QUALITATIVE DATA ANALYSIS
SESSION OBJECTIVES
• Identify analytic approaches for qualitative
research
• Understand code development
• Understand coding data
• Apply QDAS for managing qualitative data
analysis
BRAINSTORMING
• When Qualitative Data Analysis begins?
• Processes to qualitative data analysis?
ANALYSIS
• Data collection and analysis in qualitative research
should be done simultaneously.
• Qualitative data analysis in qualitative research is
both art and science.
PROCESS OF DATA ANALYSIS
1. Inventory & organize data files
2. Develop verbatim transcripts /translate data
3. Clarify purpose of analysis
4. Select analytic approach
5. Read data, write memos, & develop codes
6. Conduct inter-coder agreement
7. Code entire data set
8. Use analytic tools
• Description
• Comparison
• Categorization
• Explanation
9. Validate findings with data
TRANSCRIPTION
• Verbatim transcription (gold standard)
• Captures every word (questions and responses)
• Reflects true speech (repeated words, laughter,
grammatical errors, etc.)
• Time-consuming
• 1 hour audio = 4-5 hours to transcribe & 30+
pages of text
• Transcription key
• I=interviewer P=participant
THEMATIC ANALYSIS
Key features: Involves identifying themes in textual data
• Identify themes & code data by theme
• Analysis involves description of themes, pattern, and
context
• Often part of other analytic approaches
• Grounded theory
• Narrative analysis
GROUNDED THEORY
• Key features:
• Approach to both data collection and analysis
• Outcome is theory/explanatory models
• Analysis involves: identifying themes, coding,
conceptual categories, and linking these into
an explanatory framework
NARRATIVE ANALYSIS
Key features:
• Case-based analysis (non-cross sectional)
• Analysis of whole case/interview (not segmented)
• Data maintained in context to retain narrative flow
• Analysis examines
• Structure of narrative (sequence, plot, actors, how told)
• Core components of narrative (e.g. events, effect, meaning)
• Compare narratives across data to develop a common/core
narrative.
SELECTING ANALYTIC APPROACH
• What type of data do you have?
• Textual interviews/narratives
• Visual data
• What is the focus of your analysis?
• Substantive content/issues (what is said) vs. Structure of talk
(how it is said)
• What is the purpose of your analysis?
• Build an explanation/theory about behavior (grounded theory)
• Descriptive account/life story (life history analysis)
• Identify issues to design a survey, intervention, etc. (thematic
analysis)
DIVERSITY & OUTLIERS ARE
VALUED
Diversity is valued
• People’s experiences, perceptions & behavior are different
• Analysis focuses on identifying & explaining variation
• Captures variation & relevance of results
Exceptional cases (outliers) are important
• Understand subgroups or atypical behavior
• Identify context of outliers (how & why different?)
• Explaining difference provides nuance to analysis
DEVELOPING CODES
CONCEPTS…
▪ Code: a label of meaning unit- in relation to
the text.
▪ Category: coded meaning units with equal
content are structured in categories- based on
similarities and differences.
▪ Theme: A tread of underlying meaning- Not
necessarily mutually exclusive.
CODING- HEALTHY GREEN PLANTS
Codes category Main Main
category meaning
Apple Sweet Fruit Vitamin
Orange sources
Mango
Lemon Sour
Carrot Root Vegetables Vitamin
Beet root sources
Lettuce Leafy
Cabbage
cauliflower
CODING-EXAMPLE
Codes Subcategory Category
Diagnosis scared Being afraid Fear
Worried about Being worried
deterioration
Terrible to see the Being afraid
deterioration
HOW TO DEVELOP CODES
• Qualitative analysis software will NOT develop codes
for you.
• YOU develop codes
ACTIVE READING OF DATA
Developing codes begins with reading data:
• Active reading involves interacting, questioning, reflecting on
data
• Identify concrete issues and underlying concepts
• Think critically about data as you read
• What is the issue? Why is it relevant?
• What is said directly/indirectly?
• What is present/absent in the passage?
• What is the context? Participant characteristics?
• How does this compare to how the issue is described in
other interviews?
• Write memos as you read (annotate transcripts)
WRITING MEMOS
Memos store your ideas for potential codes
“Memo writing provides a space to become actively
engaged in your materials, to develop your ideas”
TYPES OF CODES
Inductive codes:
• Intense reading and memo writing
• Notice repetition
• Topic changes
• Invivo phrases(name codes based on words and phrases utilized by the participant)
Conceptual codes:
✓ More abstract issues or underlying concepts
✓ Group of issues may represent a concept
Deductive codes:
• Topics on interview guide
• Concepts from research literature
• Personal/professional experience
• Validate with data
Usually have mix of code types in a project
MAKING A CODEBOOK
Codebook
• Ensures consistent application of codes
• Beneficial for working in teams
Define each code
• Code name
• Clear description
• When should I use each specific code?
• What type of text should be coded? What is excluded?
• Give example of text relevant to a code
HOW MANY CODES ARE
ENOUGH?
• Until saturation is reached:
• Depends on data richness
• Depends on purpose of research
• Varies by project (e.g. 30-100+ codes)
• Codes may change during project (merged or split)
CODING DATA
CODING DATA
Coding involves indexing whole data set to provide a
framework for analysis
• Poorly developed codes and inconsistent coding
reduce quality of analysis
Coding involves interpretation
• Interpretation is subjective – check evidence in data
for applying a code
PROCESS OF CODING
➢Close reading of each text segment
➢Identify issues & interpret data
➢Apply relevant code(s) to text segment
➢Systematically code entire data set
➢Modify codes, if needed (add, collapse,
revise)
25
COMPUTER-ASSISTED
QUALITATIVE DATA ANALYSIS
SOFTWARE (C AQDAS)
WHAT C AN C AQDAS DO?
CAQDAS PROGRAMS • CAQDAS Programs Can…
CAN’T….
• Facilitate data management (store all project files)
• Do your analysis automatically
• Provide systematic analysis
• Create codes or code data
• Provide tools to search, retrieve, compare data
• Identify fruitful comparisons
• Facilitate analytic thinking (e.g. exploring
• Interpret data relationships in data, focused searches)
• Generate output of findings • Facilitate team- based research
C AQDAS PROGRAMS
• NVivo
• MAXQDA
• ATLAS-ti
• QDA Miner
• OpenCode
• Dedoose
• And more……
Choosing a CAQDAs Program
Overview of Functions of CAQDAS Programs:
http://www.surrey.ac.uk/sociology/research/researchcentres/caqdas/files/2009Cho
osingaCAQDASPackage.pdf
Choosing an Appropriate CAQDAS Program:
http://www.surrey.ac.uk/sociology/research/researchcentres/caqdas/support/choos
ing/
ASSIGNMENT
• Read about thick description and how it is
different from thin description
• Work on your qualitative data analysis plan
Thank you!
