"Data is not the plural of anecdote!" If you've heard this, you're probably a qualitative researcher, and you've been wondering how to inject more rigor into your methodology. This workshop, presented at Syracuse University and at ATTW 2016, discusses the principles of modeling qualitative data. It covers three main types of models and variations, discussing what they're for and how they can be used to more rigorously compare, understand, and interpret your data.
2. What we’ll do
Discuss principles of systematically
analyzing qualitative data
Discuss types of models for
systematically analyzing data
Discuss how to interrelate these
models
Apply principles to your own data
Before discussing your projects...
4. “Data is not the
plural of
anecdote”
Qualitative data should...
relate to a research question or
concern
relate to each other
help us to draw testable inferences
provide evidence for claims
5. Ways to analyze qualitative data
Triangulating
Coding
Memoing
Modeling
7. Three kinds of
models
(Depending on how you count
them)
network: for nonsequential
relationships
flow: for sequential relationships
matrix: for ordered comparisons
8. Models in qualitative research...
Models: visual representations that allow you to abstract
relations at each level and see patterns.
Not always used
But: useful for visualizing and exploring specific types of
relationships in the data
Specifically, useful for spotting, testing, verifying, and
elaborating patterns in the data
And consequently, for developing further hypotheses
9. Network
diagrams
The point: Nonsequential
relationships
The payoff: You can see how
different things relate along
specific lines (e.g., where they
are coordinated, where they
contact each other)
Examples from “Chains and
Ecologies”: Genre Ecology
Models (Resource Maps)
10. Flow diagrams
The point: Sequential relationships.
The payoff: You can see sequences
and decision points.
Examples from “Chains and
Ecologies”: Communication
Event Models (Handoff Chains)
11. Matrixes (tables)
The point: Ordered comparisons
The payoff: You can compare things
(in rows) using the same criteria
or characteristics (columns)
Examples from “Chains and
ecologies”: STG Tables
(Triangulation Tables)
Prepare for report Write report Deliver report
Elizair previous month’s report, highlighting
and annotations on previous
month’s report, emails with client,
spreadsheet of projects, IMs and
talks with Craig, WikiAnswers
emaiils from customer,
BRILLIANCE, report template,
notes, email to Sonia
Final draft of report, client
presentation, PowerPoint slides
Craig previous month’s report, highlighting
and annotations on previous
month’s report, emails with client,
keyword logs, text file listing
projects, IMs and talks with Dani
Emails from customer,
BRILLIANCE, report template,
notes, email to Sonia
Final draft of report, client
presentation, PowerPoint slides
Dani previous month’s report, highlighting
and annotations on previous
month’s report, emails with client,
notebook listing projects, IMs and
talks with Craig
Emails from customer,
BRILLIANCE, report template,
notes, email to Sonia
Final draft of report, client
presentation, PowerPoint slides
Sonia Email from Elizair, emails with
customer, talk with Elizair
Final draft of report, Cover email to
client, client presentation,
PowerPoint slides
12. Other sorts of
models?
Think in terms of other relationships
you could explore in your qualitative
data:
Heat maps?
Word clouds?
Traffic flow?
Combinations of other models?
14. Interrelating
models
Done well, this can provide
further insights
Each model lets you visualize and
test a relationship.
Can you interrelate the insights
from different models?
(See “Chains and Ecologies”)
16. Please introduce
yourselves and
your projects
Your name
Your project (in a sentence)
Your research question/concern
The kind of data you’re collecting
The data you brought today
In a few sentences...
17. As we look at
each project...
We’ll collectively answer these
questions.
What is the research
question/concern? What
relationships should we explore to
get to it?
What relationships might we
explore with network diagrams, flow
diagrams, and/or matrixes?
Are there relationships we can’t
model with them? How might we
model these?
What actionable next steps should
the researcher take?