Introduction

The research onion was developed by Saunders et al. (2007). It illustrates the
stages that must be covered when developing a research strategy.When viewed
from the outside, each layer of the onion describes a more detailed stage of the
research process (Saunders et al., 2007). The research onion provides an
effective progression through which a research methodology can be designed. Its
usefulness lies in its adaptability for almost any type of research methodology
and can be used in a variety of contexts (Bryman, 2012). This essay will examine
and describe the different stages of the research onion, and explain the concepts
at each stage.
1.1: Understanding the Research Process
The research onion was developed by Saunderset al. (2007) in order to describe
the stages through which the researcher must pass when formulating an
effective methodology. First, the research philosophy requires definition. This
creates the starting point for the appropriate research approach, which is
adopted in the second step. In the third step, the research strategy is adopted,
and the fourth layer identifies the time horizon. The fifth step represents the
stage at which the data collection methodology is identified. The benefits of the
research onion are thus that it creates a series of stages under which the
different methods of data collection can be understood, and illustrates the steps
by which a methodological study can be described.
Figure 1: The Research Onion

(Source: Institut Numerique, 2012, n.p.).

1.2: Research Philosophy
A research philosophy refers to the set of beliefs concerning the nature of the
reality being investigated (Bryman, 2012). It is the underlying definition of the
nature of knowledge. The assumptions created by a research philosophy provide
the justification for how the research will be undertaken (Flick, 2011). Research
philosophies can differ on the goals of research and on the best way that might
be used to achieve these goals (Goddard & Melville, 2004). These are not
necessarily at odds with each other, but the choice of research philosophy is
defined by the type of knowledge being investigated in the research project
(May, 2011). Therefore, understanding the research philosophy being used can
help explain the assumptions inherent in the research process and how this fits
the methodology being used.

Two main ontological frameworks can inform the research process: positivism
and constructionism (Monette et al. 2005). These frameworks might be described
differently (such as empiricism and interpretivism) but the underlying
assumptions are broadly similar (Bryman, 2012). Positivism assumes that reality
exists independently of the thing being studied. In practice this means that the
meaning of phenomena is consistent between subjects (Newman, 1998).
Conversely, constructionism suggests that the inherent meaning of social
phenomena is created by each observer or group (stlundet al. , 2011). In this
philosophy, one can never presume that what is observed is interpreted in the
same way between participants and the key approach is to examine differences
and nuances in the respondents understanding.
Despite the inherent differences between these two practices, it is not
necessarily the case that they form an inherent belief by the researcher that is
then applied to all research contexts. One philosophy is not inherently better
than the other, although researchers may favour one over the other (Podsakoffet
al., 2012). The philosophy simply provides the justification for the research
methodology. The methodology should be informed by the nature of the
phenomena being observed.
1.3: Research Approaches
Two types of approaches are outlined here: the deductive and the inductive
approach.
1.3.1: Deductive Approach
The deductive approach develops the hypothesis or hypotheses upon a preexisting theory and then formulates the research approach to test it (Silverman,
2013). This approach is best suited to contexts where the research project is
concerned with examining whether the observed phenomena fit with expectation
based upon previous research (Wiles et al., 2011). The deductive approach thus
might be considered particularly suited to the positivist approach, which permits
the formulation of hypotheses and the statistical testing of expected results to an
accepted level of probability (Snieder & Larner, 2009). However, a deductive
approach may also be used with qualitative research techniques, though in such
cases the expectations formed by pre-existing research would be formulated
differently than through hypothesis testing (Saunders et al., 2007). The
deductive approach is characterised as the development from general to
particular: the general theory and knowledge base is first established and the
specific knowledge gained from the research process is then tested against it
(Kothari, 2004).
1.3.2: Inductive Approach
The inductive approach is characterised as a move from the specific to the
general (Bryman & Bell, 2011). In this approach, the observations are the
starting point for the researcher, and patterns are looked for in the data (Beiske,
2007). In this approach, there is no framework that initially informs the data
collection and the research focus can thus be formed after the data has been
collected (Flick, 2011). Although this may be seen as the point at which new
theories are generated, it is also true that as the data is analysed that it may be
found to fit into an existing theory(Bryman & Bell, 2011).

This method is more commonly used in qualitative research, where the absence
of a theory informing the research process may be of benefit by reducing the
potential for researcher bias in the data collection stage (Bryman & Bell, 2011).
Interviews are carried out concerning specific phenomena and then the data may
be examined for patterns between respondents (Flick, 2011). However, this
approach may also be used effectively within positivist methodologies, where the
data is analysed first and significant patterns are used to inform the generation
of results.
1.3.3: The Quantitative Approach
As the name suggests, this approach is concerned with quantitative data (Flick,
2011). It holds a number of accepted statistical standards for the validity of the
approach, such as the number of respondents that are required to establish a
statistically significant result (Goddard & Melville, 2004). Although this research
approach is informed by a positivist philosophy, it can be used to investigate a
wide range of social phenomena, including feelings and subjective viewpoints.
The quantitative approach can be most effectively used for situations where
there are a large number of respondents available, where the data can be
effectively measured using quantitative techniques, and where statistical
methods of analysis can be used (May, 2011).
1.3.4: The Qualitative Approach
The qualitative approach is drawn from the constructivist paradigm (Bryman &
Allen, 2011). This approach requires the researcher to avoid imposing their own
perception of the meaning of social phenomena upon the respondent
(Banister et al., 2011). The aim is to investigate how the respondent interprets
their own reality (Bryman & Allen, 2011). This presents the challenge of creating
a methodology that is framed by the respondent rather than by the researcher.
An effective means by which to do this is through interviews, or texts, where the
response to a question can be open (Feilzer, 2010). Furthermore, the researcher
can develop the questions throughout the process in order to ensure that the
respondent further expands upon the information provided. Qualitative research
is usually used for examining the meaning of social phenomena, rather than
seeking a causative relationship between established variables (Feilzer, 2010).
1.4: Research Strategy
The research strategy is how the researcher intends to carry out the work
(Saunders et al., 2007). The strategy can include a number of different
approaches, such as experimental research, action research, case study
research, interviews, surveys, or a systematic literature review.
Experimental research refers to the strategy of creating a research process that
examines the results of an experiment against the expected results (Saunders et
al., 2007). It can be used in all areas of research, and usually involves the
consideration of a relatively limited number of factors (Saunders et al., 2007).
The relationship between the factors are examined, and judged against the
expectation of the research outcomes.
Action research is characterised as a practical approach to a specific research
problem within a community of practice (Bryman, 2012). It involves examining
practice to establish that it corresponds to the best approach. It tends to involve

reflective practice, which is a systematic process by which the professional

practice and experience of the practitioners can be assessed. This form of
research is common in professions such as teaching or nursing, where the
practitioner can assess ways in which they can improve their professional
approach and understanding (Wiles et al., 2011).
Case study research is the assessment of a single unit in order to establish its
key features and draw generalisations (Bryman, 2012). It can offer an insight into
the specific nature of any example, and can establish the importance of culture
and context in differences between cases (Silverman, 2013). This form of
research is effective in financial research, such as comparing the experiences of
two companies, or comparing the effect of investment in difference contexts.
Grounded theory is a qualitative methodology that draws on an inductive
approach whereby patterns are derived from the data as a precondition for the
study (May, 2011). For example, interview data may be transcribed, coded and
then grouped accordingly to the common factors exhibited between respondents.
This means that the results of the research are derived fundamentally from the
research that has been completed, rather than where the data is examined to
establish whether it fits with pre-existing frameworks (Flick, 2011). Its use is
common in the social sciences (Bryman, 2012).
Surveys tend to be used in quantitative research projects, and involve sampling
a representative proportion of the population (Bryman & Bell, 2011). The surveys
produce quantitative data that can be analysed empirically. Surveys are most
commonly used to examine causative variables between different types of data.
Ethnography involves the close observation of people, examining their cultural
interaction and their meaning (Bryman, 2012). In this research process, the
observer conducts the research from the perspective of the people being
observed, and aims to understand the differences of meaning and importance or
behaviours from their perspective.
An archival research strategy is one where the research is conducted from
existing materials (Flick, 2011). The form of research may involve a systematic
literature review, where patterns of existing research are examined and summed
up in order to establish the sum of knowledge on a particular study, or to
examine the application of existing research to specific problems. Archival
research may also refer to historical research, where a body of source material is
mined in order to establish results.
1.5: Choices
The choices outlined in the research onion include the mono method, the mixed
method, and the multi-method (Saunders et al., 2007). As the names of these
approaches suggest, the mono-method involves using one research approach for
the study. The mixed-methods required the use of two or more methods of
research, and usually refer to the use of both a qualitative and a quantitative
methodology. In the multi-method, a wider selection of methods is used (Bryman,
2012). The main difference between the mixed and the multi-method is that the
mixed-method involves a combined methodology that creates a single dataset
(Flick, 2011). The multi-method approach is where the research is divided into
separate segments, with each producing a specific dataset; each is then

analysed using techniques derived from quantitative or qualitative

methodologies (Feilzer, 2010).
1.6: Time Horizons
The Time Horizon is the time framework within which the project is intended for
completion (Saunders et al., 2007). Two types of time horizons are specified
within the research onion: the cross sectional and the longitudinal (Bryman,
2012). The cross sectional time horizon is one already established, whereby the
data must be collected. This is dubbed the snapshot time collection, where the
data is collected at a certain point (Flick, 2011). This is used when the
investigation is concerned with the study of a particular phenomenon at a
specific time. A longitudinal time horizon for data collection refers to the
collection of data repeatedly over an extended period, and is used where an
important factor for the research is examining change over time (Goddard &
Melville, 2004). This has the benefit of being used to study change and
development. Furthermore, it allows the establishment of some control over the
variables being studied. The time horizon selected is not dependent on a specific
research approach or methodology (Saunders et al., 2007).
1.7: Data Collection and Analysis
Data collection and analysis is dependent on the methodological approach used
(Bryman, 2012). The process used at this stage of the research contributes
significantly to the studys overall reliability and validity (Saunders et al., 2007).
Regardless of the approach used in the project, the type of data collected can be
separated into two types: primary and secondary.
1.7.1: The Primary Data
Primary data is that which is derived from first-hand sources. This can be
historical first-hand sources, or the data derived from the respondents in survey
or interview data (Bryman, 2012). However, it is not necessarily data that has
been produced by the research being undertaken. For example, data derived
from statistical collections such as the census can constitute primary data.
Likewise, data that is derived from other researchers may also be used as
primary data, or it may be represented by a text being analysed (Flick, 2011).
The primary data is therefore best understood as the data that is being analysed
as itself, rather than through the prism of anothers analysis.
1.7.2: Secondary Data
Secondary data is that which is derived from the work or opinions of other
researchers (Newman, 1998). For example, the conclusions of a research article
can constitute secondary data because it is information that has already been
processed by another. Likewise, analyses conducted on statistical surveys can
constitute secondary data (Kothari, 2004). However, there is an extent to which
the data is defined by its use, rather than its inherent nature (Flick, 2011).
Newspapers may prove both a primary and secondary source for data,
depending on whether the reporter was actually present. For a study of social
attitudes in the Eighteenth Century, or for a study of the causes of fear of crime
in present day UK, newspapers may constitute primary data. Therefore, the most
effective distinction of the two types of data is perhaps established by the use to

which it is put in a study, rather than to an inherent characteristic of the data

itself.
1.8: Research Design
The research design is the description of how the research process will be
completed. It is a framework which includes the considerations that led to the
appropriate methodology being adopted, the way in which the respondents were
selected, and how the data will be analysed (Flick, 2011). There are a number of
different characteristic research designs, namely the descriptive, explanatory,
and the exploratory.
The descriptive research design relates to reflecting the experiences of
respondents. It is thus related closely to ethnographic studies, but a quantitative
framework is also an appropriate framework; for example, the demographic
characteristics of a population subgroup can be reported (Bryman, 2012). An
explanatory research design is focused on how to effectively explain the
characteristics of a population or a social phenomenon (Saunders et al., 2007).
This may be seen as effective where using a quantitative framework, where the
influence of one variable on another can be established (Kothari, 2004). The
exploratory study is an exploration of an issue that takes place before enough is
known to conduct a formulaic research project. It is usually used in order to
inform further research in the subject area (Neuman, 2003).
1.9: Samples
A sample is a representative segment of a larger population (Bryman, 2012). In
quantitative research, the sample size and how it is selected can be used to
establish the reliability of the results of the study. In qualitative research, the
sample characteristics are also important, but much smaller samples tend to be
used.
1.9.1 Sample Size
The sample size represents the number of respondents selected from the overall
population that are used in the research (Newman, 1998). In quantitative
research, the size of the sample is essential in determining the reliability of the
results of a study. Sample sizes of much less than 30 will tend to produce results
where individual respondents may skew the results. In such cases, the larger the
sample size the more reliable will be the results (Flick, 2011). In qualitative
research, the size of the sample is less important, and the concept of
representativeness is not as strong a guideline for the validity of the research.
1.9.2: Sampling Techniques
Sampling techniques are the ways in which an appropriate sample size is
selected for the wider study (Bryman, 2012). There are a number of accepted
techniques that can be used. A random sample represents individuals within a
larger population who are chosen at random. However, this can result in random
distribution, which can mean significant skewing resulting from the random
nature of sample selection (Neuman, 2003). For example, a random sample may
result in more males than females being represented in a sample, or an unequal
distribution across ages. A stratified sample may then be used to ensure that the
representatives of the population in the sample reflect the significant

characteristics of the wider population, such as making sure that the

demographic characteristics of age and gender are reflected in the sample
(Newman, 1998). A convenience sample is where the sample is taken from an
existing framework, such as an educational institution, given that the ways in
which respondents may be recruited is relatively straightforward. This may be
appropriate if a study is concerned with students views, and it proved
convenient to sample just one educational institution; it may be considered
unlikely that significant variation in students characteristics will occur between
institutions or that those characteristics will have a significant effect on the
results of a study.
Conclusions
In this study, the different stages of the research onion were described. Given
the research onion comprises different stages of many research projects and can
be effectively adapted to different models, this report has necessarily been
summative and restricted in depth. However, the stages defined by Saunders et
al. (2007) have been expounded upon, and the usefulness of the staged
development of the onion demonstrated. The most effective model of its
effectiveness, however, lies in its use.
References
Banister, P., Bunn, G., Burman, E., & Daniels, J. (2011). Qualitative Methods In
Psychology: A Research Guide. London: McGraw-Hill International.
Beiske, B. (2007). Research Methods: Uses and limitations of questionnaires,
interviews and case studies, Munich: GRIN Verlag.
Bryman, A. (2012). Social research methods (5th ed.). Oxford: Oxford University
Press.
Bryman, A., & Allen, T. (2011). Education Research Methods. Oxford: Oxford
University Press.
Bryman, A., & Bell, E. (2011). Business Research Methods (3rd ed.) Oxford: Oxford
University Press.
Feilzer, M. Y. (2010). Doing mixed methods research pragmatically: Implications
for the rediscovery of pragmatism as a research paradigm. Journal of Mixed
Methods Research, 4(1), pp.6-16.
Flick, U. (2011). Introducing research methodology: A beginner's guide to doing a
research project. London: Sage.
Goddard, W. & Melville, S. (2004). Research Methodology: An Introduction,
(2nd ed.) Oxford: Blackwell Publishing.
Gulati, P. M. (2009). Research Management: Fundamental and Applied Research,
New Delhi: Global India Productions.
Institut Numerique, (2012). Research Methodology, http://www.institutnumerique.org/chapter-3-research-methodology-4ffbd6e5e3391 [retrieved 3rd
October, 2014].

Kothari, C. R. (2004). Research methodology: methods and techniques. New

Delhi: New Age International.
May, T. (2011). Social research: Issues, methods and research. London: McGrawHill International.
Monette, D.R., Sullivan, T. J., & DeJong, C. R. (2005). Applied Social Research: A
Tool for the Human Services, (6th ed.), London: Brooks Publishing.
Neuman, W. L. (2003). Social Research Methods: Qualitative and Quantitative
Approaches, London: Allyn & Bacon.
Newman, I. (1998). Qualitative-quantitative research methodology: Exploring the
interactive continuum. Carbondale: Southern Illinois University Press.
stlund, U., Kidd, L., Wengstrm, Y., & Rowa-Dewar, N. (2011). Combining
qualitative and quantitative research within mixed method research designs: a
methodological review. International Journal of Nursing Studies, 48(3), pp. 369383.
Podsakoff, P. M., MacKenzie, S. B., & Podsakoff, N. P. (2012). Sources of method
bias in social science research and recommendations on how to control it. Annual
Review of Psychology, 63, pp.539-569.
Rowley, J. (2012). Conducting research interviews. Management Research
Review, 35(3), pp.260-271.
Saunders, M., Lewis, P., & Thornhill, A. (2007). Research Methods for Business
Students, (6th ed.) London: Pearson.
Silverman, D. (2013). Doing Qualitative Research: A practical handbook. London:
Sage.
Snieder R. & Larner, K. (2009). The Art of Being a Scientist: A Guide for Graduate
Students and their Mentors, Cambridge: Cambridge University Press.
Wiles, R., Crow, G., & Pain, H. (2011). Innovation in qualitative research methods:
a narrative review. Qualitative Research, 11(5), pp.587-604.