The TIDESS team has begun qualitative analysis on the data we collected from the tabletop user study to try to characterize the processes of how people learn from data visualizations on interactive tabletop displays. We collected both audio and video recordings, as well as logs of all the touch interactions (gestures) participants did with the prototype, during the study sessions of the user study.

To start, the team transcribed the user study session videos so that the transcriptions can be coded, or labeled with interesting behaviors and spoken utterances that occurred. We then timestamped the transcripts so that the participants’ words could be matched up with the gesture data that came from the application on the display. The team constructed a code book for the utterances by reviewing prior literature in learning sciences and collaboration learning, as well as some insights from our own past work and the goals of this study itself. The codes in the code book allow the team to characterize the group dynamics, collaborative work, and group meaning making. To refine the codebook, all of the team members coded sample transcripts, and we discussed any disagreements during our team meetings until we agreed on an initial coding procedure.

We are using MaxQDA, a qualitative data analysis program, to facilitate the coding. MaxQDA allows us to take all of the codes from the code book and use those codes to classify the speech and actions in the transcripts. We’ve already begun coding the transcripts, and then next we will analyze the coded transcripts to see where the interactions with the prototype helped or hindered group learning.

I am a 3rd year Computer Science student at Brooklyn College, in the INIT lab for the summer. It has been really interesting to see how people interacted with the interactive touchscreen tabletop display. Working on this team helping to analyze the data has been fun, and I look forward to continuing the analysis of the data.

by Jeremy Alexandre

0