Prezentacija

Research
questions
Research defined
 We can define research as an activity of systematic enquiry

that seeks answers to a problem.
Research in our everyday
lives
 We all encounter research in our daily lives. The results of
research on many topics are presented to us in the form of
newspaper articles, books, reports, and television
programmes. For example, crime level figures are presented
to us by television news reports and some topics, such as
diet and health, are very popular with magazines as well as
television programmes.
lives
 Thus, through various media we have become accustomed
to seeing, reading and hearing about research and although
we may not be aware of it, we are used to making our own
judgements about research findings. If we take the example
of diet and health, many of us have taken into consideration
information that has been presented on ‘healthy eating’ and
have made our own choices about diet within the context of
our own lives.
lives
 We assess the information with which we are presented, form
an opinion as to the validity and relevance of the research, and
come to our own conclusions based on considerations such as:
  what we understand is the question or problem to be
answered
 how the findings are presented
  why the research was conducted
  what we know about who conducted the research
 what other people think about the information
  how the research findings relate to us
Research as a process
 Research can be seen as a series of linked activities moving

from a beginning to an end. Research usually begins with
the identification of a problem followed by formulation of
research questions or objectives.
 Proceeding from this the researcher determines how best to

answer these questions and so decides what information to
collect, how it will be collected, and how it will be analysed
in order to answer the research question.
 Described in this way the research process is given the

impression of linearity, yet research investigation is often an
iterative process whereby the process of conducting the
research will give rise to new ideas which, in turn, feed back
into the data collection and analysis stage. Decisions made
early in the research process are often revisited in the light
of new insights or practical problems encountered along the
way.
Cyclical or iterative research
process
process
 (1) Through the process of designing your research, consideration of
both practical and conceptual issues may force you to reconsider your
original research question.
 (2) Difficulties with access to research sites or participants may cause
you to reconsider your questions or your methods.
 (3) Issues arising during data collection may suggest that additional
data are required or reveal problems with the original research
question.
 (4) Problems or new questions arising from analysis of data collected
so far may result in a need to collect more data, sample elsewhere or
employ a different technique.
 (5) Finally you progress to the report writing stage.
process
 Regardless of the route taken subsequently, research should
start with the problem and the research questions. If the
intention of research is to answer your questions, it follows
that choice of method should develop from your question:
choose the method that can best provide the information you
need to answer your research question given the resources
available to you. This is one reason why it is very important
to be clear as to what you are asking.
process
 As you can see there are numerous choices to be made
within the research process. Planning your research involves
the consideration of four overlapping themes.
  The conceptual approach – the philosophical
underpinnings of research
  Research design – how data collection is organised
  Data collection techniques – how data are collected
  Sampling – from whom data are collected
process
 These aspects of research planning can be represented as
forming different layers of a research ‘onion’.
process
process
 Choices in research planning must also take into account the
following:
  the types of information outputs required – who needs the
information and for what purposes
  research resources – time, funds, facilities, staff, and
access
  ethical considerations – for example, within the research
plan is it possible to obtain informed consent from all
participants, does the plan involve any risks to the safety of
the researcher, can the researchers assure the confidentiality
of all information given?
process
 You will commonly encounter the following terms, some of
which may seem somewhat interchangeable.
  ‘Research topic’ often used interchangeably or to mean the
same thing as others’ use of the terms ‘research problem’ and
‘research situation’.
  ‘Research strategy’ often used interchangeably or to mean
the same thing as others’ use of the term ‘research approach’.
  ‘Method’ can refer to a broad strategy of data collection or a
specific tool for collecting data. For example ‘survey’ or
‘ethnography’ are sometimes defined as methods, as are
‘interview’ and ‘observation’. For this reason we will limit the
use of the word method.
process
 The various ways in which approaches and designs are
classified by authors are not always distinct, but overlap and
can be used in different combinations. Consistent use of
terminology is also complicated by overlap between some
concepts or components of the research process (for
example, according to our definitions below ‘experiment’ is
both a research design and a research strategy).
 This reflects both the differing use of the same terminology
and the fact that designs may be combined.
process
 Plan
 General plan of how research questions will be answered,
this includes the approach and design.
 Research approach
 This is the theoretical or conceptual basis for the research.
For example: positivist, interpretivist, realist etc.
process
 Research design
 How data collection is organised in order to answer the research
question. Basic design types are: (1) Situation, ‘snap-shot’ or Baseline
(sometimes called case-study); (2) Cross-sectional comparison; (3)
Longitudinal; (4) Longitudinal comparison; (5) Experiment.
 In a longitudinal study, researchers repeatedly examine the same
individuals to detect any changes that might occur over a period of
time.
 Longitudinal studies are a type of correlational research in which
researchers observe and collect data on a number of variables without
trying to influence those variables.
 While they are most commonly used in medicine, economics, and
epidemiology, longitudinal studies can also be found in the other social
or medical sciences.
process
 Research strategy
 Refers to a methodological practice or tradition: For
example: experiment, survey research, or case studies.
 Data collection techniques
 How data are collected: questionnaire; interview;
observation; documentary analysis.
process
 Whatever approach or strategy is followed, the research
process can usually be summarised as follows:
  Identify and formulate the research topic or problem
 Select, narrow and formulate the topic or problem to be
studied and conduct preliminary literature search.
process
  Literature search and review
 Read around the subject to help clarify your research topic,
questions, and methods. Critically review literature to
compare your research with what has already been done,
and to give context. This stage interacts with other stages.
  Research objectives, questions, and hypotheses
 Define clear questions and/or hypotheses.
process
  Research approach, design, and strategy
 Select a research approach and design that will make it
possible to answer research questions and plan the overall
research strategy.
 Identify the data you want to record and from whom/where
you are going to collect it (sampling).
process
  Data collection
  Data analysis
 The data collected are prepared in such a way that they
describe and highlight what was found in the research.
Analytical tools are used to describe the data and measure or
explore relationships between the subjects or items of
interest.
process
  Generalisation and write-up
 The researcher relates the evidence collected to the research
question(s), draws conclusions about the question(s) or
hypotheses, and acknowledges limitations of the research.
What makes research
scientific?
 We place special emphasis on the process of research
because it is the rigour with which this is carried out (the
scientific method) that distinguishes scientific research from
other forms of enquiry, and scientific knowledge from other
kinds of knowledge.
What makes research
scientific?
 Scientific method is one means by which knowledge is
created; however, it is not the only way we know or
understand our world. Three other modes of knowing in
human societies can be identified:
  Authoritarian – individuals serve as sources of

knowledge by virtue of their social or political position.
These individuals may be religious or political leaders,
kings or ‘experts’ such as respected scientists.
What makes research
scientific?
 Mystical – knowledge is sought from the supernatural
world.
 Rationalistic – within the school of rationalism knowledge
can be derived from the rules of logic and without reference to
the empirical world.
What makes research
scientific?
 In contrast to the modes listed above, scientific knowledge
about the world is based upon empirical observation.
Observation is used to develop theory to help us to describe,
understand, and predict how our world works. The
procedures by which observations are gathered, evaluated,
and used to produce new knowledge are termed
methodology.
What makes research
scientific?
 Research methodologies are the rules and procedures by
which knowledge is generated and shared. They allow
research and therefore knowledge claims to be evaluated.
The following criteria are commonly used to evaluate
scientific research:
  Reliability - Is the research study repeatable? – that is:

are the measures used reliable and consistent. If I go back
and repeat the measurements in the same conditions will I
get the same results?
What makes research
scientific?
  Replication - This refers to the idea that the procedures
(methodology) employed in the study are reported in
sufficient detail that a second researcher could repeat the
study.
  Validity – This concerns the integrity of conclusions that
are generated through a research study. There are a number
of issues raised here including (1) does the measure
employed accurately reflect the concept under investigation;
(2) is the causal relationship robust – can we be sure that X
is the cause of Y? (3) Can we be confident that we can
extrapolate our findings beyond the research context?
What makes research
scientific?
 In certain instances and particularly where a research study
is not seeking to extrapolate statistical findings beyond the
research context, the criteria of trustworthiness has been
suggested as a means to ensure the integrity of conclusions.
What makes research
scientific?
 All the criteria above are judged by perceived rigour in
method; that is confidence that the researcher has followed
accepted procedures to ensure to the fullest that the
conclusions reached are robust. To achieve confidence in
the results of a study demands ‘disciplined inquiry’ such
that the data, arguments, and reasoning are able to withstand
examination by other members of the scientific community.
Why research? The role of
information
 Research begins when we want to know something.
Research is concerned with increasing our understanding.
Research provides us with the information and knowledge
needed for problem solving and making decisions.
 Research is sometimes divided into pure (or basic) and
applied research in order to make a distinction between
research that is carried out to further our knowledge and that
which seeks to apply pre-existing knowledge to real world
problems. Our focus here is applied research for decision-
making for public policy. In this context the purpose of
research is ‘problem solving’.
Research for problem solving
 Problem solving can be broken down into a number of

separate components, each of which requires information
and analysis:
  identification of problems
  diagnosis of causes
  identification of potential solutions
  decision for action
  monitoring and evaluation of action and outcomes
 Information for policy-making will therefore serve one or

more of the following functions.
  Description – to provide baseline data or simply a
picture of how things are.
  Explanation (analytical) – to understand why things are
the way they are, what factors explain the way things are.
  Prediction – to predict how systems will change under
alternative scenarios (modelling).
  Prescription and planning (decision-making) –
prescription and planning relating to changes in existing
systems.
  Monitoring and evaluation – monitoring and evaluation

of the effects of changes during and after they have been
made. Investigations may be made to compare results in
practice with predictions, or to monitor the effects of a
policy, management technique or treatment.
Examples of research serving
the functions
the functions
 Current observations, knowledge, and policy concerns
therefore tend to influence both the selection of topics to be
researched and the way that they are researched. How may
this relationship between policy and research limit the
benefits of research for policy-makers? How may this
relationship between policy and research pose problems for
researchers?
the functions
 Research can limit policy change or stimulate it. Thus if
current policy sets the agenda for research, then research
that conforms to the broad objectives and assumptions of
current policy and uses easily available data may not ask
enough questions to challenge the fundamental approach of
policy. It may then promote more effective design and
implementation of an ineffective or inappropriate policy,
without questioning the policy or looking for alternatives.
the functions
 This is an important danger that researchers must be aware
of. We sometimes need to question basic assumptions, and
break out of the ‘box’ of current thinking. Political currents
and pressures, the demands of clients willing to fund
research, ‘fashions’ in development thinking, an emphasis
on technology or modernisation, our own experience, and
the availability of data and information can all, consciously
or unconsciously, strongly affect our choice of problems to
research and the questions we address in our research. This
is a danger that all researchers have consciously to guard
against.
The research–client
relationship
 Research clients usually have a say in the topic of research
and in the definition of research objectives. These objectives
will normally be related to overcoming problems that the
clients define as important.
relationship
 If certain categories of stakeholders (in particular the
socially disadvantaged) are not considered as research
clients, their particular interests and problems (as they
define them) are unlikely to be investigated. This will then
affect the focus of the research, how it is conducted and the
research findings. The cycle of learning and doing, research
and action, may thus exclude particular groups of people
and their interests. This may occur even when individual
research studies are conducted using apparently sound and
objective research methodologies.
relationship
 There are methods that explicitly recognise this relationship
between research, information, and development, these are
participatory research methods. They attempt to involve
different stakeholders (and, particularly, more vulnerable
groups) in the definition of research problems and
objectives and in the interpretation of research findings. The
process of different stakeholders reaching agreement about
information on problems and their solutions makes
participatory research an implicit part of a participatory
development process.
relationship
 Clients (often the research funders) are key drivers in
determining the focus, methods, and findings of research.
Consequently, research may not necessarily focus on issues
that researchers may consider most important, but rather
what the client is willing to fund. Research findings should
be considered in the light of the client–researcher
relationship. Clients and researchers may have a vested
interest in the outcome of the research and results may be
presented to prove a point .
relationship
 For example, a client may commission an experiment to
show that the insecticide they manufacture is more effective
than one manufactured by another company. The results
may show that their insecticide is indeed more effective
than that of their competitor under the specific test
conditions. However, are the results repeatable and do the
insecticides perform in the same manner under different test
conditions?
relationship
 The results should be critically assessed bearing in mind the
focus and presentation.
 This discussion shows that applied research topics and
methods are closely related to the philosophy, objectives,
and approaches of researchers and their clients.
Research and decision-
making
Research and decision-
making
 Picture on previouse slide represents the transformation of
data into knowledge for decision-making. In this context,
data refer to raw, unanalysed material; information is
analysed data; knowledge the subsequent absorption,
assimilation, and understanding of that information. This
diagram draws attention to the steps that link the situation
being studied to acquisition of data; data to information;
information to knowledge; and knowledge to decision-
making.
 Clearly there is scope for error, influence or bias, and delay
at each of these steps.
Information for decision-making,
what can go wrong?
 Bias: data and information may reflect one aspect of a
situation but not others. Bias may arise by accident or
deliberate manipulation of the data or information. Bias may
arise from faulty problem specification, research design,
sampling, data collection, analysis, interpretation, and
presentation – that is at any stage of the research process. It
can be very difficult to detect and prevent as we all have
inadequate understandings which bias our research.
Recognising, and as far as possible reducing, bias is a
critical research skill and a major explicit and implicit focus
of this module.
Information for decision-making, what
can go wrong?
 Recording and editing: recording, transcription, and

editing errors may be introduced at the observational stage,
during recording, collating, and analysis.
 Selection: data and information selected may not be
relevant to the decision for which the information is
required. Information for one decision-maker may be data to
another  depending upon the decisions they are concerned
with.
Information for decision-making, what
can go wrong?
 Time: if each step takes too much time, then the whole
process may not deliver information to the relevant
decision-makers in time to be used in decision-making.
 Analysis and interpretation: particular problems may arise
in analysis and interpretation, such as the use of an
inappropriate analysis technique leading to invalid
interpretation.
What are the qualities of ‘good’ or
useful information?
 The definition of ‘good’ information varies between
different users of information. Therefore it may be helpful
to consider who might be the users of information generated
by research?
  planners have clear information requirements as they
seek to identify and solve problems as outlined above
  implementers of projects need to monitor what they are
achieving and the impacts of their activities, in order to
adjust their activities
useful information?
  policy-makers, like planners, must be informed about

problems, their causes, and means of overcoming them
  donors seek to take actions to support other stakeholders
and at the same time pursue their own, sometimes
contradictory, objectives
  service agencies (such as extension and research
organisations, and input and output marketing companies,
for example) need to make decisions about how to invest
and what activities to engage in
  academics can be major users of information
useful information?
 Information should generally be

  comprehensive (in its coverage of issues important for a particular
decision)
  consistent with previous studies (as regards methods or coverage) and
internally consistent as regards facts and conclusions
  clearly presented and easy to comprehend without excessive detail
  relevant to the problem under consideration
  reliable and accurate
  representative of the situation as a whole, not biased or one-sided in its
coverage
  timely, with regard to the timing of decisions
  generalised and applied only to situations similar to those from which
it was obtained
useful information?
 directed and delivered to the relevant decision-makers

 cost-effective, providing information as cheaply as
possible at costs in proportion to the potential benefits gained
from improved decision-making
There are often conflicts between, for example, cost-
effectiveness and timeliness on the one hand, and
comprehensiveness and reliability on the other. Considerable
skill is required in research design and management; there is
usually a trade-off between these parameters.
How knowledge is used in
decision-making
 The utilisation of research for decision-making is mediated
by social and political factors. Research findings do not
always feed directly into decision-making for policy and
practice. However, research may influence the policy
process and the actions of practitioners even if not used
directly.
How knowledge is used in
decision-making
 Four main types of research utilisation have been
identified :
 (1) Instrumental use – research feeds directly into decision-
making
 (2) Conceptual use – research changes practitioners’
understanding providing new ways of thinking
 (3) Mobilisation of support – research as an instrument of
persuasion
 (4) Wider influence – research findings may come into use
through networks of practitioners and researchers and alter
policy paradigms or belief communities
Where do research ideas
come from?
 Ideas for research problems or topics can arise from a range
of sources such as personal or professional experience, a
theory, the media, or other research studies.
Personal or professional
experience
 Everyday personal or professional experience may lead us
to identify a problem for which we would like a solution.
Alternatively, we may encounter a question or questions
that we would like to try and answer.
 For example, on a personal level, you may prefer the taste
of organically produced vegetables and thus wonder if
people in general prefer the taste of organically produced
vegetables to those produced non-organically.
Personal or professional
experience
 The research topic is a study into taste preferences and the
question ‘do people in general prefer the taste of organically
produced vegetables to those produced non-organically?’
Alternatively, for example, as a professional nature reserve
warden you may want to encourage the establishment and
spread of a particular plant species because you know it is a
food source for a rare butterfly. The research problem may
be, ‘how do I encourage the spread of the plant species of
interest?’
Theory
 Theories are ideas about how things relate to each other.

Theories may be general, commonly held beliefs (such as,
domestic cats are the cause of a decline in bird numbers in
UK gardens) or more technical ideas (for example, that
global warming is causing a change to the timing of the
seasonal responses of the flowering cherry tree in the UK).
Theory
 There are many ways of expressing theories, some are very

formal, others are informal. Here are some examples:
  Keynes’ statement that ... ‘men are disposed as a rule and
on average, to increase their consumption as their income
increases, but not as much as the increase in their income ...’
is a theory.
  The idea that distance learners have different needs than
on-campus students is a theory.
Theory
  A hunch that crossing two particular strains of maize will

produce a more drought-tolerant variety is a theory.
  The assumption that every species has a fundamental
niche, is a theory.
 Theories may be useful in suggesting interesting questions
and generally guiding fieldwork, but should not restrict us
from exploring alternative explanations. The end result of
the research process is knowledge.
Literature and the media
 There are many sources of literature, such as books, journal

articles, and newspapers. When searching and reading
literature it is possible to encounter gaps in information and
knowledge, and problems for which there is currently no
solution. These may provide a good basis for research. We
are also flooded with information presented by the media,
such as television, which again might give rise to research
ideas.
Formulating the research
problem
 Once the general topic or problem has been identified, this
should then be stated as a clear research problem, that is,
taken from just a statement about a problematic situation to
a clearly defined researchable problem that identifies the
issues you are trying to address.
Formulating the research
problem
 It is not always easy to formulate the research problem
simply and clearly. In some areas of scientific research the
investigator might spend years exploring, thinking, and
researching before they are clear about what research
questions they are seeking to answer. Many topics may
prove too wide-ranging to provide a researchable problem.
Choosing to study, for instance a social issue such as child
poverty, does not in itself provide a researchable problem.
The problem is too wide-ranging for one researcher to
address. Time and resources would make this unfeasible and
the results from such a study would consequently lack depth
and focus.
Statement of research
problem
 An adequate statement of the research problem is one of the
most important parts of the research. Different researchers
are likely to generate a variety of researchable problems
from the same situation since there are many research issues
that can arise out of a general problem situation. Your
research will be able to pursue only one in depth.
problem
 For a problem statement to be effective in the planning of
applied research it should have the following :
 (1) The problem reflects felt needs
 (2) The problem is non-hypothetical, ie it must be based on
factual evidence
 (3) It should suggest meaningful and testable hypotheses –
to avoid answers that are of little or no use to the alleviation
of the problem
 (4) The problems should be relevant and manageable
problem
 Formulating the research problem allows you to make clear,
both to yourself and the reader, what the purpose of your
research is. Subsequent elaboration of method should be
oriented to providing information to address that problem.
The problem statement is therefore a very important device
for keeping you on track with your research. It is also one
means by which your research will be evaluated – does the
research address the problem as stated.
How do we get from the research
problem to researchable questions?
 Arriving at specific research objectives, questions or

hypotheses from an idea or problem is a highly personalised
activity – there are different ways of doing it and we all do
it differently. Below is one suggestion based around the idea
of ‘brainstorming’. The results of this process can be
displayed in the form of a ‘spider diagram’ or mental map
of ideas and themes related to your research idea. The
resulting conceptual map can serve both as a starting point
and as a conceptual framework for your investigation.
Conceptual frameworks
 A common tactic here is to ‘unpack’ your idea or problem

thus generating a range of possibilities before narrowing
down on one or two themes. Following the suggestions of
steps could be:
 (1) write down the all the concepts involved, and all the
sub-questions you can think of pertaining to the issue.
Reading around your research idea will help to generate
questions and information and to identify themes and
potential information sources
Conceptual frameworks
 (2) subdivide your questions where possible; split wide

general questions into smaller ones
 (3) begin to order questions and develop focus: group
questions together under common themes, separate general
and specific questions
 (4) start to trim by selecting those questions that you wish to
deal with, consider the resources that will be available to
you
 (5) collate these thoughts within a loose conceptual
framework – this shows how questions and themes are
related and may help guide your thinking at a later stage.
Example
 The example below is taken from a suggested student report

on the theme of contract farming in peri-urban Lusaka. The
student is aware that the government is seeking to promote
contract farming as a development activity and would like
to do some research on the issue. The questions, groupings,
and linkages suggested here are not meant to be exhaustive,
I am sure you will be able to think of more – this is
presented to give an idea of process.
Example
 Step 4. Identify the questions that you think you would like
to deal with. This student has identified certain areas he
would like to focus his investigation on:
  What is the present situation and how do contract
schemes in Lusaka operate?
  Does participation in contract schemes benefit
participating households?
 He also is curious about the idea that lack of trust has

proved a particular problem for schemes in the past and
would like to explore this further.
What is a hypothesis?
 A hypothesis is a unproven answer to a research

problem that is advanced so that it can be tested.
 It is appropriate to use a hypothesis when you are testing a
theory. Your immediate answer to this may be ‘I’m not
testing a theory’; however, remember that our definition of
theory is very broad – ‘an idea about how things relate to
each other’. If you have an expectation of how your
research question will be answered (the outcome) then it is
fair to say you have a theory in mind.
 If you ask of your research question ‘What is the expected
outcome?’ and have an answer, you can ask why? What is
my thinking behind this prediction? This is essentially the
theory that you will be testing.
 If you are not able to predict the answer to your question
then your approach is not one of theory testing and you
should not proceed with developing hypotheses to test.
Your research questions remain as such. This will be the
case if your research is descriptive or exploratory in nature.
Example
 - Below research question is suitable for testing as a

hypothesis- there is a narrow question to be addressed to
which a unproven answer could be suggested:
 When is the best time of year to translocate a meadow
grassland from its original site in Surrey, south-east
England to a new site?
Example
 The question below to be a more exploratory research

question. However, if we have knowledge of some likely
outcomes, these could be stated as hypotheses and tested.
 What are the implications of de-regulating a previously
controlled market for a main food on producers and
consumers in a named region of a country?
Developing a hypothesis from
a research question
 Our definition of a hypothesis stresses that it can be tested. To
meet this criterion the hypothesis must be operationalised – that
is the concepts employed in the hypothesis must be
measurable.
 Developing hypotheses requires that you identify one character,
variable or descriptor of a sampling unit that causes, affects, or
has an influence on, another character, variable or descriptor of
the same or other sampling units. The character, variable or
descriptor that affects other variables or sampling units is called
the independent variable. The character, variable or descriptor
which is affected by the independent variable is called the
dependent variable or response variable.
Good hypotheses
 There are two criteria for good hypotheses. One, hypotheses

are statements about relationships between variables. Two,
hypotheses carry clear implications for testing the stated
relationships. These criteria mean, then, that hypothesis
statements contain two or more variables that are
measurable or potentially measurable and that they specify
how the variables are related.
Diagramming hypotheses
 Diagramming hypotheses is a
useful technique to help
clarify your thinking.
 Usually a hypothesis takes
the form ‘X causes Y’ or ‘X
is related to Y’.
 For example, the first

hypothesis stated in previouse
slide could be represented by
a diagram as follows
 The two variables, or concepts are in boxes that are linked

by an arrow going from one concept to the other. The arrow
indicates that one variable (financial resources) does
something to the other variable (adoption of new
technology).
 The plus sign indicates that the relationship is seen as
positive, that is more of the one will lead to more of the
other. Not all concepts have a positive relationship.
 Once you get used to forming

hypotheses and making
diagrams then you can
explore new patterns
involving more than two
concepts. For example:
 In this case two concepts, finance and distance from market,

are related as independent concepts to the dependent
concept, adoption of technology. One of the independent
concepts is positively related and the other negatively
related to the dependent concept.
 There are endless possibilities. Most research projects deal
only with one small area of the diagram. But it is often
useful to make a diagram of more than you plan to study in
order to show where your research fits into the larger frame
of things and to help you to identify factors which may have
to be taken into account (these could be integrated into your
conceptual map).
Research without hypotheses
 In exploratory research our base knowledge of a subject

may be so low that we cannot formulate meaningful
hypotheses. Nonetheless, exploratory research should be
guided by a clear sense of purpose. Instead of hypotheses,
the design for the exploratory study should state its
purpose, or research objectives as well as criteria by
which the exploration will be judged successful.
 For example, if we are trying to encourage farmers to make use

of compost, we may first need to know the social structure or
social norms of the farming community before we can begin
making meaningful hypotheses about which individuals will
influence the decision and the factors they consider when
making their decision. We can state that our exploratory study
would have the purpose of generating hypotheses about
personal characteristics which correlate with the
adoption/rejection of composting, the composition of the
decision-making unit, and the factors which influence the
decision either to adopt or reject. Success would be measured in
terms of generating testable hypotheses.
 Interpretative research, which seeks to develop knowledge

through understanding meaning, does not usually proceed
with hypotheses.
What are the research
objectives?
 In general, research objectives describe what we expect to
achieve by a project.
 Research objectives are usually expressed in lay terms and
are directed as much to the client as to the researcher.
Research objectives may be linked with a hypothesis or
used as a statement of purpose in a study that does not have
a hypothesis.
 Even if the nature of the research has not been clear to the
layperson from the hypotheses, s/he should be able to
understand the research from the objectives.
objectives?
 A statement of research objectives can serve to guide the
activities of research. Consider the following examples.
  Objective: To describe what factors farmers take into
account in making such decisions as whether to adopt a new
technology or what crops to grow.
  Objective: To develop a budget for reducing pollution by
a particular enterprise.
  Objective: To describe the home of the jambo panda in
China.
objectives?
 In the previouse examples the intent of the research is
largely descriptive.
  In the case of the first example, the research will end the
study by being able to specify factors which emerged in
household decisions.
  In the second, the result will be the specification of a
pollution reduction budget.
  In the third, creating a picture of the home of the jambo
panda in China.
objectives?
 These observations might prompt researchers to formulate
hypotheses which could be tested in another piece of
research.
 So long as the aim of the research is exploratory, for
example to describe what is, rather than to test an
explanation for what is, a research objective will provide an
adequate guide to the research.
Examples of research
statements
 From research problem to hypothesis, a natural science
example
objectives?
 These two examples are set out as primarily natural and social
science, respectively, note that this assumes certain prior
research choices (as in the research onion). Thus in the first
example ‘best’ appears to be judged from a technical
perspective of most successful meadow growth (however that is
defined) without reference to more ‘social’ questions about cost
or the desirability of meadow establishment by particular times
of year. In the second example it is assumed or already known
that the new technologies are effective (and this has both
natural/ technical and social science elements in terms of (a)
physical/biological input/output relations and (b) financial and
non-financial costs and benefits).
Key Points
 To summarise, the key learning points from this unit are as follows.
  Research is systematic investigation to find answers to a problem.
  When we collect data we are collecting raw or analysed facts or
figures. Information is data that has been analysed, that is often
presented in a form specifically for a decision-making task. Knowledge
is assimilated information, an appreciation and understanding of that
information.
  Information that is generated from research is used for problem-
solving and decision-making.
  Research clients have a major role in determining the focus,
methods, and funding of research. Participatory methods are those that
also attempt to involve stakeholders in the process.
Key Points
  Research does not always have an impact on policy-making. There

are many potential reasons for this, including the lack of political will;
inappropriate research questions; or poor dissemination of results to
policy-makers (for example, through information overload or poor
presentation).
  There are three main sources from which we can identify research
topics or research problems: personal experience, theory or literature.
  Research questions ask, how, who, what, why, and when, whilst a
hypothesis is a tentative answer to a research problem that is advanced
so that it can be tested.
  Research objectives describe what we aim to achieve by a project
and may be linked to hypotheses or used for exploratory study without
hypotheses.
Business research
Business research: Definition
 Business research is a process of acquiring detailed

information of all the areas of business and using such
information in maximizing the sales and profit of the
business. Such a study helps companies determine which
product/service is most profitable or in demand. In simple
words, it can be stated as the acquisition of information or
knowledge for professional or commercial purpose to
determine opportunities and goals for a business.
 Business research can be done for anything and everything.

In general, when people speak about business research it
means asking research questions to know where the money
can be spent to increase sales, profits or market share. Such
research is critical to make wise and informed decisions.
 For example: A mobile company wants to launch a new

model in the market. But they are not aware of what are the
dimensions of a mobile that are in most demand. Hence, the
company conducts a business research using various
methods to gather information and the same is then
evaluated and conclusions are drawn, as to what dimensions
are most in-demand, This will enable the researcher to make
wise decisions to position his phone at the right price in the
market and hence acquire a larger market share.
Business research: Types and m
ethodologies
 Business research is a part of the business intelligence
process. It is usually conducted to determine whether a
company can succeed in a new region, to understand their
competitors, or to simply select a marketing approach for a
product. This research can be carried out using qualitative
research methods or quantitative research methods.
Quantitative research
methods
 Quantitative research methods are research methods that
deal with numbers. It is a systematic empirical investigation
using statistical, mathematical or computational techniques.
Such methods usually start with data collection and then
proceed to statistical analysis using various methods. The
following are some of the research methods used to carry
out business research.
1. Survey research
 Survey research is one of the most widely used methods to

gather data especially for conducting business
research. Surveys involve asking various survey
questions to a set of audiences through various types
like online polls, online surveys, questionnaires, etc.
Nowadays, most of the major corporations use this method
to gather data and use it to understand the market and make
appropriate business decisions.
1. Survey research
 Various types of surveys like cross-sectional surveys which

are needed to collect data from a set of audience at a given
point of time or longitudinal surveys which are needed to
collect data from a set of audience across various time
duration in order understand changes in the respondents’
behavior are used to conduct survey research. With the
advancement in technology, now surveys can be sent online
through email or social media.
1. Survey research
 For example: A company wants to know the score for their

website i.e. how satisfied are people who are visiting their
website. An increase in traffic to their website or the
audience spending more time on a website can result in
higher rankings on search engines which will enable the
company to get more leads as well as increase its visibility.
 Hence, the company can ask people who visit their website
with a few questions through an online survey to understand
their opinions or gain feedback and hence make appropriate
changes to the website to increase satisfaction.
2. Correlational research
 Correlational research is conducted to understand the

relationship between two entities and what impact each one
of them has on the other. Using mathematical analysis
methods, correlational research enables the researcher to
correlate two or more variables. Such research can help
understand patterns, relationships, trends, etc. Manipulation
of one variable is possible to get the desired results as well.
Generally, a conclusion cannot be drawn only on the basis
of correlational research.
2. Correlational research
 For example: A research can be conducted to understand the

relationship between colors and gender-based audiences.
Using such research and identifying the target audience, a
company can choose the production of particular color
products to be released in the market. This can enable the
company to understand the supply and demand
requirements of its products.
3. Causal-Comparative
research
 Causal-Comparative research is a method based on the
comparison. It is used to deduce the cause-effect relationship
between variables. Sometimes also known as quasi-
experimental research, it involves establishing an independent
variable and analyzing the effects on the dependent variable. In
such research, manipulation is not done; however, changes are
observed on the variables or groups under the influence of the
same changes. Drawing conclusions through such research is a
little tricky as independent and dependent variables will always
exist in a group, hence all other parameters have to be taken
into consideration before drawing any inferences from the
research.
3. Causal-Comparative
research
 For example: A research can be conducted to analyze the
effect of good educational facilities in rural areas. Such a
study can be done to analyze the changes in the group of
people from the rural areas when they are provided with
good educational facilities and before that.
 Another example can be to analyze the effect of having
dams and how it will affect the farmers or production of
crops in that area.
4. Experimental research
 Experimental research is based on trying to prove a theory.

Such research may be useful in business research as it can
let the product company know some behavioral traits of its
consumers, which can lead to more revenue. In this method,
an experiment is carried out on a set of audiences to observe
and later analyze their behavior when impacted with certain
parameters.
 For example: Experimental research was conducted recently

to understand if particular colors have an effect on its
consumers’ hunger. A set of the audience was then exposed
to those particular colors while they were eating and the
subjects were observed. It was seen that certain colors like
red or yellow increase hunger. Hence, such research was a
boon to the hospitality industry. You can see many food
chains like Mcdonalds, using such colors in their interiors,
brands, as well as packaging.
 Another example of inferences drawn from experimental

research, which is used widely by most bars/pubs across the
world is that loud music makes a person drink more in less
time. This was proven through experimental research and
was a key finding for many business owners across the
globe.
5. Online research /
Literature research
 Literature research is one of the oldest methods available. It
is very economical and a lot of information can be gathered
using such research. Online research or literature research
involves gathering information from existing documents
and studies which can be available at Libraries, annual
reports, etc. Nowadays, with the advancement in
technology, such research has become even more simple
and accessible to everyone.
Literature research
 An individual can directly research online for any
information that is needed, which will give him in-depth
information about the topic or the organization. Such
research is used mostly by marketing and salespeople in the
business sector to understand the market or their customers.
Such research is carried out using existing information that
is available from various sources, although care has to be
taken to validate the sources from where the information is
going to be collected.
Literature research
 For example: A salesperson has heard a particular firm is
looking for some solution which their company provides.
Hence, the salesperson will first search for a decision maker
from the company, investigate what department he is from
and understand what the target company is looking for and
what are they into. Using this research he can cater his
solution to be spot on when he pitches it to this client. He
can also reach out to the customer directly by finding a
mean to communicate with him by researching online.’
Qualitative research methods
 Qualitative research is a method that has a high importance

in business research. Qualitative research involves obtaining
data through open-ended conversational means of
communication. Such research enables the researcher to not
only understand what the audience thinks but also why he
thinks it. In such research, in-depth information can be
gathered from the subjects depending on their responses.
Qualitative research methods
 There are various types of qualitative research methods such

as interviews, focus groups, ethnographic research, content
analysis, case study research that are widely used. Such
methods are of very high importance in business research as
it enables the researcher to understand the consumer. What
motivates the consumer to buy and what does not is what
will lead to higher sales and that is the prime objective for
any business.
 Following are a few methods that are widely used in today’s
world by most businesses.
1. Interviews
 Interviews are somewhat similar to surveys, like sometimes

they may have the same questions used. The difference is
that the respondent can answer these open ended
questions at a length and the direction of the conversation or
the questions being asked can be changed depending on the
response of the subject. Such a method usually gives the
researcher, detailed information about the perspective or
opinions from its subject. Carrying out interviews with
subject matter experts can also give important information
critical to some businesses.
1. Interviews
 For example: An interview was conducted by a telecom

manufacturer, with a group of women to understand why do
they have less number of female customers. After
interviewing them, the researcher understood that there
were less feminine colors in some of the models, hence
females preferred not purchasing them. Such information
can be critical to a business such as a telecom manufacturer
and hence it can be used to increase its market share by
targeting women customers by launching some feminine
colors in the market.
1. Interviews
 Another example would be to interview a subject matter

expert in social media marketing. Such an interview can
enable a researcher to understand why certain types of
social media advertising strategies work for a company and
why some of them don’t.
2. Focus groups
 Focus groups are a set of individuals selected specifically to

understand their opinions and behaviors. It is usually a
small set of a group that is selected keeping in mind, the
parameters for their target market audience to discuss a
particular product or service. Such a method enables a
researcher with a larger sample than the interview or a case
study while taking advantage of conversational
communication.
2. Focus groups
 Nowadays, focus groups can be sent online surveys as well

to collect data and answer why, what and how questions.
Such a method is very crucial to test new concepts or
products before they are launched in the market.
 For example: A research is conducted with a focus group to
understand what dimension of screen size is preferred most
by the current target market. Such a method can enable a
researcher to dig deeper if the target market focused more
on screen size, features or colors of the phone. Using this
data, a company can make wise decisions to its product line
and secure a higher market share.
3. Ethnographic research
 Ethnographic research is one of the most challenging

research but can give extremely precise results. Such
research is used quite rarely, as it is time-consuming and
can be expensive as well. It involves the researcher to adapt
to the natural environment and observe its
target audience to collect data. Such a method is generally
used to understand cultures, challenges or other things that
can occur in that particular setting.
3. Ethnographic research
 For example: The worldly renowned show “Undercover boss”

would be an apt example of how ethnographic research can be
used in businesses. In this show, the senior management of a
large organization works in his own company as a regular
employee to understand what improvements can be done, what
is the culture in the organization and to identify hard-working
employees and to reward them. It can be seen that the
researcher had to spend a good amount of time in the natural
setting of the employees and adapt to their ways and processes.
While observing in this setting, the researcher could find out
the information he needed first hand without loss of any
information or any bias and to improve certain things that
would impact his business.
4. Case study research
 Case study research is one of the most important in business

research. It is also used as marketing collateral by most
businesses to land up more clients. Case study research is
conducted to assess customer satisfaction, document the
challenges that were faced and the solutions that the firm
gave them. Using these inferences are made to point out the
benefits that the customer enjoyed for choosing their
specific firm. Such research is widely used in other fields
like education, social sciences, and similar. Case studies are
provided by businesses to new clients to showcase their
capabilities and hence such research plays a crucial role in
the business sector.
4. Case study research
 For example: A services company has provided a testing

solution to one of its clients. A case study research is
conducted to find out what were the challenges faced during
the project, what was the scope of their work, what
objective was to be achieved and what solutions were given
to tackle the challenges. The study can end with the benefits
that the company provided through their solutions, like
reduced time to test batches, easy implementation or
integration of the system, or even cost reduction. Such a
study showcases the capability of the company and hence it
can be stated as empirical evidence to the new prospect.
5. Website visitor
profiling/research
 Website intercept surveys or website visitor
profiling/research is something new that has come up and is
quite helpful in the business sector. It is an innovative
approach to collect direct feedback from your website
visitors using surveys. In recent times a lot of business
generation happens online and hence it is important to
understand the visitors of your website as they are your
potential customers. Collecting feedback is critical to any
business as without understanding a customer, no business
can be successful. A company has to keep its customers
satisfied and try to make them loyal customers in order to
stay on top.
5. Website visitor
profiling/research
 A website intercept survey is an online survey that allows
you to target visitors to understand their intent and collect
feedback to evaluate the customers’ online experience.
Information like visitor intention, behavior path, satisfaction
of overall website, can be collected using this.
 Depending on what information a company is looking for,
multiple forms of website intercept surveys can be used to
gather responses. Some of the popular ones are Pop-ups or
also called Modal boxes and on-page surveys.
5. Website visitor
profiling/research
 For example: A prospective customer is looking for a
particular product that a company is selling. Once he is
directed to the website, an intercept survey will start noting
his intent, and path. Once the transaction has been made a
pop-up or an on-page survey is provided to the customer to
rate the website. Such research enables the researcher to put
this data to good use and hence understand the customers’
intent, his path and improve any parts of the website
depending on the responses, which in turn would lead to
satisfied customers and hence, higher revenues and market
share
Advantages of Business
research
 Business research helps to identify opportunities and
threats.
 It helps identify problems and using this information, wise
decisions can be made to tackle the issue appropriately.
 It helps to understand customers better and hence can be
useful to communicate better with the customers or
stakeholders.
 Risks and uncertainties can be minimized by conducting
business research in advance.
 Financial outcomes and investments that will be needed can
be planned effectively using business research.
Advantages of Business
research
 Such research can help track competition in the business
sector.
 Business research can enable a company to make wise
decisions as to where to spend and how much.
 Business research can enable a company to stay up-to-date
with the market and its trends and appropriate innovations
can be made to stay ahead in the game.
 Business research helps to measure reputation
Disadvantages of Business
research
 Business research can be a high-cost affair
 Most of the time, business research is based on assumptions
 Business research can be time-consuming
 Business research can sometimes give you inaccurate
information, because of a biased population or a small focus
group.
 Business research results can quickly become obsolete
because of the fast-changing markets
Importance of Business
research
 Business research is one of the most effective ways to
understand customers, the market and competitors. Such
research helps companies to understand the demand and
supply of the market. Using such research will help
businesses reduce costs, and create solutions or products
that are targeted to the demand in the market and the correct
audience.
research
 In-house business research can enable senior management
to build an effective team or train or mentor when needed.
Business research enables the company to track its
competitors and hence can give you the upper hand to stay
ahead of them. Failures can be avoided by conducting such
research as it can give the researcher an idea if the time is
right to launch its product/solution and also if the audience
is right.
research
 It will help understand the brand value and measure
customer satisfaction which is essential to continuously
innovate and meet customer demands. This will help the
company grow its revenue and market share. Business
research also helps recruit ideal candidates for various roles
in the company. By conducting such research a company
can carry out a SWOT analysis, i.e. understand the
strengths, weaknesses, opportunities, and threats. With the
help of this information, wise decisions can be made to
ensure business success.
research
 Business research is the first step that any business owner
needs to set up his business, to survive or to excel in the
market. The main reason why such research is of greatest
importance is that it helps businesses to grow in terms of
revenue, market share and brand value.
What is a statistical analysis
 Having a set of interpretations about the results of an

investigation is a great tool that statistical analysis provides
us with, it also allows us to develop statistical models,
design surveys and research studies, especially when we are
talking about large volumes of data.
 We know data analysis involves an in-depth review of each

part of a whole to understand its structure and interpret its
operation. Statistics, on the other hand, is the science that
uses probabilities as a basis to influence the possible
outcomes of situations that are determined by numerical
data when collecting, interpreting and determining their
validity.
 Therefore, statistical analysis occurs when we collect and

interpret data with the intention of identifying patterns and
trends; this means that while it’s a type of data analysis by
itself, it is approached with an interpretive angle that is
useful for making specific decisions and understanding a
company’s potential customers, their behavior and
experiences.
 “Today, statistics is a tool that cannot be lacking in the

analysis of data from an investigation, because from the
conception of the idea of what is going to be investigated,
through the definition of objectives, hypotheses, variables,
collection of the data, organization, review, classification,
tabulation and production of the results for their analysis it
is important to know how to give an appropriate use to the
different measures and statistical models for the analysis.
 When it is accomplished, the results obtained represent a

true contribution to solving the problems essential to the
field where the activities basic to the different investigations
are carried out. ”
With a statistical analysis we can answer questions such as the

following:
Who are our clients?
How much does a client pay in a visit?
What is the age of our clients?
How can we categorize our types of clients?
What types of experience do our clients enjoy?
 Identifying patterns of behavior or different trends in a set

of data helps companies observe and record the buying
behavior of their customers, both to improve their products
or services and to facilitate an updated and improved
shopping experience, obtaining satisfied customers and a
great brand awareness as a result.
How to perform a functional
statistical analysis
 In order to perform a statistical analysis, we need to collect
and review the data samples available in the results of the
study to be analyzed.
 Although there is no single way to carry out an interpretive
analysis, there are practices that can be replicated in any
study if they are carried out in the appropriate way to the
information provided. These tips will allow us to carry out a
useful statistical analysis.
Uses of statistics in data
analysis
 By making a clearer interpretation of a set of data
from qualitative data to quantitative data, through its
analysis and categorization, we will be able to manipulate
and adjust situations in certain contexts with appropriate
decision-making.
 When we have a broader notion of the trends around our
market, we are ahead, since we can use statistical analysis to
predict future behaviors as the result of the implementation
of appropriate strategies for risk management.
Uses of statistics in data
analysis
 In addition, we will have specific data on consumer
behavior to identify their preferences, the products or
services that are most and least attractive to them and how
to approach them to make a purchase.
 There’s always new trends and behaviors among clients or
even employees that we need to constantly review, for this
reason, to carry out more complex data with specific tools
and functions to design a statistical analysis, we recommend
using a survey software such as QuestionPro.
Using samples to collect data
 When there is no appropriate secondary data, you have to

collect your own primary data. You do this from the relevant
population. Here we are using ‘population’ in the statistical
sense of all people or entities that share some common
characteristic. For instance, when Royal Mail wants to see how
long it takes to deliver first-class letters, the population is all
the letters that are posted first-class; a consumer organisation
testing the quality of Whirlpool dishwashers would define the
population as all the dishwashers made by Whirlpool; a toy
manufacturer getting reactions to a new game might define the
population of potential customers as all girls between the ages
of 6 and 11.
 Obviously, it is important to identify the appropriate population

because a mistake here makes all the subsequent data collection
and analysis pointless.
 But this is not as easy as it seems. The population for a survey
of student opinion is clearly the students – but does this mean
only full-time students, or does it include part-time, day-
release, short-course and distance-learning students? What
about students who are doing a period of work experience,
school students and those studying but not enrolled in courses?
If the population is ‘imported cars’, does this include those
where components are imported but assembly is done in this
country, or those where almost finished cars are imported for
finishing here or those where components are exported for
assembly and the finished car is then brought back?
 Even when we can identify a population in principle, there can

be difficulties translating this into actual sources of data. For
instance, it is easy to describe the population of houses with
broadband connections, but it is much more difficult to identify
them all. Similarly, it would be difficult to identify populations
of people who bought an imported television set within the last
five years, or people who use a particular supermarket or
people who caught a cold last winter. The answer is, generally,
to ask a very large number of people and then ignore everyone
who does not have the features you want – but this approach
raises other problems with bias, efficiency and reliability.
 When you can get a complete list of a population, you have

a sampling frame. Some common sources of sampling
frames include electoral registers, memberships of
organisations (such as the Automobile Association), lists of
employees, customer loyalty cards, account holders,
registered users, website addresses and credit rating
agencies. But even when you have a sampling frame your
troubles are not over because a common problem concerns
the size of the population. The sampling frame of, say,
‘households that use electricity in Germany’ has over 60
million entries.
 Then there are two alternatives:

 a census collecting data from every entry in the sampling
frame – which is the whole population
 a sample collecting data from only a representative sample
of entries in the sampling frame.
Types of sample
 When a population is small and the results are important, it

is worth doing a census and collecting data from every
member of the population. A company might run a census
of every person working in a department to get their views
on a proposed reorganisation, a service might ask all of its
users to comment on its quality and a housing association
might ask all of its tenants about a proposal.
Types of sample
 A census clearly gives the most accurate results – but it is

never completely accurate because there are inevitably
errors, misunderstandings and omissions.
 A census of people living in Copenhagen, for example, will
always find that some people are ill, on holiday, travelling,
or cannot answer or simply refuse to answer. So a census is
difficult, time-consuming, expensive – and still not entirely
reliable.
Types of sample
 Sampling is inherently less accurate – but it is also easier,

cheaper and faster. Of course, the main difficulty is identifying
a sample that gives a fair representation of the whole
population. This difficulty highlights the major weakness of
samples, which is the uncertainty about results that can never
give an entirely accurate picture of the population. When your
breakfast cereal contains 20% fruit, you do not expect every
spoonful to contain exactly this amount. However, we can
choose samples carefully to make them more reliable.
 In particular, the samples should be big enough to give fair
representation of the population, but small enough to be
practical and cost effective.
Random sample
 A random sample is the most common type, and has the

essential feature that every member of the population has
exactly the same chance of being chosen to supply data. If
you randomly choose one member of a football team, it
means that all 11 members form the population and each
has exactly the same chance of being chosen.
Random sample
 But this does not mean that a random sample is disorganised

or haphazard.
 If you want some data about tinned soup, you could go to a
supermarket and buy the first dozen tins of soup you see.
This is haphazard – but it is certainly not random becuase
tins of soup in every other shop have no chance of being
chosen. When a television programme asks people to phone
in to give their views, only certain types of people bother to
respond, and they form nothing like a random sample of
viewers.
Random sample
 With a raffle you can take a genuinely random sample by

putting numbers into a hat and choosing one without
looking. On a bigger scale, national lotteries usually have
some mechanism for selecting numbered balls at random.
Random sample
 In business, a more common approach uses random

numbers, which are a stream of random digits – such as 5 8
645301172...
Random sample
 Suppose that you want to collect data from a random sample

of people visiting an office. Using the string of random
digits above, you could stand by the door and interview the
fifth person to pass, then the eighth person after that, then
the sixth after that, then the fourth after that, and so on. The
result is a completely random sample, which should give a
fair representation of the population. The underlying
assumption is that your sample is big enough to give this
accurate representation.
Random sample
 If a sample does not reflect the population exactly it is said to be

biased.
 If you know that 10% of shops in a town centre opened within the past
2 years, but find that a sample contains 30% of shops opened in this
time, then your sample is biased in favour of newer shops. It can be
difficult to avoid such bias. If you decide to save time by simply
writing down a series of digits that looks random, you will always
introduce bias – perhaps reflecting your preferences for even numbers,
sequences that are easy to type on a keyboard or numbers that are
under stronger fingers on a number pad. Similarly, if you ask
interviewers to select people at random they will give a biased sample
because they approach people they find attractive, and avoid those they
find unattractive, very tall people, people in an obvious hurry, people
in groups and so on.
Random sample
 Even a well-organised random sample can be affected by a

few atypical results. A survey of the amount people spend
on transport is biased if one randomly chosen person is a
film star who just bought a Boeing 747.
 We can avoid problems like this by adding some structure
to the sample. The results are not entirely random, but they
maintain significant random elements – and they aim at
giving results of equivalent accuracy, but with a much
smaller and more convenient sample.
Systematic sample
 One way of organising a non-random sample is to collect

data at regular intervals with a systematic sample. For
example, you might interview every tenth person using a
service, weigh every twentieth unit from a production line
or count the people in every sixth car passing. Clearly this is
not a random sample because every member of the
population does not have the same chance of being chosen –
and if you interview every tenth person using a service, then
members 11, 12, 13 and so on have no chance of being
selected.
Systematic sample
 Unfortunately, a systematic sample can introduce bias.

Checking the contents of every twentieth bottle filled by a
bottling machine is unreliable if every twentieth bottle is
filled by the same head on the machine; collecting data from
every thirtieth person leaving a bus station introduces bias if
buses carry about 30 people, and you are always
interviewing the people who get off last.
Stratified samples
 When there are distinct groups – or strata – in the

population, it is a good idea to make sure that members
from each stratum are fairly represented in a sample. For a
stratified sample you divide the population into strata and
then take a random sample from each, with the number
chosen from each stratum ensuring that the overall sample
contains the right mix. For example, 60% of people working
in a company might be women. To get a stratified sample of
views, you divide the population of employees into two
strata – women and men – and randomly select 60% of your
sample from women and 40% from men.
Stratified samples
 A problem with stratified samples appears with small

groups. In Worked example in previouse slide, if
Westmorefield had very few transport operators, the strata
sample would not have suggested collecting data from any
of them – but their views might still be important. We could
get around this by increasing the sample size, but the sample
becomes very large if we include every possible stratum. An
alternative is simply to collect views from all strata, even if
they are very small. Then small groups are over-represented
and the sample is biased – but it does include contributions
from all parts.
Quota samples
 Quota samples extend the idea of stratified sampling by

adding a more rigid structure. They look at the
characteristics of the population, and then specify the
characteristics needed in the sample to match this exactly.
Suppose you want to see how people will vote in an
election. For a quota sample you choose a sample that
contains exactly the same proportions of people with
different characteristics as the population of people eligible
to vote. If the population consists of 4% of men who are
over 60, retired from manual work and living alone, then the
sample is chosen to also have this proportion.
Quota samples
 Quota sampling often uses interviewers who are given a

number of people with different characteristics to interview.
Each interviewer has to fill their quotas, but they choose the
actual people, so there is still a significant random element.
However, the sample is clearly not random because
interviewers who have filled one quota do not interview any
more people in it, and they have no chance of being chosen.
Multi-stage samples
 Suppose that you want a sample of people who subscribe to

a particular magazine. If you take a random sample, you
will probably find that they are spread over a wide
geographical area, and it is inconvenient and expensive to
travel to interview them. A cheaper option is to use multi-
stage sampling, which makes sure that a sample is confined
to a smaller geographical area.
Multi-stage samples
 The usual approach is to divide the country into a number of

geographical regions, such as television or local radio
regions. Then select some of these regions at random, and
divide them into smaller subdivisions – perhaps
parliamentary constituencies or local government areas.
Then select some of these subdivisions at random and again
divide them into smaller areas – perhaps towns or
parliamentary wards. Continue in this way until you have
small enough areas, and then identify a sample of
individuals from within these areas.
Cluster sampling
 Cluster sampling chooses the members in a sample not

individually, but in clusters. This is useful when a
population naturally falls into distinct groups, and then you
randomly select a number of groups and collect information
from some, or all of, the members. For example, you might
divide the population of a town into households, then select
households at random and collect information from
everyone within the chosen households.
Cluster sampling
 Or you might find that each file of invoices is representative

of the whole population, so it is easier to select a file at
random and then collect data from random entries within
this. This has the benefits of reducing costs and being
convenient to organise – and it is especially useful for
surveying working practices, when individual companies
form the clusters.
Organising data collection
 After identifying an appropriate sample, the next stage is to

collect the data.
 There are two ways of doing this. Firstly, you can use direct
observation to see what is happening, and secondly you can
ask people questions.
Observation
 When the population consists of machines, animals, files,

documents or any other inanimate objects, the only way to
collect data is by direct observation.
 Then an observer watches some activity and records what
happens, typically counting a number of events, taking
some measurements or seeing how something works. But
the observers do not have to be human because automatic
recorders are better for simple tasks like recording Internet
use or monitoring traffic conditions.
Observation
 Observation is usually more reliable than asking for data –

but human observers get tired, make mistakes, get distracted
and misunderstand, while automatic observers break down
and develop faults.
Questionnaires
 When you cannot collect data by observation, you have to ask people
to supply it. This means interviewing the sample and eliciting data by a
series of questions. According to the Gallup organisation,5 such
interviews can reveal:
 whether or not a respondent is aware of an issue (‘Do you know of any
plans to develop . . . ?’)
 general feelings for an issue (‘Do you think this development is
beneficial . . . ?’)
 views about specific points in an issue (‘Do you think this
development will affect . . . ?’)
 reasons for a respondent’s views (‘Are you against this development
because . . . ?’)
 how strongly these views are held (‘On a scale of 1 to 5, how strong
are your feelings about this development . . . ?’).
Questionnaires
 A major problem with asking questions is reliability because

people tend to give the answers they feel they ought to give – or
the answer the interviewer wants – rather than the true answer.
For example, fewer people say that they use their mobile phone
while driving than is found from direct observation – and more
people claim to wash their hands after using a public lavatory,
eat healthy food, do exercise, give more to charity, recycle,
read books and so on. There are also problems with emotional
responses, so asking customers how they liked the food in a
restaurant is likely to get replies based on the whole experience,
including who they were with, how they felt, what the weather
was like, how attractive the servers were and so on.
Questionnaires
 The usual way of asking questions is to present them in a

questionnaire – which is an ordered list of questions. There
are several arrangements for administering a questionnaire
including personal interview, telephone interview, postal
survey, e-mail survey, panel survey and longitudinal survey.
 Sometimes the questionnaire is given to people to complete
themselves (particularly by post or e-mail), and sometimes
it is completed by an interviewer.
Personal interviews
 These have an interviewer directly asking questions to

respondents, and they can be the most reliable way of
getting detailed data. They have a high response rate – only
about 10% of people refuse to answer on principle, but this
depends on circumstances and few people will agree to a
long, complicated or inconvenient interview. Quota
sampling needs some assessment and selection of the people
questioned, so it inevitably uses personal interviews.
Personal interviews
 In principle, collecting data by personal interviews is easy

because it only needs someone to ask questions and record
the answers. The reality is more complicated and depends
on skilled interviewers. For instance, they must be careful
not to direct respondents to a particular answer by their
expression, tone of voice or comments. And they should
help to sort out questions that are unclear and ask follow-up
questions – but they should not explain the questions or
offer any help because this would introduce bias.
Personal interviews
 The main drawback with personal interviews is the high

cost. Each interviewer has to be trained, taken to the right
place, given somewhere to work, fed, given overnight
accommodation and so on. Typically an interviewer spends
40% of their time in travel, 25% in preparation and
administration and only 35% in asking questions.
Telephone interviews
 These can be used for the 95% of people who own a

telephone, and it is cheap and easy, involves no travel and
gets a high response rate. On the other hand, it has the
disadvantage of bias because it uses only people with
telephones, who accept anonymous calls and are willing to
give honest answers over the phone. Other weaknesses are
that observers cannot see the respondents and phone calls
annoy people who object to the intrusion.
Telephone interviews
 The usual procedure for telephone interviews has a

computer selecting a phone number at random from a
directory listing. Then an interviewer asks the questions
presented on a computer screen and types in the answers.
This allows the computer to analyse answers interactively
and choose an appropriate set of questions – and it prevents
errors during the transfer of data.
Postal surveys
 These send questionnaires through the post, and ask people

to complete it and return the answers. They are cheap and
easy to organise, and are suitable for very large samples.
But there are drawbacks, such as the inability to observe
people or clarify points, and the difficulty of getting a
questionnaire to the right people. There are also problems
with bias because only certain types of people bother to
reply, and those who have had bad experiences are more
likely to respond than those who have had good
experiences.
Postal surveys
 The major problem with postal surveys is the low response

rate, which is usually lower than 20% and can approach
zero. This might be raised by making the questionnaire
short and easy to complete, sending it to a named person (or
at least a title), enclosing a covering letter to explain the
purpose of the survey, including a pre-paid return envelope,
promising anonymity of replies, using a follow-up letter or
telephone call if replies are slow and promising a summary
of results.
Postal surveys
 Many surveys try to increase the response rate by offering

some reward – typically a small gift, discount on a future
purchase or entry to a prize draw – but this again introduces
bias because respondents now feel more kindly towards the
questionnaire.
E-mail surveys
 These are an extension of postal surveys, and they can

contact very large numbers of people at almost no cost. But
there are obvious problems with bias because they are
limited to people who regularly use e-mail, publish their
address, accept unsolicited messages and want to reply.
Spam is a huge problem on the Internet and most people use
filters that would not allow random questionnaires through.
An alternative is to open a website and ask people to visit
and give their views. But clearly the replies are rarely
representative of the population.
Panel surveys
 These assemble a representative panel of respondents who

are monitored to see how their opinions change over time.
For example, you could monitor the political views of a
panel during the lifetime of a government or their awareness
of a product during an advertising campaign. Panel surveys
are expensive and difficult to administer, so they use very
small samples.
Panel surveys
 One interesting problem is that members of the panel can

become so involved in the issues that they change their
views and behaviour. For instance, a panel that is looking at
the effects of a healthy eating campaign might become more
interested in health issues and change their own habits.
 Another problem is that some panel members inevitably
have to leave, and the remainder become less representative
of the population.
Longitudinal surveys
 These are an extension of panel surveys that monitor a

group of respondents over a long period. For example,
studies routinely monitor the effects of lifestyles on health
over many decades – and find that exercise reduces heart
disease, alcohol increases liver disease, smoking reduces
life expectancy and so on. Оne tv, called Granada TV has
been monitoring the progress of a group of 14 children –
then adults – since 1964. These studies need a lot of
resources and they are generally limited to studies of
sociological, health and physical changes.
Design of questionnaires
 Designing a good questionnaire is far from easy – and many

surveys fail because they asked the wrong questions, or
asked the right questions in the wrong way. Even subtle
differences in wording and layout can have unexpected
effects. A lot of research into the design of questionnaires
has led to useful guidelines, illustrated in the following list.
Many of these are common sense, but they are often
overlooked.
 A questionnaire should ask a series of related questions and

should follow a logical sequence.
 Make the questionnaire as short as possible. People will not
answer long or poorly presented questionnaires, and
unnecessary questions cost more to collect and analyse.
 Questions should be short, simple,definite, easy to

understand and phrased in everyday terms – if people do not
understand a question they will give any response they think
of, or none at all.
 Even simple changes to phrasing can give very different
results. For example, people are more impressed by a
medical treatment described as having a 60% success rate
than by the same treatment described as having a 40%
failure rate.
 People are not always objective, so asking ‘Do you think

that prison sentences would deter speeding drivers?’ gets a
different response from ‘If you are caught driving too fast,
should you go to prison?’
 Avoid leading questions such as ‘Do you agree with the
common view that NBC news is more objective than Fox
news?’ Such questions encourage conformity rather than
truthful answers.
 Use phrases that are as neutral as possible – rephrasing ‘Do

you like this cake?’ to ‘How do you rate the taste of this
cake on a scale of 1 to 5?’
 Phrase all personal questions carefully – with ‘Have you
retired from paid work?’ being more sensitive than ‘Are you
an old-age pensioner?’
 Do not give warnings – a question that starts ‘We
understand if you do not want to answer this, but ...’ will
discourage everyone from answering
 Avoid unclear questions like ‘Do you usually buy more meat
than vegetables?’
 This raises a series of questions – what does ‘usually’ mean?
What is ‘more’? Do frozen meals count as meat or vegetables?
 Ask positive questions like ‘Did you buy a Sunday newspaper
last week?’ rather than the less definite ‘Has the number of
Sunday newspapers you buy changed?’
 Avoid hypothetical questions such as ‘How much would you
spend on life insurance if you won a million euros on a
lottery?’ Any answer will be speculative and probably not
based on any real thought.
 Avoid asking two or more questions in one, such as ‘Do you

think this development should go ahead because it will
increase employment in the area and improve facilities?’
This will confuse people who think the development should
not go ahead, or those who think it will increase
employment but not improve facilities.
 Open questions such as ‘Have you any other comments?’
collect general views, but they favour the articulate and
quick thinking, and are difficult to analyse.
 Ask questions with precoded answers, with respondents

choosing the most appropriate answer from a set of
alternatives. Be prepared for unexpected effects, such as
sensitivity to the colour and format of the questionnaire, or
different types of interviewer getting different responses.
 Always run a pilot survey before starting the whole survey.
This is the only way to identify problems and improve the
questionnaire design.
Non-responses
 Even the best questionnaire will not get a response from

everyone in the sample. There are several reasons for this,
including:
 People may be real unable to answer the questions, perhaps
because of language difficulties or ill health – or they
simply do not know the answer.
Non-responses
 They are out when the interviewer called – but careful

timing of calls, appointments and repeat visits can reduce
this problem.
 They are away for a period, with holidays and business
commitments making surveys in the summer particularly
difficult.
Examples of precoded
questions
Non-responses
 They have moved and are no longer at the given address –

in which case it is rarely worth following up at the new
address.
 They refuse to answer – about 10% of people refuse to
answer on principle, and nothing can be done about these.
Non-responses
 You might be tempted to ignore non-responses. But then

you are assuming that the non-respondents make no
difference – meaning that actual respondents fairly represent
the sample, and this, in turn, represents the population
fairly. This is not necessarily true and there may be a
systematic reason for the non-responses. Suppose you run a
survey and start with the question ‘Does your company
have any strategic alliances?’
Non-responses
 Companies that answer ‘No’ to this question are unlikely to

be interested enough to complete the rest of the
questionnaire, so replies are biased towards companies that
actually have alliances. In an extreme example a postal
questionnaire has been used to ask about literacy skills (in
the way that people with reading difficulties can pick up
information packs when visiting their local library).
Non-responses
 To avoid these effects you should always follow up non-

respondents.
 Another well-timed visit, telephone call or letter might
encourage nonrespondents to reply, but realistically you will
get limited success. Then the only option is to examine non-
respondents to make sure they do not share some common
characteristic that introduces bias.
Summary of data collection
 1 Define the purpose of the data.

 2 Describe the data you need to achieve this purpose.
 3 Check available secondary data to see how useful it is.
 4 Define the population and sampling frame to give primary data.
 5 Choose the best sampling method and sample size.
 6 Identify an appropriate sample.
 7 Design a questionnaire or other method of data collection.
 8 Run a pilot study and check for problems.
 9 Train interviewers, observers or experimenters.
 10 Do the main data collection.
 11 Do follow-up, such as contacting non-respondents.
 12 Analyse and present the results.
 This seems rather complicated, and you may be tempted to

take short-cuts.
 Remember, despite the fact that every decision in an
organisation depends on available information – and this, in
turn, depends on reliable data collection.
 Unfortunately even careful planning cannot eliminate all
errors, which typically arise from:
 failure to identify an appropriate population

 choosing a sample that does not represent this population
 mistakes in contacting members of the sample
 mistakes in collecting data from the sample
 introducing bias from non-respondents
 mistakes made during data analysis
 drawing invalid conclusions from the analysis.

Prezentacija

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Prezentacija

Uploaded by

Copyright:

Available Formats

Research

 We can define research as an activity of systematic enquiry

 Research can be seen as a series of linked activities moving

 Proceeding from this the researcher determines how best to

 Described in this way the research process is given the

  Authoritarian – individuals serve as sources of

  Reliability - Is the research study repeatable? – that is:

 Problem solving can be broken down into a number of

 Information for policy-making will therefore serve one or

  Monitoring and evaluation – monitoring and evaluation

 Recording and editing: recording, transcription, and

  policy-makers, like planners, must be informed about

 Information should generally be

 directed and delivered to the relevant decision-makers

 Theories are ideas about how things relate to each other.

 There are many ways of expressing theories, some are very

  A hunch that crossing two particular strains of maize will

 There are many sources of literature, such as books, journal

 Arriving at specific research objectives, questions or

 A common tactic here is to ‘unpack’ your idea or problem

 (2) subdivide your questions where possible; split wide

 The example below is taken from a suggested student report

 He also is curious about the idea that lack of trust has

 A hypothesis is a unproven answer to a research

 - Below research question is suitable for testing as a

 The question below to be a more exploratory research

 There are two criteria for good hypotheses. One, hypotheses

 For example, the first

 The two variables, or concepts are in boxes that are linked

 Once you get used to forming

 In this case two concepts, finance and distance from market,

 In exploratory research our base knowledge of a subject

 For example, if we are trying to encourage farmers to make use

 Interpretative research, which seeks to develop knowledge

  Research does not always have an impact on policy-making. There

 Business research is a process of acquiring detailed

 Business research can be done for anything and everything.

 For example: A mobile company wants to launch a new

 Survey research is one of the most widely used methods to

 Various types of surveys like cross-sectional surveys which

 For example: A company wants to know the score for their

 Correlational research is conducted to understand the

 For example: A research can be conducted to understand the

 Experimental research is based on trying to prove a theory.

 For example: Experimental research was conducted recently

 Another example of inferences drawn from experimental

 Qualitative research is a method that has a high importance

 There are various types of qualitative research methods such

 Interviews are somewhat similar to surveys, like sometimes

 For example: An interview was conducted by a telecom

 Another example would be to interview a subject matter

 Focus groups are a set of individuals selected specifically to

 Nowadays, focus groups can be sent online surveys as well

 Ethnographic research is one of the most challenging

 For example: The worldly renowned show “Undercover boss”

 Case study research is one of the most important in business

 For example: A services company has provided a testing

 Having a set of interpretations about the results of an

 We know data analysis involves an in-depth review of each

 Therefore, statistical analysis occurs when we collect and

 “Today, statistics is a tool that cannot be lacking in the

 When it is accomplished, the results obtained represent a

With a statistical analysis we can answer questions such as the