H0 : = X = 30.5 i.e. the sample mean X and population mean () do not differ.
H1 : < 30.5 years (left tail test)

X 28.8 30.5 1.7 2.677
Z 6.35
Computation of test statistic: S 0.635
n 100
/Z/ = /-2.677/ = 2.677
Tabulated value of Z for left tailed test at 0.05 level of significance is 1.645.
Decision: Since the computed value of /z/ is greater than its tabulated value. H0 is rejected i.e.
the insurance agents claim that the average age of policyholders who insure through him is less
than the average for all agents accepted with 95% confidence.

b. Test of significance of difference between two means


The step in testing significance of difference between two means fro large sample (n30) are as
follows:
Step 1. Formulate null (H0) hypothesis and Alternative Hypothesis (H1).
H0 : 1 = 2
i.e. two samples have been drawn from the same parent population or two samples have
been drawn from two different parent populations have the same mean and standards.
H1 : 1 2 (Two- tailed test)
Step 2. Under H0, compute the test statistic

(X1 X 2 )
Z . (1)
2 2
1 2

n1 n2
where n1 30 and n2 30, Z- N(0, 1). If 1 and 2 are not known then their
2 2

2 2
estimates are provided by the corresponding sample variances S1 and S 2 ,
respectively. Equation (1) reduce to:

(X1 X 2 )
Z . (2)
2 2
S1 S 2

n1 n2
Note that S and S are unbiased estimate of population SDs. If two samples are drawn
from the same population then 1 = 2 and 1 = 2 = says, then
2 2 2
79


(X1 X 2 )
Z
2 2
1 2

n1 n2

( X1 X 2 )
Z (3)
1 1
( )
2

n1 n 2
Step 3. Select the level of significance, . The level of significance which indicates whether the
probability of difference of means is small or large, is generally fixed at 5% (0.05).
Step 4. Write down the critical value of Z at predetermine level of significance .
Z0.05 = 1.96 For two-tailed test
Z0.05 = 1.65 For right-tailed test
Z0.05 = - 1.645 For left-tailed test
Step 5. Make the decision by comparing computed and tabulated value of Z. If the computed
value of Z is less than its tabulated value Z, the null hypothesis H0 is accepted otherwise
H0 is rejected.

c. Test of significance of difference between two Standard Deviation for large


samples:
The steps for the test of significance of difference between two standard deviation are as follows
Step 1. Formulate the null (H0) and alterative (H1) hypothesis.
H0 : 1 = 2 i.e. sample SDs do not differ significantly. Two samples are drawn from the
same population.
H1 : 1 2 i.e. the sample SDs differ significantly. Two samples are not drawn from the
same population (Two- tailed test).
Step 2. Compute the test statistic given by
( S1 S 2 ) ( S1 S 2 )
Z
1 2
2 2 2 2
S1 S2

2n1 2n2 2n1 2n2
S1 and S2 are unbiased estimate of 1 and 2, respectively. Z- N(0, 1).
Step 3. Select the level of significance, (0.05).
Step 4. Write the critical value Z, according to the type of tests- two- tail and one-tail.
Step 5. Make the decision. If Z<Z, accept H0. If Z> Z , reject H0 and accept H1.

Example 6: In a survey of incomes of two classes of workers, two random samples gave the
following details. Examine whether the different between the (i) means (ii) the SDs, are
significant.

Sample Size Mean annual income SD (in Rs.)


I 100 Rs. 582 24
II 100 Rs. 546 28

Solution,
n1 = 100 n2 = 100

X 1 = Rs. 582 X 2 = Rs. 5546
S1 = Rs. 24 S2 = Rs. 28
80

(i) H0 : 1 = 2 i.e. there is no difference between sample means


H1 : : 1 2 i.e. there is a significance difference between sample means. (Two-tailed
test)

Computation of test statistics:



(X1 X 2 ) (582 546) (582 546)
Z = (582 546) = 36
2
S1 S 2
2 242
28 = 576 784 =
2
3.69
=9.76
5.76 7.84
100 100 100 100
n1 n2
Tabulated value of Z at 0.05 level of significance is 1.96.
Decision: Since the computed value of Z is greater than tabulated value the null hypothesis is
rejected i.e. there is significance difference between the two sample means.

(ii) H0 : 1 = 2 i.e. there is no significance difference between the sample standard


deviations.
H1: 1 2 i.e. there is a significance difference between the sample standard
deviations (two- tailed test)
Computation of test statistic:
( S1 S 2 ) (24 28)
4 4
24 2 28 2
2 2
Z S1 S2 1.35
2.88 3.92 2.61
2n1 2n2 200 200
Tabulate value at 0.05 level of significance is 1.96.
Decision: Since the computed value of Z less than its tabulated value, the null hypothesis is
accepted i.e. there is no significant difference between the two sample standard deviation.

Example 7. A potential buyer of light bulbs bought 50 bulbs each of two brands. Upon testing
these bulbs, he found that brand A had a mean life of 1282 hours, with SD of 80 hours whereas
brand B had a mean life of 1208 hours with SD of 94 hours. Can the buyer be quite certain that
the two brands do differ in quality?

Solution:
Given,
Brand A Brand B
n1 = 50 n2 = 50

X 1 = 1282 hrs X 2 = 1208 hrs
S1 = 80 S2 = 94
2 2
S1 80 x80 S2 94 x94
128 176.72
n1 50 n2 50
H0 : 1 = 2 i.e. there is no significant difference between the two brands in quality.
H1 : 1 2 i.e. the two brand differ in quality (two tailed test)

Computation of test statistic:


81


(X1 X 2 ) 1282 1208
Z
74

74
4.24
2 2
S1 S 2 128 176.72 304.72 17.456

n1 n2
Tabulated value of Z at 0.05 level of significance is 1.96.
Decision: Since the computed value of Z is greater that its tabulated value, the null hypothesis
H0 is rejected i.e. there is significance difference between the two brands of bulbs in quality.

d. Test of significance of a sample proportion


The steps in testing the significance of sample proportion or percentage are as follows:

Step 1. Formulate the null (Ho) and alternative (H1) hypotheses.


Ho : P=Po
i.e. , the sample has been drawn from a population with proportion P.
H1 : P Po (two tail test)
Step 2. Compute the test statistics is
pP
Difference
Z= = PQ
S .E
n
Where P = population proportion, Q = 1 P
p = sample proportion
P (1 P ) PQ
S.E (p)
n n
Note that if we have sampling from a finite population of size N, then
N n P (1 p ) N n PQ
S .E.( p ) =
N 1 n N 1 n
Step 3. Obtain the level of significance be = 0.05 unless otherwise stated.
Step 4. Write down the critical value Z for two tail test.
Step 5. Make the decision. If Z < Z accept H0. If Z > Z reject H0 and accept H1 at level of
significance.

Example 8: A furniture store with a loose credit policy expects that 8% of its credit accounts
will default on payments. Looking at the 500 accounts sold to last year, however we see that 49
have default on payments. Looking at the 500 accounts sold to last year, however we see that 49
have defaulted. Using a 5% level of significance do you think the store has reason to believe that
estimate of 8% is too low ?
Tabulated value for 5% and 10% level of significances are 1.96 and 1.645, respectively.

Solution:
H0 : P = 0.08 i.e. the proportion of the defaulted credit accounts on payments is 8%.
H1: P 0.08 (two tail)
pP
Computation of test statistic : Z = PQ / n
49
p = sample proportion of defective credit accounts = = 0.098
500
82

P = 0.08, Q = 1 0.08 = 0.92


0.098 0.08 0.018
Z= = = 1.488
0.08 x 0.92 / 500 0.0121
Tabulated value of Z at 5% and 10% levels of significance are 1.96 and 1.645, respectively.
Decision : Since the computed value of Z is less than tabulated values at 5% and 10% levels of
significance, the null hypotheses H0 is accepted i.e. the store has not reason to believe that
estimate of 8% is too low.

Example 9: A manufacture claimed that at least 90% of the machine parts that it supplied to a
factory conformed to specifications. An examination of 200 such parts revealed that 160 parts
were not faulty. Determine if the manufactures claim is legitimate at 1% level of significance.

Solution:
H0 : P = 0.90 i.e. the proportion of the machine parts confirming to specifications in the
lot is 90%.
H1 : P < 0.90 (left tail)
pP
Computation of test statistic: Z PQ / n
p = sample proportion of machine parts conforming to specifications
160
= = 0.80
200
P = 0.90, Q = 1 P = 1 0.90 = 0.10
0.80 0.90 0.10
Z= 0.90 x 0.10 / 200
=
0.0212
= - 4.72

Tabulated value of Z at 1% level of significance for left tailed test is 2.326.


Decision: since the computed value of /Z/ is greater than the tabulated value, the null hypothesis
is rejected at 1% level of significance i.e. the proportion of the machine parts conforming to
specifications in the lot is no 90%, it is less than 90%.

Example 10: A sample of 600 persons selected randomly from a large city gives the result that
males are 53%. Is there reason to doubt the hypothesis that males and females are in equal
number in a city?

Solution:
H0 : P = 0.5 i.e. both males and females are equal in numbers in the city.
H1 : P 0.5 i.e., both males and females are not equal in numbers in the city (Two tail
test)
pP
Computation of test statistic: Z = PQ / n
p = Sample proportion of males = 53% = 0.53
0.53 0.50 0.03 0.03 0.03
Z= 0.5 x 0.5 / 600
=
0.25 / 600
= =
0.000417 0.02
= 1.5
Tabulated value of Z for two tail test at 5% level of significance is 1.96.
Decision: Since the computed value of Z is less than tabulated value the null hypothesis is
accepted i.e. males and females are in equal numbers in city.

e. Test of significance Between Two Sample Proportions


The steps in testing the significance of difference between two sample proportions are:
Step 1. formulate the null (H0) and alterative (H1) hypotheses
83

H0 : P1 = P2
i.e., there is no significance difference between the two population proportions or
percentages or, the two samples have been drawn from the same population.
H1 : P1 P2 (two tailed test)
Step : 2. compute the test statistic. Under H0, the test statistic is given by
P1 P2
Z= P1Q2 P2 Q2

n1 n2
P1Q1 P2 Q2
Where S.E. (p1 p2) =
n1 n2
Q1 = 1 P1, Q2 =1 - P2

Note that if P1 and P2 are not known, each of them is estimated by


^ n1 p1 n2 p2
P
n1 n2
And the test statistic is given by

p1 p2 p1 p2
Z
^ ^ ^ ^

PQ PQ PQ 1 1
= ^^

n1 n2 n n 1 2

Also not that the sample proportions can be substituted for their corresponding
population proportions,
p1 q1 p 2 q 2
Hence S.E. (p1 p2) =
n1 n
where, q1 = 1- p1 and q2 = 1 p2
Step 3. Fix the level of significance (usually = 0.05)
Step 4. Write down the tabulated value of Z at the level of significance .
Step 5. Make decision. If the computed value of Z is less than its tabulated value Z, H0 is
accepted otherwise it is rejected.

Example11. Random samples of 200 bolts manufactured by machine A and 100 bolts
manufactured by machine B showed 19 and 5 defective bolts respectively. Is there a significant
difference between the performances of the two machines?

Solution:
H0 : P1 = P2 i.e. there is no significant difference between the performance of the two
machines.
H1 : P1 P2 (two tailed)

Computation of test statistic:


84

p1 p2

Z ^^ 1 1
PQ n1 n2
P1 = Proportion of defective in the machine A = 19/200 = 0.095
P2 = Proportion of defective in the machine B = 5/100 =0.05

n1 p1 n2 p2 200 .095 100 .05 19 5


^

P n1 n2 200 100 300 0.08


^ ^
Q 1 P = 1- 0.08 = 0.92

0.095 0.05
0.045 0.045
Z= 0.08 0.92
1

1
0.01104

0.0332
1.355
200 100
Tabulated value of Z at 5% level of significance for two tailed test is 1.96.
Decision: since the computed value of Z is less than tabulated value at 5% level of significance
the null hypothesis is accepted i.e. there is no significant difference between the performances of
the two machines.

Example13. At a certain date in a large city 400 out of a random sample of 500 men were found
to be smokers. After the tax on tobacco had been heavily increased another random sample of
600 men in the same city included 400 smokers. Was the observed decrease in the proportion of
the smokers significant? Test at 5% level of significant.

Solution:
H0: P1 = P2, i.e. there is no significant difference between the proportion of the smokers
before and after the increase in the tax.
H1 : P1 >P2 (Right tailed)
Computation of test statistic:

p1 p2

Z ^^ 1 1
PQ n1 n2
400
P1 = Sample proportion of smokers before increase in tax.= 0.8
500
85

400
P2 = Sample proportion of smokers after increase in tax. = 0.667
600

n1 p1 n2 p2 500 .800 600 0.667 8


^

P n1 n2 500 600 11
^
8 3
Q 11 11
1

0.8 0.667
Z= 8 3 1

1

11 11 500 600
0.133 0.133 0.133
4.926
= 24 11 264 / 363000 0.027

121 3000

Tabulated value of Z at 5% level of significance for right tail test is 1.645.


Decision: Since the computed value of Z is greater than the tabulated value the null hypothesis
is rejected i.e. there is a significant decrease in the proportion of the smokers after the increase in
the tax.

Example12. A sample poll of 300 voters from district A and 200 from district B, showed that
56% and 48% respectively were in favor of a given candidate At a level of significance of 0.05,
test the hypothesis that
(a) There is a difference between the districts
(b) the candidate is preferred in district A
Solution:
H0 : P1= P2 i.e. there is no significance difference between the district.
H1: P1 P2 i.e. there is a significant difference between the districts.
Computation of test statistic:

p1 p2

Z ^^ 1 1
PQ n1 n2
P1= Proportion of voters of district A who are in favor of candidate = 0.56
P2 = Proportion of voters of district B who are in favor of candidate = 0.48

n1 p1 n2 p2 300 0.56 200 0.48 168 96


^

P n1 n2 300 200 500 0.528


86

^ ^

Q (1 p)1 .0 528 .0 472


0.56 0.48 0.08
1.75
Z= 0.528 0.472
1

1

500
300 200
(a) Tabulated value of Z at 5% level of significance for two tail test is 1.96.
Decision: Since the computed value of Z is less than tabulated value, we accept the null
hypothesis i.e. there is no significant difference between the districts.
(b) Tabulated value of Z at 5% level of significance is 1.645 for one tail test. If we want to
determine whether the candidate is preferred in district A, alternative hypothesis must be
(H1: P1>P2) used which involves a one tailed test.
Decision: On the basis of a one tail test at 5% level of significance, we reject the null hypothesis
because computed value of Z is greeted than tabulated value of Z and it may be concluded that
the candidate is preferred in district A.
87

Exercise for practices

1. Explain the general procedure followed in testing a hypothesis. Point out the difference
between one-tail and two-tail tests.
2. Distinguish between
(a) Null hypothesis and alternative hypothesis
(b) Critical region and acceptance region.
(c) Type I and type II errors.
(d) Level of significance and confidence limits.

3. (a) How do you test the significance of the difference between the means of the two
large samples.
(b) Explain the procedure for Z-test.
4. An auto company decided to introduce a new six cylinder car whose mean petrol
consumption is claimed to be lower than that of the existing auto engine. It was found
that the mean petrol consumption for the 50 cars was 10 km per liter with a standard
deviation of 3.5 km per liter, test for the company at 5% level of significance, whether
the claim the new car petrol consumption is 9.5 km. per liter on the average is
acceptable.
5. A random sample of size 100 from a large population give the distribution
Value: 10-20 20-30 30-40 40-50 50-60
Frequency: 13 20 45 13 9
Test the hypothesis that this sample comes from a population with mean 40, given that
the population SD is 10.
6. A sample of size 400 was drawn and the sample mean was found to be 99. test whether
this sample could have come from a normal population with mean 100 and variance 64 at
5% level of significance.
7. A random sample of 200 tins of coconut oil gave an average weight of 4.95 kgs. With a
standard deviation 0.21. Do we accept the hypothesis of net weight of 5 kgs. Per tin, at
1% level?
8. A pharmaceutical firm maintains the mean time for a drug to take effect is 24 minutes. In
a sample of 400 trails the mean time is 26 minutes with a standard deviation of 4 minutes
against the alternative hypothesis that it is not equal to 24 minutes. Use 0.05 level of
significance.
9. an ambulance service claims that it takes, on the average, 9.89 minutes to reach its
destination in emergency calls. To check on this claim, the agency which licenses
ambulance services has them timed on 50 emergency calls, getting mean of 9.3 minutes
with a standard deviation of 1.8 minutes. At the level of significance 0.05, does this
constitute evidence that the figure claimed is too low ?
10. a sample of 100 iron bars is said to be drawn from a large number of bars whose lengths
are normally distributed with mean 4 ft. and standard deviation 0.6 ft. if the sample mean
is 4.2 ft., can the sample be regarded as a truly random sample ?
11. intelligence test on two groups of boys and girls gave the following results:
Mean SD N
Girls 75 15 150
Boys 70 20 250
Is there a significant difference in the mean scores obtained by boys and girls?

12. The average height of 50 students who showed athletic interest was 68.2 inches with a
standard deviation of 2.5 inches, while another set of 50 students who showed no athletic
88

interest had the average height of 67.5 inches with a standard deviation of 2.8 inches.
Test the hypothesis that athletic interest makes a students taller.

Level of significance 1% 2% 5% 10%


Tabulated value 2.58 2.33 1.96 1.65

13. In a survey of buying habits, 400 women shoppers are chosen at random in super market
A located in a certain section of a city. Their average monthly food expenditure is Rs.
250 with a SD of Rs. 40. for 400 women shoppers chosen at random in super market B in
another section of the city, the average monthly food expenditure is Rs. 220 with a SD of
Rs. 55 Test at 1% level of significance whether the average food expenditure of the two
population of shoppers from which the samples were obtained are equal.
14. The mean produce of wheat of a sample of 100 fields is 200 lbs. per acre with a SD of 10
lbs. another sample of 150 fields gives the mean at 220 kbs. With SD of 12 lbs. assuming
the SD of the mean field at 11 lbs for the universe, find at 1% level if the two results are
consistent.
15. A random sample of 100 mill workers at Birgunj showed their mean wage to be Rs. 350
per month with a SD of Rs. 28. A sample of 150 mill workers in Nepalgunj showed the
mean wage to be Rs. 390 per month with a SD of Rs. 40. On the basis of the data, would
you say the mean wages of moll workers in Nepalgunj are higher than those at Birgunj.
16. The mean population of a random sample of 400 villages in district A was found to be
400 with a SD of 12. The mean population of a random sample of 400 villages in district
B was found to be 395 with a SD of 15 Is the difference between the two means
statistically significant?
17. A campus conducted both day and evening classes intended to be identical. A sample of
100 day students yields examination results as under:
X 1 = 72.4, 1 = 14.8
A sample of 200 evening students yields examination results as under:
X 2 = 73.9 2 =17.9
Are the two means statistically equal?
18. Two randomly selected groups of 50 employees each of a very large firm are taught an
assembly operation by two different methods and then tested for performance if the fist
group averaged 140 points with a SD of 10 points while the second group averaged 135
points with a SD. of 8 points, test at 0.05 level whether the difference between their
mean scores is significant?
19. You are working as a purchase manager for a company. The following information has
been supplied to you by two manufacturers of electric bulbs:
Company A Company B
Mean life in hours: 1300 1248
Standard deviation in hours: 82 93
Sample size: 100 100
Which brands of bulbs are going to purchase if you desire to take a risk of 5%.
20. A coin was tossed 400 times and head turned up 216 times. Test the hypothesis that the
coin is unbiased.
21. In 600 throws of six faced die, odd points appeared 360 times. Would you say that the die
is fair at 5% level of significance?
22. In a random sample of 1000 persons from a village in district, 560 are found to be rice
eaters and the rest wheat eaters. On the basis of this data, can it be concluded that both
the food articles are equally popular?
23. A manufacturer claimed that at least 95% of the equipment which he supplied to a
factory conformed to specifications. An examination of a sample of 200 pieces of
89

equipment revealed that 18 wee faulty. Test his claim at a significance level of (a) 0.01
(b) 0.05.
24. An auditor claims that 15% vouchers of a firm contain some mistakes or the other. A
random sample of 500 vouchers is taken. If 100 vouchers are found to contain a mistake,
do you believe the claim of the auditor to be correct. Show you calculations in arriving at
this conclusion.
25. An airline must allocate available seating space between first class passengers and
economy class passengers. The null hypothesis is that 20% of the passengers fly first
class, but the management recognizes the possibility that the percentage could be more
or less. A random sample of 400 passengers includes 70 passengers holding first class
tickets. Can the null hypothesis be rejected at 10% level of significance?
26. The owner of a wholesale distributing firm would like to know the proportion of
accounts receivable that are more than 60 days past due. The owner estimates that in the
past the proportion has remained stable at 15%. A random sample of 200 current
accounts receivable revealed that are more than 60 days past due has changed.
27. In a simple random sample of 600 men taken from a big city 400 are found to be
smokers. In another simple random sample of 900 men taken from another city 450 are
smokers. Do the data indicate that there is a significant difference in the habit of smoking
in the two cities?
28. A machine produced 20 defective articles in a batch of 500. After overhauling it
produced 3 defective articles in a batch of 100. Has the machine improved?
29. In a sample of 300 units of a manufactured product, 65 units were found to be defective
and in another sample of 200 units, there were 35 defective. Is there significant
difference in the proportion of defectives in the samples at 5% level of significance?
30. In a village A out of a random sample of 1000 persons 100 were found to be
vegetarians while in another village B out of 1500 persons 180 were found to be
vegetarians. Do you find a significance difference if the food habits of the people of the
two villages?
31. A certain manufacturing process is expected to produce only 1% defective items. In one
sample of 134 of these items, 3 were found to be defective, and in another sample of size
66, 2 were found defectives significant?
32. Random sample of 100 bolts manufactured by machine A and 50 bolts by machine B
showed 10 and 6 defective bolts respectively. Is there a significant difference in the
performance of the machines at 5% level o significance?