Exploring Chat Data Through the Lens of Facilitation

• In Event: Advancing Equity in Assessment Through Small-Team Collaboration in Mathematics

Abstract

Objectives
We are exploring the nature of facilitation as applied to the discourse of students collaborating on mathematics tasks in small teams through a text-based chat. Adult mentors served as facilitators. To analyze the chat logs, we developed a coding scheme using both theory-driven and data-driven approaches. Our goal is to identify the frequencies of chat moves used by facilitators and how this supports student mathematical discourse.
Theoretical Framework
Talk moves (Michaels & O’Connor, 2012) is a framework that was developed for teachers to support student academic discourse. The framework consists of 9 talk moves organized into four goals. The four goals include: “Individual students share, expand, and clarify their own thinking,” “Students listen carefully to one another,” “Students deepen their reasoning,” and “Students think with others,” (Michaels & O’Connor, 2012, p. 11). Example talk moves under the students deepen their reasoning goal are “How did you get that anser [sic]” or “Why not the last one?”
Methods
The research team prepared training sessions for facilitators, which included a research overview, training on the content, the learning progression (Authors, 2021; Authors, 2021), the Algebra Project’s 5-step curricular process (Moses et al., 1989; Moses et al., 2009) and talk moves. In addition, mock sessions were conducted in which facilitators practiced using talk moves in a chat with each other.
On Day 2, 20 teams participated (12 facilitated and 8 unfacilitated). Facilitated teams included an adult mentor and observer in the chat, while unfacilitated teams included students only. We focus on the facilitator chat from the facilitated teams.
In developing the coding scheme, the theory-driven approach was based on talk moves (Michaels & O’Connor, 2012), and the data-driven approach was based on our observations of moves used by facilitators that supported the student chat but that did not map to the talk moves framework.
Data Sources
Our data consist of the chat logs from facilitated teams. Only the facilitators’ chat turns will be coded but attending to the student chat assists interpretation of facilitator moves.
Analysis
The coding framework was developed based on an earlier, preliminary data collection. Each of the four goals from the talk moves framework was mapped to a main code, and each talk move was mapped to a subcode. Chat turns in the data that did not correspond to any of the talk moves informed the development of additional main codes. There were seven such codes, for a total of 11 main codes and 9 subcodes. In the new data set, each facilitator chat turn will be coded with a main code and subcode, if applicable. We will explore frequencies of chat moves and compare these frequencies to what we observed based on our preliminary data collection.
Scholarly Significance
Although there has been much work on collaborative problem solving (see Andrews & Rapp, 2015), the role of human facilitation in fostering productive academic discourse among students is less understood. This work could uncover the moves facilitators tend to make, and how those moves are related to student performance.

Authors

• Arianna M Gaston, The Young People's Project

• Edith Aurora Graf, Educational Testing Service