Transcript Analyzing Qualitative Data

Analyzing Qualitative Data
Berg, Chapter 11
1
Social Anthropological Approach
• Participated in the community or group.
• More than an observer. An observer will have a
different perspective than a participant.
• Understand welfare recipients by “being one”
• Risk of “going native” means will be an
advocate for the group rather than trying to
understand the group
2
Social Anthropological Approach
• Multiple sources of data including
•
•
•
•
•
•
•
Diaries
Field notes
Interviews
Observations
Photographs
Artifacts
Observations
3
Social Anthropological Approach
• Multiple sources of data including
•
•
•
•
•
•
•
Diaries
Field notes
Interviews
Observations
Photographs
Artifacts
Observations
4
Social Anthropological Approach
• Analyze by describing regularities in
everyday life
• Rituals
• Verbal and nonverbal communication
• Symbols
• Relationships
• How do people come to understand things?
5
Collaborative Research Approach
• Understand to work with participants for
change, problem solving.
• Reflexive. Share understanding, get
feedback, revise ideas. An iterative process
• Participants are “Stakeholders”
6
Overview of Qualitative analysis
• Collected data is transcribed into text/video
• Codes are inductively identified in data and
affixed to transcription.
• Codes are generalized into themes
• Data is sorted by these themes
• Meaningful patterns are identified in each theme
• Patterns compared to previous research, extend
understandings.
7
Quantitative Analysis
• During colonial period newspapers called the U.S.
the “united” States. Gradually became the
“United” States. Historians counted the frequency
the word united was capitalized.
• Number of positive statements a person makes
about his/her partner in a 15 minute discussion
prior to marriage is good predictor of likelihood of
divorce.
8
Quantitative Analysis
• Software can search text for key words and their
synonyms.
• This can show patterns of love, hostility,
support, trust.
• Interviews with clients can show that words
associated with distrust are extremely common
in one program but not another.
• A simple word count, how often each word
appears, may suggest important differences,
say, in how men and women describe marital
strengths or weaknesses.
9
Quantitative Analysis
• Software can search text for key words and their
synonyms.
• This can show patterns of love, hostility, support, trust.
• Interviews with clients can show that words associated
with distrust are extremely common in one program but
not another.
• A simple word count, how often each word appears,
may suggest important differences, say, in how men and
women describe marital strengths or weaknesses.
• Imagine doing this on news coverage or editorials about
a war that becomes unpopular. Terrorist vs. Insurgents
10
Quantitative Analysis
• Software can search text for key words and their
synonyms.
• This can show patterns of love, hostility,
support, trust.
• Interviews with clients can show that words
associated with distrust are extremely common
in one program but not another.
• A simple word count, how often each word
appears, may suggest important differences,
say, in how men and women describe marital
strengths or weaknesses.
11
Quantitative Analysis
• Examples of quantitative analysis
• Imagine doing this on news coverage or
editorials about a war that becomes
unpopular. Terrorist vs. Insurgents
• Violence in TV. How many people has a
5-year old seen being murdered?
12
Qualitative Coding
• Codes and Categories can be
• Deductive from an a priori list—Watch
discussion and count positive and negative
statements about a partner
• Inductive from understanding you get reading a
transcript—Discover that most welfare mothers
express desire for being independent. Call this
“Grounded Research.”
• Both—Usually the case
13
Coding: What Counts?
• Words—easilly quantified
• Themes—qualitative interpretations
• Character reference—How often is partners name used?
How often is a person mentioned?
• Latent Concepts—A variety of words, phrases, nonverbal
behaviors could fall under the concept of expressions of
love
• Semantics. Meanings of affect, strength. Video much better
than transcription. Statement of “I love school” may be
sarcastic, deeply emotional, etc.
14
Coding: What Counts? Classes.
• Common Classes—Common distinctions such as male vs.
female, demographics.
• Special Classes—Special to your group
• Groups may have common distinctions you discover.
Understanding of these is critical.
• Incarcerated women may have these for one another.
• In-group vs. out-group distinctions. Think adolescent
females or males.
• Theoretical Classes—You may identify different classes
based on goals, motivations, shared rituals—Valley girls
15
Coding: How???
• Ask consistent set of questions of data, but be
ready to modify the question.
• Study of cost-effectiveness resulted in repeated
reports of changes in attitudes and behavior.
• Look for unanticipated consequences.
• Animals for inmates  less recividism?
• May find it leads to less guard turnover
because guards consistently report fewer
problems.
16
Coding: How???
• Minutely analyze data.
• Think of funnel like literature review.
• Start trying to code practically
everything.
• Drop/combine codes the second, third,
etc. reading.
• Come back for fresh start after you’ve
generated “final” codes.
17
Coding: How???
• Write theoretical notes
• These are inductive ideas.
• Consistent self serving comments by
clients may lead you to think there is a
problem of trust. A program may be
failing because clients do not trust this
even though this is never explicit.
18
Coding: How???
• Don’t assume importance of common categories
(sex, race). Look for new categories.
• Common categories are confounded with
stereotypes.
• Studying whites who work with African
Americans may lead you to feel position in
hierarch is key category. African American
supervisors may change attitudes of Whites much
more than African American subordinates.
19
20
Forms of Reporting
• Most reports are textual accounts.
• These are greatly strengthened with
verbatim reports of the participants that let
them speak in their own voice.
• Pictures and video clips, for example, as
part of a web report are a great strength.
• Innovative reporting could be a play, such
as the “Vagina Monologue”
21