The University Writing Center (UWrite)

PhD: Research Design

The interrelationship between the building blocks of research1

Epistemology Methodology Methods

Ontology Sources
What and How can we Which precise
What's out Which data
how can we go about tools can we
there to can we
know about acquiring that use to acquire
know? collect?
it? knowledge? it?

Research design can be considered as a plan to collect and analyse data systematically to address research
questions. This includes establishing the nature and types of the data to be collected, the methods to be
used, and the procedures to be followed when analyzing the data.

Both research methodology and research design are informed by philosophical considerations. The decision
to conduct an experiment as opposed to an interview is not simply a technical question, but it also
dependent upon your philosophical approach to the social world (your worldview).

Research paradigms allow researchers to consider their basic set of beliefs, or in other words their
philosophical worldview. This paradigm or worldview will influence the way that somebody approaches
research and their selection of a theoretical framework; additionally, it will help to explain the choice of
certain research approaches: qualitative, quantitative, or mixed methods.

Troudi (2010, p. 1) provides a detailed definition of research paradigm (worldview) based on the work of
Crotty (1998): “A research paradigm or approach is a wide world view that includes major assumptions about
the nature of knowledge, known as epistemology, the nature of reality and social reality, known as ontology,
and assumptions about research methodology.”2

In order to understand the broad concept of research design, it is important to define some key terms:

 Ontology: The study of how one construes existence, a person's conception of the world and what is
entailed in such an existence.3 (What is being?) – The study of the nature of being.
 Epistemology: The theory of knowledge, especially with regard to its methods, validity, and scope,
and the distinction between justified belief and opinion.4 (What is knowing?) – The study of the
nature of knowledge.

1
Source: Adapted from Grix, J. (2002). Introducing students to the generic terminology of social research. Politics, 22(3),
175-186.
2
Troudi, S. (2010). Paradigmatic nature and theoretical framework in educational research. In M. Al-Hamly, M. et al.
(Ed.), English in Learning: Learning in English (pp. 315-323). Dubai: TESOL Arabia Publications.
3
(Grand Canyon University) [Link]
experience/v1.1/[Link]#ontology
4
(Oxford Dictionary) [Link]
  Inductive Research Approach (Bottom up): An inductive approach is where the researcher begins
with as few preconceptions as possible, allowing theory to emerge from the data. This approach
is concerned with the creation of theory.
 Deductive Research Approach (Top down): In deductive research a hypothesis is derived from
existing theory and the empirical world is then explored, and data are collected, in order to test the
hypothesis.5This approach is concerned with testing existing theory.
 Theoretical Framework: A theoretical framework is a frame of reference that is a basis for
observations, definitions of concepts, research designs, interpretations, and generalizations, much as
the frame that rests on a foundation defines the overall design of a house (Haber & LoBiondo-Wood,
1998, p. 141)6. A theoretical framework allows you to examine a topic from a particular (theoretical)
perspective/ lens, which will inform all aspects of your thesis.
 Methodology: The broad philosophical underpinning to your chosen research methods, including
whether you are using qualitative or quantitative methods, or a mixture of both, and why. 7 The
methodology is your broad methodological approach, which is influenced by you theoretical
framework and worldview.
 Methods: Methods are the specific techniques and procedures used to collect and analyse data
(Crotty, 1998, p. 3)8. Are you collecting data using focus groups, questionnaires, participant
observation, and/ or using software packages such as SPSS for analysis?

Research Paradigms (Worldviews)

Three of the most common research paradigms are summarised below in simple terms:

Positivism
Positivism in general refers to philosophical positions that emphasize empirical data and scientific methods.
This tradition holds that the world consists of regularities, that these regularities are detectable, and, thus,
that the researcher can infer knowledge about the real world by observing it. The researcher should be more
concerned with general rules than with explaining the particular. 9

Constructivism/ Interpretivism
Social constructivists believe that individuals seek understanding of the world in which they live and work.
Individuals develop subjective meanings of their experiences – meanings directed towards certain objects or
things. These meanings are varied and multiple, leading the researcher to look for the complexity of views
rather than narrowing meanings into a few categories or ideas. The goal of the research is to rely as much as
possible on participants’ views of the situation being studied (Cresswell, 2014, p. 8) 10.

5
Inductive and deductive. (2009). In O'Reilly, K. Sage key Concepts: Key concepts in ethnography (pp. 104-109). London:
SAGE Publications Ltd. doi: 10.4135/9781446268308
6
Haber, J., & LoBiondo-Wood, G. (1998). Nursing research: Methods and critical appraisal for evidence-based practice.
St. Louis: Mosby
7
(Skills You Need) [Link]
8
Crotty, M. (1998). The foundations of social research: Meaning and perspective in the research process. Thousand
Oaks, CA: Sage.
9
(Popular Social Science) [Link]
10
Cresswell, J. W. (2014). Research design: Qualitative, quantitative, and mixed methods approaches. Thousand Oaks,
CA: Sage.
Pragmatism
Pragmatism as a worldview or philosophy arises out of actions, situations, and consequences rather than
antecedent conditions…. There is a concern with applications – what works – and solutions to problems.
Instead of focusing on methods, researchers emphasise the research problem and use all approaches
available to understand it (Cresswell, 2014, p. 245).

The properties of these paradigms and the assumptions that underlie them are outlined in the table below:11

Research Research Ontology Epistemology Question Methodology/

Paradigm Approac Methods
/ h
Worldview
Positivism Deductive Objective. Reality is What works Quantitative (e.g.
There is a measurable. and what questionnaire/
single reality Focus on reliable does not experiments/
or truth. and valid tools to work? statistical
uncover rules. analysis).

Constructivism/ Inductive Subjective. Reality needs to Why do you/ Qualitative

Interpretivism People/ be interpreted. people act this
(e.g. observation/
groups create Discover the way?
qualitative
their own underlying
interviews).
realities. meaning of
events and
activities.

Pragmatism Deductive/ Objective or Problems/ issues Will this Quantitative and/

Inductive Subjective. need to be intervention/ or Qualitative
Truth/ reality solved using the change
(e.g. interviews
is constantly best/ most improve the
and experiments).
renegotiated/ appropriate situation?
interpreted in method.
relation to its Identifying the
usefulness in issue is the
specific means, change is
situations. the underlying
aim.

Ontology
At one end of the ontological spectrum, we can have a completely mechanistic view of the social world in
which any event that occurs is an ‘effect’ which is produced by a clear ‘cause’. Anything that happens occurs
as a consequence of how the parts (of the machine) function together. If we can measure how they work
sufficiently well, then we can model and predict any given event. At the other end of the spectrum is a
holistic or organismic world-view. Rather than being constant or fixed, everything is in motion, with fluid

11
Adapted from: (Universitat Oberta de Catalunya) [Link]
2013
relationships springing up and falling apart on a contingent basis. Like they say, “you never step into the
same river twice.” As we begin with the idea of continuous change and of processes that interact together in
non-deterministic ways, rather than try to predict single events, we have to take a broadly stochastic or
probabilistic view of the kinds of events that might be possible given the processes involved. In practice,
most research falls somewhere between these two absolute positions on ontology.

Epistemology
Epistemology can be considered along a similar spectrum. The positivist world-view holds that it is possible
to produce new knowledge by making a series of systematic observations of any given phenomenon. For
example, understanding why a financial crash happens might be done by collecting data on stock
movements and investor confidence levels over the years preceding and succeeding the crash. These raw
data are treated as a set of ‘facts’ that can serve as the basis for modelling and theorizing. Developing better
techniques for making observations, such as producing new ways to measure investor confidence, leads to
advances in knowledge. Therefore, positivism is strongly related to a mechanistic ontology. Quantitative
data and statistical analysis are accordingly fundamental to this approach.

At the other end of the spectrum, the constructivist or interpretivist world-view emphasizes that knowledge
is the product of human activity. Constructivism is strongly related to holistic or organismic ontology. It
treats the world as a living, developing process, where social actors actively try to make sense of one
another’s actions. Understanding the world starts with collecting the different ways that actors engage in
sense-making, which are gathered until the analyst feels that he/ she is able to build a coherent picture of
what happened. Constructivist social science offers possible scenarios based on its understanding of past
events and ongoing processes. Qualitative analysis is appropriate because it allows for the collection of data
of sufficient complexity to generate strong analytic accounts.

Although researchers are often more to one side than the other, they tend not to be at either of the
extremes. So, in your own research proposal, it’s best to consider the general approach your research will
take rather than tie yourself entirely to one position.

Research design
Research Design is the overall framework that sets out the terms for your data collection and analysis. ‘Data’
can be either quantitative or qualitative.

They can be information collected by the researcher (‘primary data’) or that already exists in some form
(‘secondary data’). Such material might be in the form of novels, films, exhibitions, artefacts, individual’s
knowledge and opinions.

Research design will obviously influence the type of data collected and the questions asked. Most designs
usually involve a mixture of different data in order to address the specific research questions involved – it is
unusual and inadvisable to rely on only one data type.

Quantitative and qualitative research design

Quantitative data is numerical and provides a dataset (statistics) that can then be analysed. Using the
collected data, you will be able to prove or disprove a hypothesis. In other words, you are using numerical
data to explain a phenomena. In a quantitative design, variables must be defined (i.e. the things to be
measured or estimated) that will be used to organize the data. Variables are typically constructed on the
basis of prior theories taken from the literature. In this way, the majority of quantitative research designs
are
deductive in their approach – they seek to gather data to test hypotheses that formulated from existing
theoretical propositions.

For quantitative research designs, sample size is based on an estimate of the overall population from which
the data are selected. In addition, more advanced inferential statistical tests will require a specific minimum
amount of data in order to deliver results. In general, quantitative research tends to work with relatively
large overall sample sizes.

In contrast to quantitative research, qualitative research does not require numerical data. While quantitative
data can give us the statistics, qualitative data can help us better understand the how and why. In other
words, qualitative methods provide rich and detailed data that help researchers better understand a
phenomenon. In the majority of qualitative designs, the approach taken is inductive, rather than deductive.
It is a more subjective approach and could be affected by motivation, perception or emotion. Issues relating
to bias can therefore be an issue when using qualitative methods. Moreover, as it usually deals with smaller
samples, its results are less generalizable. There is usually no straightforward technical means of calculating
how large the sample ought to be, with adequate size usually being determined against the convention
adopted within a particular research tradition.

When analyzing the data, which is often in the form of words, the researcher tends to focus on identifying
commonalities and differences, or simply themes. Language-based data, for example, can be analysed in a
number of ways, but the most common systems are content analysis, grounded analysis, social network
analysis, discourse analysis, narrative analysis, and conversation analysis. It is your job as the researcher to
select the most appropriate data analysis method, and to justify it accordingly. Refer to research
methodology texts and/ or consult your supervisor in order to make an informed decision.

However you design your research project, you will need to demonstrate clearly how your approach,
methodology and methods specifically help you to answer the research question(s) and objectives that you
have set for your doctoral research.

Research methodology
The key question to ask is whether or not a given tool is actually the right one for the task you have in mind.
For example, although the 5-Point Likert scales is a very convenient method for collecting large quantities of
quantitative data, it does not allow respondents to answer the question on their own terms, or to describe
contradictory feelings. Therefore, it is not suitable (alone) for capturing the complexities of the social world.
Additional interviews might help you better accomplish your objective.

Common methods used in quantitative research include:

 Surveys/Questionnaires:
o Pros – Good at pulling together standardized responses across large samples of respondents
o Cons – Weaker at capturing complexity in responses and context
 Experiments:
o Pros – Good at providing high levels of control over variables
o Cons – Weaker at external and ecological validity
 Analysis of secondary data:
o Pros – Good at providing complex material for multiple forms of analysis
o Cons – Bad at allowing the researcher direct control over what information is collected
 ’Big Data’ analysis:
o Pros – Good at providing ‘real world’ information
o Cons – Weaker at accessing participants own reasoning
Common methods used in qualitative research include:
 Interviews:
o Pros – Good at providing in-depth information
o Cons – Weaker at delivering immediate comparable results
 Focus Groups:
o Pros – Good at gaining quick access to opposing points of view
o Cons – Weaker at controlling for self-presentation issues
 Ethnography/Participant Observation:
o Pros – Good at resourcing thick accounts of social life
o Cons – Weaker at gaining a wide spread of data beyond the immediate population
 Historical/Archival Analysis: Good at capturing context and temporal dimension; Weaker at allowing
a researcher to control data collection
o Pros – Good at capturing context and temporal dimension
o Cons – Weaker at allowing a researcher to control data collection
 Analysis of ‘naturally occurring data’ (i.e. recordings made of social practices in action): Good at
allowing access to what people actually do rather than what they say they do; Weaker at allowing
access to participants’ own experiences
o Pros – Good at allowing access to what people actually do rather than what they say they do
o Cons – Weaker at allowing access to participants’ own experiences

Choosing your research method or methods

Some PhD projects are built exclusively around a single method. Ethnographic research, for example,
involves such considerable effort and immersion that there is no need for additional methods. However,
many PhD students utilise a combination of methods. For example, a quantitative project might use a
mixture of secondary data along with survey material, whilst a qualitative project might include a mixture of
interviews with focus groups.

It is possible to combine methods as ‘Mixed Methods’ research. This is a way of synthesising the best aspects
of each approach e.g. the large samples of quantitative approaches with the depth of qualitative approaches.
For others, it is a means of combining several different kinds of empirical material comparatively within the
same overall project. Designing a mixed methods study in a way that works can be quite difficult to
accomplish and requires a detailed understanding of the philosophical underpinnings of each. Nevertheless,
mixing methods does raise the possibility of extending the spread of possible issues in a research project. It
also enables the research to engage with different research traditions.

Source:
This document is based around the key concepts introduced in the following course:
(University of Leicester) Discovering Your PhD Potential: Writing a Research Proposal. Available on Future
Learn [Link]