Research Design & Research

Methods
Prof Veruscha Fester
16 September 2020
Quantitative research
• Quantitative research is the systematic empirical investigation of
observable phenomena via statistical, mathematical, or computational
techniques
• The process of measurement is central to quantitative research because it
provides the fundamental connection between empirical observation and
mathematical expression of quantitative relationships.
• Quantitative data is any data that is in numerical form such as statistics,
percentages, etc
• The researcher analyses the data with the help of statistics and hopes the
numbers will yield an unbiased result that can be generalized to some
larger population
Qualitative research
• Qualitative research relies on unstructured and non-numerical data.
• The data include field notes written by the researcher during the
course of his or her observation, interviews and questionnaires, focus
groups, participant-observation, audio or video recordings carried out
by the researcher in natural settings, documents of various kinds
(publicly available or personal, paper-based or electronic records that
are already available or elicited by the researcher), and even material
artifacts.
Research Design vs Research Methods
• The meanings of “Research design” and “Research
methodology” can be confusing and are sometimes used
interchangeably (Hofstee, 2006). In this course they are
defined as:
• Research design – a general technique which is not focused on any
specific problem, but can be applied to a range of problems
• Research method – the way of going about determining the
answers to the research question, within the chosen research
design technique
Research design
• There are many accepted, established ways to design a research project,
each of which works well for different types of research (for example see
the list in Hofstee (2006):120).
• Remember that all research designs have some inherent assumptions and
specific requirements to work well. This means they all have limitations
and that you must investigate and think carefully about which research
design will work best for your specific research problem and question.
• You then need to explain your research design and how it will apply in your
work, giving due consideration to its strengths and weaknesses. Don’t try
to use a research design that is radically different from those known to
work for your type of problem. Adapt and apply one of the techniques
described below. Note that your methodology will fall within one of these
design techniques:
Possible research designs
• Survey based research
• Correlation based research
• Comparative research
• Experimental research
• Secondary data analysis
• Simulations and statistical modeling
(i) Survey based research
• In survey based research you need to obtain information from a
limited number of people, whom you presume have the information
you want, are willing to give it to you and are representative of the
larger group you wish to investigate. Implementing such surveys
generally ranges from the use of carefully structured questionnaires
to (unstructured) in-depth interviews
• The important considerations are the type of questions asked, how
they are asked, the sample size and how representative it is, time and
cost to conduct the survey and any ethical issues that may be
relevant. Surveys can be used to get opinions or simply factual
information
(ii) Correlation based research
• In this type of research two or more variables are compared statistically to
establish if any relationship exists between them. Bear in mind that
variables not under consideration can influence the results, reliable data
(including sample size, accuracy and how representative it is) can be
difficult to get and there is a danger of over-simplifying the relationship(s).

• This type of research is done to prepare for future experimentation or

when experiments cannot be done for whatever reason. Confidence levels
must be well known, and generalisations must be made carefully. Again,
this type of research generally is not encouraged for this course.
(iii) Comparative research
• Comparative research must be systematic and focussed. Generally two
products or processes for example are investigated in depth and results
compared to establish the reasons for similarity or difference. The broader
the comparison the more difficult this approach becomes. In this type of
research you need to make a case for comparing the items, keep focussed
and remember that variables outside the study can influence it, making
conclusions more difficult.

• Comparative research is better suited to situations where most variables

can be controlled by the researcher e.g. experimental work. It requires a
very specific research question, tight focus and carefully defined and
executed methodology if the results and conclusions are to be accepted.
(iv) Experimental research
• Experiments can be done in a laboratory (more control of environment) or in the
field (less control of environment), to test a hypothesis or theory or to observe
behaviour. Experimental research needs to record and explain results, and implies
that if others follow the same procedures they will get the same results. Care
must be taken in identifying and controlling variables that are outside the scope
of the experiments, as they can influence the results. The accuracy of
measurements must be determined and things such as complexity, cost,
interpretation of results and in some cases, ethics, have to be considered.

• This research technique is generally taken as the classic research design because,
if done properly, allows others to repeat the tests for the same results. However,
it can be difficult and expensive to conduct experiments that simulate real
practical problems, and to meaningfully interpret the results. Both of these are
needed.
(v) Secondary data analysis
• In this type of research design data collected by other researches is
studied and analysed to check some aspect of their work or to answer
some additional question, and requires that your analysis of the data
has some significance. There is much data available on a wide range
of subjects, so this can be a good approach as long as you can obtain
the data you need and determine its reliability and limitations.
• You also need to ensure the data are suitable for your use, that you
employ suitable analysis methods and that the cost of getting the
data (if relevant) is feasible. If these criteria can be met, the fields in
which this approach can be used are almost unlimited.
(vi) Simulations and statistical modelling
• Simulations and statistical models attempt to represent practical situations or
processes mathematically using appropriate key variables. A simulation considers
similar scenarios while a statistical model tries to capture and describe the
situation or process. Both can be very useful and have large practical and
theoretical implications, but remember that simplifications are always necessary,
so these models have limitations.

• This approach can be difficult. This difficulty can be compounded by insufficient

reliable data to build and evaluate models and by the need to include more
variables than originally envisaged. This could require simplifying assumptions
resulting in modelling errors. It is therefore an approach that should be
considered very carefully. However, if you have the skills and the necessary data,
it can be rewarding and interesting. Many real world problems exist, so the areas
in which such an approach can be applied are wide. If done successfully the
significance of the results is large.
Research Methodology
Research methodology
• Research methodology – the way of going about determining the
answers to the research question, within the chosen research design
technique
• Having finalised your research question and design, you must determine how
you are going to conduct your research to get to your answers.
• You must define what data are needed,
• how they will be obtained and analysed, and
• how results will be presented in order to extract the required meaning.
• This must be done in sufficient detail to enable anyone else to conduct a
similar study by reading this part of your proposal.
Components of research method
• Research instruments
• Data to be collected
• Analysis
• Presentation of results
Research instruments
• Anything you use to get your data
• For example laboratory equipment or a questionnaire.
• You must describe the design of the instruments, their specific purpose and
their reliability and validity,
• In order to show that you will get correct, reliable, believable data.
• Here you also need to mention any limitations of the instruments, and
understand the difference between simply collecting data and knowing
exactly what data you will collect.
• Even if your particular research doesn’t need instruments, you still need
data, so you must explain and justify here the source of your data.
Data
• This refer to the measurements you will make
• Or the questions you will ask in interviews
• You must describe these in detail
• Referring to the journal article you’ve studied last week, what data
did they collect?
Analysis of results
• Data on its own won’t help you –
• you have to analyse it and extract some meaning.
• You must explain here how you intend to do this i.e.
• what type of analysis will be applied to the acquired data and
• how the results will be presented.
• This may well require some brief detail on for example, statistical methods to
be used,
• Whatever you intend doing, you must show that the analysis will lead
to the desired information.
• Do not make statements such as “will plot graphs using Excel”.
The criteria to consider are:
• Will the method allow you to completely answer the research question and make
valid conclusions?
• Do you believe in the method? Is it applicable to your problem?
• Are you familiar and comfortable with what needs to be done?
• Will the method result in reliable data? This is the first place to look for a
weakness in the proposal (and finally the dissertation)
• How difficult will it be to acquire the primary data? Collecting reliable data can be
very challenging
• How difficult will it be to analyse the data? Do you have the skills and resources
to correctly analyse the data you collect?

• Do you have access to the resources needed for the method? Are they available
when you need them? Can you afford them?
• Will the method allow you to keep within your timeframe? Consider such
things as how long it will take to collect primary data, get permissions,
develop software, develop equipment etc.? – all things that can cause
delays
• Will the method allow you to easily and thoroughly write up the body of
your dissertation (results, analysis, discussion)?
• Does the proposed method align with your own strengths and
weaknesses?

• Will complying with ethical standards be possible?

Summary
• Keep your method as simple and reliable as possible – the less complicated the better
• Clearly identify your research design and methodology. Final details of methodology may
not be available for the proposal, but give a good, provisional indication to show that
your methodology will be relevant and acceptable
• Make sure your intended research method is compatible with the aim and objectives of
the research (show means by which research objectives will be achieved), and motivate
why you think the method is appropriate
• Describe your procedures and indicate the sources of your data in enough detail to
enable other people to duplicate the work
• Plan your investigation in phases, setting measurable target dates
• Describe infrastructure needed and make sure that it will be affordable and available,
especially if you need sponsorship or to use facilities or expertise at other institutions or
companies. If you do not show adequate evidence of this in the proposal, it will be
rejected