Qualitative Data Analysis

• Qualitative researchers face the challenge of analyzing text and other data forms and representing it effectively in tables, matrices, or narrative form
• Data analysis goes beyond analyzing text and image data, involving organization, preliminary review, coding, theme organization, representation, and interpretation—all interconnected in a spiral

Chapter Focus

• Outlines three general analysis approaches from leading authors
• A visual model is presented, called a data analysis spiral, as a conceptualization tool for qualitative research steps
• Explores five inquiry approaches
• Examines specific data analysis procedures within each, comparing them
• Introduces four software programs used in qualitative analysis: MAXQDA, ATLAS.ti, NVivo, and HyperRESEARCH
• Discuss common features of software in data analysis and coding templates

Core Qualitative Data Analysis Steps

• Preparing and organizing data such as transcripts or photographs
• Reducing data into themes via coding and condensing codes
• Representing data through figures, tables, or discussion

General Analysis Procedures

• Three qualitative researchers with different perspectives illustrate typical general analysis procedures
• Madison presents an interpretive framework from critical ethnography
• Huberman and Miles adopt a systematic approach with a long history
• Wolcott uses a traditional approach to ethnographic and case study analysis

Central Steps of Coding

• Coding compresses data into meaningful segments and assigns names
• Codes combine into broader categories or themes
• Data is displayed and comparisons are made using graphs, tables, and charts
• These are core elements of qualitative data analysis

Authors

• Huberman and Miles provide detailed steps such as writing marginal notes, drafting summaries, and noting category relationships
• Madison emphasizes creating a point of view that signals the interpretive framework

Data Analysis Spiral

• Data analysis is custom-built and revised
• Data collection, analysis, and report writing are interrelated, occurring simultaneously
• Qualitative researchers learn by doing
• The data analysis process conforms to a general contour, best represented in a spiral image
• The researcher engages in moving through analytic circles rather than a fixed linear approach

Data Management

• Researchers organize data into computer files and convert them to appropriate text units
• Data is voluminous, requiring easy location within databases

Reading and Memoing

• Researchers continue by understanding the whole database
• Researchers read transcripts entirely to get a sense of the whole before breaking it down
• Memos are short phrases, ideas, or key concepts written in margins

Describing, Classifying, and Interpreting

• This step involves moving from reading and memoing to describing, classifying, and interpreting the data
• Forming codes or categories is the core of qualitative data analysis
• Researchers build detailed descriptions to develop themes, dimensions, and interpretations
• Authors describe what they see within the context of the person, place, or event

Coding

• Coding involves aggregating text or visual data into small information categories
• Researchers seek evidence for codes from various databases and assign labels
• A short list of tentative codes (25–30) matches text segments,
• Beginning researchers tend to create elaborate lists
• Expand categories while reviewing the database

Code Counting

• Counting data codes and determining frequency can be useful, suggesting preliminary insights
• May indicate frequency of occurrence, but must be used carefully to avoid overemphasizing quantity over quality
• Counting conveys a quantitative orientation contrary to qualitative research

Preexisting

• There's a mixed reaction to using pre-existing or a priori codes
• Should researchers use codes based on theoretical models or literature?
• "Prefigured" codes limit analysis

Code Origin

• Code labels can emerge from participant's exact words (in vivo codes), social or health sciences terms, or researcher-composed names

Codes

• Qualitative researchers should look for code segments describing information
• Codes can represent:
  • expected information
  • surprising information
  • conceptually interesting or unusual information

Classification

• Classification involves identifying 5-7 general themes from the categorized information
• Themes consist of several codes aggregated to form a common idea, forming a "family" of themes

Coding Perspectives

• Coding perspectives may include narrative research (stories), phenomenology (experiences), grounded theory (processes), ethnography (cultural themes), or case studies (detailed description)
• Deconstructive stance can focus on desire and power issues

Deconstruction Data Analysis Strategies

• Dismantling dichotomies exposes false distinctions
• Examining silences notes what isn't said or who isn't included
• Attending to disruptions identifies places where text fails to make sense

Focus

• Focus on alien elements: find limits of conceivable/permissible
• Interpret metaphors: a source of multiple meanings
• Analyze double entendres: reveal unconscious subtext
• Separate bias: "reconstruct" text

Interpretation

• Interpretation involves making sense of data, aka the "lessons learned"
• Abstracting beyond codes/themes unlocks larger meanings
• This begins with code development, theme formation, and organization for data sense-making
• Forms of abstraction exist: through hunches, social science constructs, ideas, or intuition

Representation

• Tentative, inconclusive, and questioning interpretations
• Representing the data involves packaging findings in text, tabular, or figure forms
• A visual image, comparison table, or matrix can be useful
• Hierarchical tree diagrams show abstraction levels

Qualitative Data

• Hypotheses/propositions specify relationship among information categories
• Metaphors analyze data, literary devices borrow from domains
• Entire studies may be shaped from metaphors

Feedback

• Feedback is obtained on initial summaries by taking information back to informants

Narrative research analysis

• Analyze story, looking at chronology of events or turning points
• Literary approach can include analysis in science education

Elements of plot structure

• Characters, setting, problem, actions, resolution
• Three-dimensional space approach: interaction, continuity, location
• Narrative analysis: field texts, elements, rewriting, place and setting

Analysis of Narratives

• Researchers can use a literary perspective for analysis, such as analyzing stories in science education by focusing on plot structure elements (characters, setting, problem, actions, resolution)
• Data analysis can follow timelines or biographical narratives

Analytic Strategies

• Thematic: analyzes what is spoken during data collection
• Structural emphasizes how a story is told via language
• Dialogic/performance: talk is interactive
• A story with images

Phenomenological Analysis

• Specific, structured methods
• Involves describing personal experiences via a modified method

Description analysis process

• First, describe personal experience with the phenomenon, setting aside experiences. This directs focus to participants
• Find statements about individual's experience. List significant statements
• Group significant statements into "meaning units" or themes
• Describe "what" participants experienced and "how."
• Provide textural & structure descriptions

Reflection Guides

• Consider space felt by individuals, physical presence, time dimensions, and relationships with others

Grounded Theory

• Grounded theory uses a 3-phase coding process: open, axial, and selective
• Develop categories (open), connect categories (axial), building a "story" (selective), theoretical propositions (selective)

Initial Categories

• Examine text, like transcripts, for salient categories
• Attempt "saturate" them, Instances must represent the category
• Subcategories define properties

Charmaz approach

• Charmaz emphasizes emergent, multiple realities; the link of facts and values; provisional information; and a narrative about social life as a process. The theory may be a figure
• Can express meanings in life

Ethnographic Research

• Wolcott says data analysis has 3 aspects: description, analysis, & interpretation
• Good starting point for ethnography is to describe the culture sharing group & setting

Description

• Description allows the reader to see through the writer's eyes
• Analysis should involve sorting, highlighting material/findings through tables, and using systematic procedures

Analysis Techniques

• Generate comparison tables: building taxonomies, etc
• search for patterned regularities
• Compare cultural group, evaluate via social standards
• Draw cultural connections: larger frameworks; critique & propose changes

Data

• The researcher speculates interpretations, draws inferences via theories
• Author personalizes interpretation/expresses emotions

Fetterman analysis

• The procedure is not fixed; data is triangulated, looking for thoughts & behavior
• Key events are focused on like ritual
• Analysis includes frequency & magnitude

Analysis

• Analysis makes a detailed description: case setting is always critical
• Stake recommends 4 data/interpretation forms

Data Interpretation Forms

• Seek "categorical aggregation" via data collection. Hoping there will be some issue/relevancy
• Look at the single instance- "direct interpretation"

Categorical Aggregation

• pulling data apart = more meaningful arrangements

Case studies

• Yin suggests cross-case analysis
• Create word table, display data, track similarity & difference
• Develop natural generalizations from data analysis in those cases

Common Analysis Features

• Five approaches follow a basic template, starting research file
• Reading of information begins, then to get data
• Approaches involve a descriptive phase

Important Differences

• Grounded theory and phenomenology use detailed process, ethnographic have common procedures

Computer programs

• Computer programs aid: storage is easy
• Material search is easier
• Think carefully with meaning
• Connect codes/themes
• Retrieve and analyze memos

Using Computer Programs

• Computer programs demand users get proper resources so the qualitative data is easier to understand

Types

• Several types exist such as MAXQDA, ATLAS.ti, NVivo, & HyperResearch