Coding tool supporting holistic characterization of online classroom discussion
Originally Published In
Presented at 8th Annual Emerging Technologies for Online Learning International Symposium, Dallas, TX, United States.
Coding based analysis of online threaded classroom discussions provides essential information regarding student interactions, development of ideas, and facilitation processes. Unfortunately, such in depth characterization is time consuming. In addition, it is challenging to holistically characterize discussion while still targeting specific focus areas. This presentation demonstrates a for-purpose coding program designed specifically to support holistic characterization of discussion as well as allow coding additions as desired.
The coding tool provides a tree interface showing the discussion threading and steps through coding post by post. The side-bar supports the tree showing the overall discussion thread and highlighting the active post. Then the main window shows the content of the post as well as check-boxes and menus for the codes. Pop-up boxes support full definitions and additional instruction for each code. Pop-up boxes also allow the previous post in the thread to be reviewed. This simple windows forms environment is intuitive and supports quick processing.
The coding tool is supported by a database which facilitates conversion of various forums and supports easy statistical analysis of the resulting data set. Currently, database conversion capabilities are available for BlackBoard and eCollege. Conversion requires downloading html files which are then automatically parsed and process. Conversion of other forums will require setting up the html parsing process but any forum then appears in the same threaded format in the coding tool. In addition, the database is structured so it can be directly loaded into common analytical tools.
Codes to holistically characterize the discussion are pre-populated in the coding tool with full instructions and definitions. These 31 definitions are aligned to the Community of Inquiry model (Weltzer-Ward, 2014). A multi-step process assessing validity, reliability, and utility of codes was employed to reduce an initial coding framework with 108 codes and 3 units of analysis to this set of 31 codes all utilizing the same post-level unit of analysis. This coding framework for asynchronous, online classroom discussion is grounded in the Community of Inquiry model, applicable to multiple discussion types, comprehensive, reliable, and efficient.
In addition to the built-in coding framework, the tool is able to accommodate additional post level codes as well. Boolean, Ordinal, and Nominal codes may be entered into a spreadsheet and the loaded into the coding tool. This supports targeted analysis of specific areas of interest to the researcher or educator.
Online classroom discussion is ubiquitous in higher education today, with both online and hybrid courses. This tool supports analysis of discussion and refinement of best educational practices in this medium. The adoption of this tool and holistic coding scheme by current research efforts will allow results from all studies employing the scheme to be easily compared, speeding the identification of best practices to improve learning for the millions of learners engaged in learning through asynchronous discussion. This presentation will demonstrate the tool and elicit discussion of next step for further automating and refining capabilities.