This document discusses research design and methods. It outlines different research philosophies like positivism and post-modernism and how they influence knowledge generation. Qualitative and quantitative approaches are compared, with qualitative focused on understanding perspectives and meanings, while quantitative prioritizes measurable and statistical data. Examples of research designs, data collection methods, and analyses are provided for both qualitative and quantitative work. The document stresses using primary data for dissertations and supplementing with secondary data.
4. Data
• Raw numbers &
facts
Information
• Useful data (that
has been analysed/
interpreted)
Knowledge
• Information that is
known by an
individual/group
Wisdom
• “Constructive” use
of knowledge
(Matthews, 1997)
• “Use of knowledge
...to achieve a
common good”
(Sternberg, 2001)
5. Different ways of viewing and
constructing knowledge...
Universal truth generated
by reducing the world to
its constituent parts to
test hypotheses
Knowledge as a social
construction leading
to multiple realities
6. Different types of knowledge...
Knowledge Type
Implicit
(not yet articulated)
Local
Informal
Novice
Tacit
(cannot be articulated)
Traditional
Generalised/Universal
Formal
Expert
Explicit
(articulated)
Scientific
Raymond CM, Fazey I, Reed MS, Stringer LC, Robinson GM, Evely AC
(2010) Integrating local and scientific knowledge for environmental management:
From products to processes. Journal of Environmental Management 91: 1766-1777
Extent to which knowledge is locally
generated/relevant versus universal
Extent to which knowledge generated
via formal, codified processes
Extent to which those generating
knowledge are regarded as experts
Extent to which knowledge is
articulated and accessible to others
Extent to which knowledge is
embedded in and reflects traditional
cultural values/norms, or in the
scientific method
7. Different types of knowledge...
Knowledge Type
Implicit
(not yet articulated)
Local
Informal
Novice
Tacit
(cannot be articulated)
Traditional
Generalised/Universal
Formal
Expert
Explicit
(articulated)
Scientific
Epistemology
Raymond CM, Fazey I, Reed MS, Stringer LC, Robinson GM, Evely AC
(2010) Integrating local and scientific knowledge for environmental management:
From products to processes. Journal of Environmental Management 91: 1766-1777
Extent to which knowledge is locally
generated/relevant versus universal
Extent to which knowledge generated
via formal, codified processes
Extent to which those generating
knowledge are regarded as experts
Extent to which knowledge is
articulated and accessible to others
Extent to which knowledge is
embedded in and reflects traditional
cultural values/norms, or in the
scientific method
Post-modern Positivist
8. Different ways of
managing
knowledge...
Knowledge
Transfer
Producers Users
Producers Users
One-way flow of
existing knowledge
Knowledge
Exchange
Producers Users
Two-way flow of
existing knowledge
Knowledge generation
Producers
Producers generate or
co-generate
knowledge together
Know-ledge
Storage
Knowledge application
Users
Users apply knowledge
gained through transfer
or exchange and provide
feedback to or become
producers of knowledge
Reed MS, Fazey I, Stringer LC, Raymond CM, Akhtar-Schuster M, Begni G, Bigas H, Brehm S,
Briggs J, Bryce R, Buckmaster S, Chanda R, Davies J, Diez E, Essahli W, Evely A, Geeson N,
Hartmann I, Holden J, Hubacek K, Ioris I, Kruger B, Laureano P, Phillipson J, Prell C, Quinn CH,
Reeves AD, Seely M, Thomas R, van der Werff Ten Bosch MJ, Vergunst P, Wagner L (2011)
Knowledge management for land degradation monitoring and assessment: an analysis of
contemporary thinking. Land Degradation & Development
10. How to choose research design
Choice influenced by:
• Research questions you want to answer
• Epistemology
• Preferences towards qualitative/quantitative
11. Designing to questions
The questions you can answer will depend on:
• Existing data availability
• Can you measure/collect relevant new data?
– Skills, equipment, time etc.
• The more focused your question, the easier it
will be to design your research
12. Epistemology
• How do you perceive knowledge, how it is
generated and what constitutes valid
knowledge?
• Positivists: define hypotheses and quantify,
proving beyond doubt
• Post-modernists: more open-ended research
questions and qualitative, providing a range of
perspectives to build credible arguments
13. Qualitative versus quantitative
• Examples of reasons to choose qualitative
versus quantitative in different contexts?
• Benefits/challenges of mixing both?
14. Qualitative or quantitative?
• Depending on research question and
epistemology, qual/quant may be obvious
• Alternatively, start with a qual/quant
preference and select research questions
accordingly
• More on choosing qual/quant later
15. Writing up research design
• Methodology chapter: difference between
research design and methods
• Create a sub-section for both
• Explain your design and methods in enough
detail for someone else to replicate
• Justify your choice – theoretically and/or
empirically
17. Should I use or collect Primary or
secondary Data?
18. Primary data
• Primary data is collected by you, first-hand
19. Secondary data
• Secondary data has been collected by someone else,
and you are using it “second-hand”
20. What should I use?
• For your dissertation it is safest to focus on
primary data collection
– Easier to demonstrate originality
– Harder to fall into trap of writing extended lit
review
• Supplement your primary data with secondary
data to check/deepen your analysis
– Handy if you don’t think you’ve got enough
primary data
22. What is qualitative?
• Understanding the quality or nature of things,
rather than their quantity
– Good for asking “why” questions and gaining an
in-depth understanding of many different
perspectives on an issue (i.e. often subjective)
– Not so suited to statistical analysis and clear-cut,
“objective” answers
– Typically use quite small sample sizes (e.g. 20
interviews and a focus group)
– Can be flexible – adapt your methods as you go
23. Examples of qualitative
– Examples of qualitative data collection methods:
• Open-ended questions in questionnaires
• Semi-structured interviews
• Focus groups
• Participant observation
• In-depth case studies
– Examples of qualitative data:
• Transcripts, audio, interview notes, documents
– Examples of qualitative analysis:
• Content analysis e.g. Grounded Theory Analysis
24. What is quantitative?
• Understanding the quantity of things – being able to
quantify relationships and describe them
mathematically or in terms of their statistical
significance
– Good when you need to be able to answer a research
question with precision, determine if there is a
relationship between two things (x varies with y) or you
need to determine something is statistically significant
– Harder to determine causality (x causes y to vary) and
answer “why” questions
– Typically large data sets (min 50 data points, ideally >100)
– Inflexible – have to stick to and replicate your method
25. Examples or quantitative
• Examples of quantitative data collection
methods:
– Ecological and soil-based survey techniques e.g.
counting plants in quadrats or along transects
– Experiments
– Closed questions in questionnaires e.g. Likert scale
and categorical or numerical questions
• Examples of quantitative analysis
– Calculating percentages, means & standard deviations
– Statistical analyses
26. Qualitative or quantitative?
– I need to ask mainly what, where and when
questions
– I need to understand exactly how something has
changed or might change in future
– I need to understand if something influences
something else
– I need to know of something is significantly greater
or lesser than something else
– The people reading my research want a precise or
“objective” answer to my research questions
– PROBABLY QUANTITATIVE
27. Qualitative or quantitative?
– I need to ask why questions
– I want an in-depth understanding of the issue
– I want to understand what happens in one
particular area in-depth
– I want to interview people
– I want to consider differing perspectives
– I don’t like numbers
– PROBABLY QUALITATIVE
30. Quantitative research design
• Representing reality
– Systematic e.g. transects
– Random and random stratified (i.e. random within
different groups such as socio-economic classes or
habitats)
31. Quantitative data collection
• Counting things…
• Closed ended question surveys with large
samples e.g. via internet
• Ecological and soil-based techniques e.g.
chemical analysis or counting plants in
quadrats
32. Quantitative data analysis
– Descriptive statistics e.g. mean, median, standard
deviation, percentages
– Parametric statistics (sample size >50, not too
much variation)
• Significant differences e.g. T-Test
• Correlations e.g. regression
• Multi-variate e.g. multiple-regression, ordination
– Non-parametric (sample size <50, lots of variation)
• Significant differences e.g. Mann Whitney U
• Correlations e.g. Pearson Product Moment Correlation
34. Qualitative research design
• Purposive sampling
– Selecting respondents on the basis of pre-defined categories
that cover key aspects of your research question
• Snowball sampling
– Keep interviewing within a category till no new ideas
– Get respondents to recommend others for you to interview
• Case studies
– Common in qualitative research
– In-depth understanding of a particular case from which you may
be able to generalise more widely
– Multiple cases representing different perspectives, locations or
components of your issue
35. Qualitative data collection
• Understanding the quality/nature of things…
• Open ended question surveys with large
samples e.g. via internet
• Semi-structured interviews with small samples
(e.g. 12-20 people)
• Participant observation – transcripts and
behaviour
• Make sure you get informed consent from
respondents
36. Qualitative data analysis
– Different types of content analysis and ways of
summarising large bodies of text
• Key word counts (aggregating synonyms)
• Coding for themes – preset or emergent (Grounded Theory
Analysis)
• Discourse analysis to capture context and power relations
• Recursive abstraction – summarising and summarising
summaries and so on, to reach core themes
37. Triangulation
• Simply “checking” your data and interpretation
of results
• Commonly used to increase the reliability of
qualitative studies
• Is there another way of collecting data to
answer the same question a different way?
– Follow your interviews with a focus group
– Follow up historical documents to check an oral
history
38. Summary
• Primary or secondary?
• Qualitative or quantitative?
– Research design
– Data collection methods
– Analysis methods
– Triangulation