Data Analysis

Luan Nguyen
 Lecture Overview
• Processing data
• Displaying data
Processing Data
 Topics covered
• Methods for processing data in quantitative
research
– Coding
– Analysis
• Methods for processing data in qualitative
researcher
– Thematic/Content analysis in qualitative research
• Computers in research
Steps in data processing
 Discussion
• Read the chapter: How can you clean your
data quickly and efficiently?
• Processing and cleaning the data. (2016). In
Toepoel, V. Doing surveys online (pp. 175-191).
London: SAGE Publications Ltd.
 Coding
• The way a variable has been measured
– Nominal
– Ordinal
– Interval
– Ratio
• Numerical values will be used to describe the
data
– Quantitative (numerical) responses
– Categorical responses
– Descriptive responses
 Steps of coding
1. Developing a code book
2. Pre-testing the code book
3. Coding the data
4. Verifying the coded data
An example of coded data on a code
sheet
 Discussion
• Read the journal article: What does it mean to
‘code’ your data and create a code book?
• Read this chapter: How do you determine
which and how many codes to use for your
data?
– Thematic coding and categorizing. (2007). In
Gibbs, G. R. Qualitative Research kit: Analyzing
qualitative data (pp. 38-55). London, England:
SAGE Publications Ltd
 Discussion
• Watch the video: How do you examine the
responses and code the themes of your data?
• Visit the website: How can you ensure your
analysis of your data’s themes is in line with
your research aims?
 Parametric and non-parametric
statistics
• Parametric statistics
– Characteristics of, or factors in, the population are
known
• Non-parametric statistics
– Characteristics of, or factors in, the population are
unknown
 Descriptive and inferential statistics
• Descriptive statistics
– To summarise features of the sample or simple responses
of the sample (e.g., frequencies or correlations).
– No attempt is made to infer or predict population
parameters.
• Inferential statistics
– to infer or predict population parameters or outcomes
from simple measures, e.g., from sampling and from
statistical techniques.
– Based on probability.
 Discussion
• Visit the website: What are the most common
types of quantitative data analysis software
for social science research, and how do they
compare?
 Six key initial questions
1. What kind (scales) of data are there?
2. Are the data parametric or non-parametric?
3. Are descriptive or inferential statistics required?
4. Do dependent and independent variables need to be
identified?
5. Do the research and data analysis need to take
account of moderating and mediating variables?
6. Are the relationships considered to be linear or non-
linear?
 Hypotheses
• Null hypothesis (H0)
• Alternative hypothesis (H1)
• The null hypothesis is the stronger hypothesis,
requiring rigorous evidence not to support it.
• One should commence with the former and cast
the research in the form of a null hypothesis, and
only turn to the latter in the case of finding the
null hypothesis not supported.
Data processing methods in qualitative
research
• Developing a narrative to describe the
situation, episode, event or instance
• Identify main themes that emerge from the
field notes or transcriptions
• In addition, quantify the frequency of
occurrence of main themes to provide their
prevalence
 Discussion
• Read the encyclopedia entry: What is
qualitative content analysis?
• Visit the website: How do you integrate
qualitative data themes into your report?
 Content analysis
Process of content analysis:
1. Identifying the main themes
2. Assigning codes to the main themes
3. Classifying responses under the main themes
4. Integrating themes and responses into the
text of the report
 Discussion
• Read the blog: When might a mixed methods
approach be appropriate?
• Visit the website: How do you synthesise
qualitative and quantitative data in an
integrated type of analysis?
 Advanced Data Analysis in
Qualitative and Mixed Research
 Interim analysis
• Data analysis begins early in qualitative and
continues throughout project.
• Interim analysis: cyclical process of collecting
and analysing data.
• Interim analysis continues until issue
understood (or run out of resources).
 Memoing
• Can include notes about anything, including
thoughts on emerging concepts, themes, and
patterns found in the data; the need for
further data collection comparisons that need
to be made in the data.
• Should not rely on memory later, write down
things as they come up.
• Include memos as additional data to be
analysed.
 Data entry and storage
• Qualitative researchers usually transcribe
their non-visual data, converting interviews,
observational notes, memos into text
documents.
• Transcribed data later analysed, typically with
qualitative data analysis computer program.
 Coding process
• Read transcribed data, line by line.
• Divide data into meaningful analytical units
(called segmenting).
• Coding: marking segments of data with
symbols, descriptive words, or category
names.
• Continue segmenting and coding until
completed initial coding.
 Coding process
• Intercoder reliability: consistency among
different coders
• Intracoder reliability: consistency within a
single individual.
 Coding process
• Co-occurring and Facesheet
Codes
• Text with overlapping codes
• Source: From M. D. LeCompte
and J. Preissle, Ethnography
and Qualitative Design in
Educational Research, p. 294,
copyright© 1993 Academic
Press. Reprinted by permission
of Elsevier and the authors.
 Data theming
• Theme: a word, or more typically, a set of
words denoting an important idea that occurs
multiple times in your data
• Thematic analysis: identification of themes in
the research findings.
• Go beyond theme identification because there
is more information to be accessed.
 Computer programs for
qualitative data analysis
• Simple or complex searches which use
Boolean operators (words such as and and or
that create logical combinations)
• Most popular: MAXQDA, hyperRESEARCH,
QDA Miner, NVivo.
• Others include Dedoose, Ethnograph, and
ATLAS.
 Data analysis in mixed research
Mixed research data analysis matrix
Analysis typesa
Data typesb One type of analysis: Monoanalysis Both types of analysis: Multianalysis
One type of data: Cell 1 Cell 2
Monodata
Monodata-monoanalysis Monodata-multianalysis
This is not a type of mixed data analysis. (a) For quantitative data: Quantitative analysis (QUAN)
and qualitative analysis of quantitative data (QUALITIZE).

OR
(b) For qualitative data: Qualitative analysis (QUAL) and
quantitative analysis of qualitative data (QUANTITIZE)

Both types of Cell 3 Cell 4

data: Multidata
Multidata-monoanalysis Multidata-multianalysis
This type is not frequently used. This is a combination of “(a)” AND “(b)” from cell 2.
Only quantitative analysis of both quantitative This is the type of analysis typically conducted in mixed
and qualitative data methods research.
OR
Only qualitative analysis of both qualitative and
quantitative data
aAnanalysis type is either quantitative (i.e., statistical) or qualitative.
bAdata type is either quantitative or qualitative.
Notation: “QUAL” stands for qualitative analysis; “QUAN” stands for quantitative analysis.
Displaying Data
 Topics covered
• Communicating and displaying analysed data
• Text
• Presenting data in tables
• Types of graphs
• Statistical measures
 Discussion
• Visit the website: How can you use language
to present data in an engaging way?
• Visit the website: How do you make your
language more decisive and use it to present
your results effectively?
 Presenting data in tables
• Most common method to accurately and
effectively present analysed data
• Frequency tables (univariate, one variable)
• Cross tabulations (bivariate, two variables)
• Multivariate containing information about
more than two variables
• Use of percentages to interpret the data
presented
Structure of a table
 Types of percentages
• Row percentage Insert Fig 16.3
• Column percentage Respondents by attitude
toward uranium mining
• Total percentage
and age (cross tabulation
– hypothetical data)
Types of percentages
 Types of graphs
• Histogram
• Bar chart
• 100% bar chart
• Frequency polygon
• Cumulative frequency polygon
• Stem-and leaf display
• Pie charts
• Line diagram or trend curve
• Area chart
• Scattergram
 Different charts
• Insert Fig 16.4 The
stacked bar chart (top
left hand side)
 Different charts
• Insert Fig 16.5 The
100% bar chart (bottom
left hand side)
 Different charts
• Insert Fig 16.7 the
cumulative frequency
polygon (top right hand
side)
 Different charts
• Insert 16.9 Two- and
three-dimensional pie
charts (bottom right
hand side)
 More charts
• Insert Fig 16.10 The line
diagram or trend curve
(top left hand side)
 More charts
• Insert Fig 16.11 The
area chart (bottom left
hand side)
 More charts
• Insert Fig 16.12 The
scattergram (on right
hand side)
 Discussion
• Visit the website: What software can help you
with simple means of displaying your data
graphically?
• Read this chapter: How do you choose a graph
type that is well-suited to your data?
• Choosing data displays. (2016). In Sue, V., &
Griffin, M. Data visualization & presentation with
microsoft office (pp. 3-12). Thousand Oaks, CA:
SAGE Publications, Inc
 Discussion
• Visit the website: Where can I find out about
basic statistical principles?
• Visit the website: Where can I get help if I
need to know more about specific types of
statistical methods?
 Reflections
• What is the main ‘learning point’ or ‘take-away’ for you?
• Which models do you use in your research? How does it help you to
reflect?
• To what extent do you think you can apply the idea from the lecture
to address your professional issue(s)?
• Challenges: Which parts of the lecture raise a red flag for you?
• Concepts: What key concepts would you pass onto someone who
has not attended the lecture?
• Changes: What changes in attitudes, thinking or action are
suggested by the lecture, either for you or others?
 Q&A
• Do you have any questions?
• See you next week