Creating, Refining, and Validating Automated Discourse Codes is an interactive CSCL 2019 half-day pre-conference workshop where participants will work together to analyze their own data using the nCoder and rho toolkits.
The goals of the workshop are for each participant to:
-
- Understand how to combine qualitative and quantitative perspectives for text analysis and use two key tools for conducting such analyses,
- Create codebooks for code validation and publication,
- Develop, test, and validate automated classifiers to code text data, and
- Find a meaningful result from their own data that they can refine in future work, ideally leading to the publication of a special issue on current approaches to automated text analyses.
This workshop will provide an overview of the methodology and hands-on experience with two tools:
nCoder is a learning analytics platform used to develop, validate, and implement automated coding schemes.
nCoder is available as an R package (ncodeR) and interactive website
Rho, a statistical method for establishing the generalizability of inter-rater reliability metrics.
rho is available as an R package (rhoR) and interactive website
This is a data-intensive workshop where you get to develop codes for your own data, and discuss your process and results with other participants and with experts in automated data analysis.