Chapter 16 - Analysis of Qualitative Data Flashcards
Why is qualitative analysis challenging?
- no universal rules for analyzing
- enormous amount of work required, LOTS of narrative data
- requires creativity, sensitivity, and strong inductive skills - must be able to discern patterns and weave them together
- challenge to reduce data for reporting purposes - maintain richness of data
Category Scheme
Coding data according to categories
careful reading of data, looking for underlying concepts and clusters of concepts
Coding Qualitative Data
Coding data for correspondence to the categories (not easy)
- categories may have been missed in initial categorizing
- materials usually are not linear (paragraphs from an interview may pertain to several different categories)
Conceptual files
Way to organize data by creating a physical file for each category and then cutting out/inserting all materials relating to that category
- labor intensive
- must include contextual information when segment is cut out
Computer-assisted qualitative data analysis software (CAQDAS)
Computer program that permit an entire data set to be entered into the computer, each portion coded, and then portions corresponding to specific code can be retrieved to be analyzed
–>can NOT do the coding, only the researcher can analyze the data
Basic Steps in Qualitative Data Management & Organization (REVIEW)
Transcribing the data
Developing a category scheme
Coding the data
Organizing the data
–>Computerized methods of organization using CAQDAS
–>Manual methods of organization (conceptual files)
Data management vs data analysis
Data Management: reductionist (making it more manageable)
Data Analysis: constructionist (putting segments together into meaning conceptual patterns)
Theme
an abstract entity that brings meaning and identity to a current experience and its variant manifestations - captures and unifies the nature or basis of the experience into a meaningful whole
-not universal (will vary amongst different groups of people)
A General Analytic Overview
- begins with a search for broad categories/themes
- seek out relationships/patterns within the data
- may make flow charts - metaphors - using figurative language to evoke a visual analogy
- validation - does the theme accurately represent the perspectives of the participants
- quasi statistics: frequency with which certain themes/insights are supported by the data - weave thematic pieces together into a cohesive whole
Qualitative Content Analysis
analyzing the content of narrative data to identify prominent themes/patterns among the themes
-breaking down content into smaller units (coding/naming units)
What are the types of units involved in data collection?
- Physical - time, length, size
- Syntactical - grammatical divisions within the data (ex. words, sentences, paragraphs)
- Categorical distinctions define units by identifying something they have in common (membership in a category)
- Thematic distinctions - delicate units according to analyses
Clustering
similarities among units of analysis and hierarchies that conceptualize the text on different levels of abstraction
Ethnographic Analysis
- ->analysis usually begins when data collection starts
- continually looking for patterns/thoughts of participants and comparing one pattern to another (maps, flowcharts, matrices
Spradley’s Research Sequence (1979)
LANGUAGE is the primary means that relates cultural meaning in a culture
Spradley’s 12 Steps
12 Steps:
- locating an informant
- interviewing an informant
- making an ethnographic record
- asking descriptive questions
- analyzing ethnographic interviews
- making a domain analysis
- asking structural questions
- making a taxonomic analysis
- asking contrast questions
- making a componential analysis
- discovering cultual themes
- writing the ethnography
Spradley: Domain Analysis
1st level of analysis
Domains: units of cultural knowledge, broad categories that encompass smaller categories
- identify relational patterns among terms in the domains that are used by members of the culture
- focuses on cultural meaning of terms/symbols and their interrelationships
Spradley: Taxonomic Analysis
2nd level of analysis
–>decide how many domains the data analysis will include
taxonomy: a system of classifying/organizing terms, developed to illustrate the internal organization of a domain and he relationship among the domain subcategories
Spradley: Componential Analysis
3rd level of analysis
–>multiple relationships among terms in the domains are examined (similarities/differences are observed)
Spradley: Theme Analysis
4th (final) level of analysis
–>cultural themes are uncovered, domains are connected in cultural themes
*want to find cultural meaning
Ethnonursing Research Method (Leininger and McFarland - 2006)
Four stage ethnonursing data analysis guide
- collect, describe, record data
- identifying and categorizing descriptors
- data analyszed to discover repetitive patterns in their context
- abstracting major themes and presenting findings
What are the three methods for analyzing phenomenological research?
- Colaizzi (1978)
- Giorgi (1985)
- Van Kaam (1966)
–>all from the Duquesne School of phenomenology
OUTCOME: description of the essential nature of an experience, often through the identification of essential themes
Colaizzi - Phenomenologic Method
- Read all protocols to get a feel for them
- Review each protocol to extract significant statements
- Spell out meaning of each significant statement
- Organize the formulated meaning into clusters of themes:
- refer clusters back to protocol to validate them
- note discrepancies among/between various clusters (do not ignore data) - Integrate results into an exhaustive description of the phenomenon under study
- Formulate an exhaustive description of the phenomenon under study in as unequivocal a statement of identification as possible - leaving no doubt
- Ask participants about the findings thus far as a final validating step
Giorgi - Phenomenologic Method
- Read the entire set of protocols to get a sense of the whole.
- Discriminate units from partisans description of phenomenon being studied
- Articulate the psychological insight in each of the meaning units
- Synthesize all of the transformed meaning units into a consistent statement regarding participants experiences (referred to as the “structure” of the experience”)
- ->can be expressed on a general or specific level
Van Kaam - Phenomenologic Model
- List and group preliminarily the descriptive expressions which must be agreed upon by expert judges. Final listing presents percentages of these categories in that particular sample.
- Reduce the concrete, vague, and overlapping expressions of the participants to more descriptive terms.
- ->intersubjective agreement among judges is needed - Eliminate elements not inherent in the phenomenon being studies or that represent blending of two related phenomenon
- Write a hypothetical identification and description of the phenomenon being studies
- Apply hypothetical description to randomly selected cases from the sample.
- if necessary, revise the hypothesized description, which must then be tested again on a new sample - Consider the hypothesized identification as a valid identification and description once preceding operations have been carried out successfully.
Differences among Colaizzi/Giorgi/Van Kamm
Colaizzi’s method is the only one that calls for validation of results by querying study participants
Giorgi’s view is that it is inappropriate either to return to participants to validate findings or to use external judges to review the analysis
Van Kamm’s method requires the intersubjective agreement be reached with other expert judges.
Van Manen - Phenomenological Model
defintion, six activities
–>combine characteristics of descriptive and interpretive phenomenology
Six Activities:
- turning to the nature of the lived experience
- exploring the experience as we live it
- reflecting on essential themes
- describing the phenomenon through the art of writing and rewriting
- maintaining a strong relation to the phenomenon
- balancing the research context by considering parts and a whole
Van Manen - Phenomenological Model (Three Methods to uncovering descriptions of experiences)
Thematic aspects of experience can be uncovered or isolated from participants descriptions of the experience by three methods:
- Holistic Approach: researchers view the text as a whole and try to capture it’s meaning
- Selective Approach: researchers highlight/pull out statements that seem essential to the experience under study
- Detailed Approach: researchers analyze every sentence
- once themes have been identified, they beam the objects of reflection and interpretation through follow-up interviews with participants
- may use an artistic approach
Heideggerian Hermeneutics (overview) - Phenomenological Method
INTERPRETIVE APPROACH
–>hermeneutic circle: signifies a methodological process by which, to reach understanding, there is continual movement between the parts and the whole of the text being analyzed
-to interpret a text, researchers cannot separate themselves from the meanings of the text and must strive to understand possibilities that the text can reveal
Hermeneutics (7 stage process of data analyzation)
Diekelmann, Allen, and Tanner
–>collaborative efforts by a TEAM of researchers
- All the interviews or texts are read for an overall understanding
- Interpretive summaries of each interview are written
- A team of researchers analyzes selected transcribed interviews/tests
- Any disagreements on interpretation are resolved by going back to the text
- Common meanings/shared practices are identified by comparing/contrasting the text
- Relationships among themes emerge
- A draft of the themes along with exemplars from texts are presented to the team. Responses/suggestions are incorporated into final draft.
Constitutive Pattern
a pattern that expresses the relationship among relational themes and is present in all the interviews/texts
*HIGHEST level of hermeneutic analysis (stage 6)
-a situation is constitutive when it gives content to a person’s self-understanding or a person’s way of being in the world
Benner’s approach to hermeneutics
ANALYTICAL APPROACH
–>three interrelated processes: the search for paradigm cases, thematic analysis, and the analysis of exemplars
Paradigm Cases: strong instances of concerns or ways of being in the world
-used early in analytical process to gain understanding
Thematic Analysis: done to compare and contrast similarities across cases
Exemplars: enhance paradigm cases and thematic analysis by illuminating aspects of case or theme
-readers are able to play a role in the consensual validation of the results by deciding whether the cases support the researchers’ conclusions
Glaser and Straus’ Grounded Theory Method
- constant comparison, fit
- coding
- concerned with the generation of categories and hypotheses rather than testing them
- product: usually a theoretical model that endeavors to explain a pattern of behavior that is relevant for study participants
- cautious with literature
- emergent fit
Constant Comparison
a method that involves comparing elements present in one data source (ex. one interview) with those in another
-continues until content of all sources has been compared so that commonalities are identified
Fit
how closely the emerging concepts fit with the incidents they are representing - depends on how thoroughly the constant comparison was done
Substantive Codes
- ->codes are used to conceptualize data into categories
1. Open Coding (Level I, II, III)
2. Selective Coding
Open coding
used in first stage of constant comparison, captures what is going on in the data
- ex. actual words participants used
- data are broken down and their similarities and differences are examined
- ends when the core category is discovered
Level I Open Coding
In vivo codes
- ->derived directly from the language of the substantive area
- have vivid imagery and “grab”
Level II Open Coding
comparison of new level I codes with previously identified ones
Level III Open Coding
theoretical constructs
–>most abstract, collapsing level II codes aids in identifying constructs
Core Category (variable)
a pattern of behavior tag is relevant and/or problematic for the study participants
1. basic social process: evolves over time in two or more phases (all BSPs are core categories, but not all core categories have to be BSPs)
Selective Coding
researcher code only those data that are related to the core category
Criteria for deciding on a core category…
Central - related to many categories
Recur frequently
Relates meaningfully and easily to other categories
Clear and grabbing implications for formal theory
Theoretical Codes
provide insights into how substantive codes relate to each other
-help grounded theorists to weave the broken pieces of data back together again
18 families (process, strategy, cutting pint, six Cs - causes, contexts, contingencies, consequences, covariances, and conditions)
Strauss and Corbin’s Approach - Grounded Theory
differs from Glaser and Strauss method in regard to method, processes, and outcomes
- stressed that to generate theory, the basic problem MUST emerge from the data (be discovered)
- research problems can come from the research itself, as well as literature or a researcher’s personal/professional experience
OUTCOME: full conceptual descriptions (contrasts original grounded theory method, which generates a theory that explains how a basic social problem that emerged from the at a is processed in a social setting)
Strauss/Corbin: Open Coding
data broken into parts and compared for similarities and differences
-researcher focuses on generating categories and their properties/dimensions
Strauss/Corbin: Axial Coding
analyst systematically develops categories and links them with subcategories
- occurs around an “axis” of a category
- paradigm helps identify linkages (conditions, actions/interactions, consequences)
Strauss/Corbin: Selective Coding
process in which the findings are integrated and refined
Strauss/Corbin: Central Category
first step in integrating the findings = deciding on a central category
–>main category of research
Constructivist Grounded Theory Approach
similar to Glaser
- ->coding generate the bones for your analysis, theoretical bones help theory to function
- coding: word by word, line by line, incident by incident
- construct grounded theory by means of their past/current involvements and interactions with individuals and research practices
initial vs. focused coding
initial: pieces of data are studied so the researcher can learn what the participants view as problematic
focused: analysis is directed toward identifying the most significant initial codes, which are then theoretically coded
