Interpretation Protocol (Word doc reference copies)


The WE1S Interpretation Protocol for topic models is implemented as a set of Qualtrics surveys with declared steps for observing and drawing conclusions from topic models. The surveys are designed to be mix-and-matched in sequence or in parallel depending on the needs of a particular research problem.

Below are Word doc copies of the surveys, which can be used for advance planning of research exercises. They can also be used to write collaborative drafts of responses to the surveys, including by members of a team not logged into the live Qualtrics survey (which can only be operated by one person or computer in a unique session.) For example, copy these docs into a Google Drive space and work on them collaboratively there.


Getting an overview of a topic model:

Specific research tasks:

Final research steps:

Module 0
User Training
Module 0...
Module 1
Overview of model
Module 1...
Module 2
Representative topics
Module 2...
Module 3.a
Analyze a topic
Module 3.a...
Module 3.b
Analyze a cluster of topics
Module 3.b...
Module 3.c
Analyze a keyword
Module 3.c...
Module 4.a
Compare sets of topics (multiple topics)
Module 4.a...
Module 4.c
Compare two keywords
Module 4.c...
Module 5
Compare two parts of corpus
     * part to whole
     * 2 metadata sets
     * 2 time ranges
     * compare to "random" corpus
Module 5...
Module Z
Add-on steps for collaborative interpretation
Module Z...
Module 7
Analysis & synthesis of interpretation results
Module 7...
Report Module --
Instructions for writing report
Report Module --...
= Created
= Created
Module 6
Compare two different
topic models
Module 6...
Module X
Make & document your
own workflow
Module X...
Report
(use Google Doc
research report template
Report...
Module Y
Document use of additional methods and tools
Module Y...
Viewer does not support full SVG 1.1