Teachers play a crucial role in supporting students' learning. It is crucial to also support teachers in their teaching. Traditional means to support teachers are highly effective but also repetitive, not personalized or not scalable, and infrequent. As classrooms become instrumented with educational technologies, opportunities emerge to provide teachers with feedback through data from these technologies. A small body of work has started to look at supporting teachers with data and feedback outside of class to promote reflection. However this work does not investigate the multiple dimensions of teachers' data needs and does not explore how data affects or can support teachers' goal-setting and behavior change. In this dissertation, I investigate how technologies can support teachers' reflection-for-action outside of class, with data from educational technologies. My aim is to help improve teachers' practices and support their long-term behavior change. I initially explore and better understand teachers' data needs and design dashboard prototypes based on those needs. I then investigate how those dashboards affect teachers' practices and how to best support their reflection-for-action through data and technology as a first step towards improved practices and behavior change.

In Part 1, I explore teacher data needs in relation to student data. Findings show that teachers manually generate student data and use it to drive instruction. Based on these findings, I designed a dashboard and investigated how it affected teachers. The dashboard influenced what teachers knew about their students, which affected their lesson plan, and in turn guided what they covered in the class session. I demonstrated that data can affect teacher knowledge, decision making, and actions in the classroom, thus leading to behavior change.

In Part 2, I explore teacher data needs in relation to their data, design a dashboard that shares with teachers their own data, and investigate how to support their reflection-for-action with motivational feedback. Findings showed that teachers are interested in their data and in how their behaviors affect their students. They reflected on their performance and set goals to improve. Through proxies for behavior change, they showed their willingness, readiness, and intentionality for behavior change, an important first step towards improving their practices. Finally, teachers who received social comparison motivational feedback scored higher in behavior change proxies compared to teachers who received such feedback through verbal persuasion.

In Part 3, I explore behavior patterns and relationships in teacher and student data. Findings showed potential for improvement in teacher behaviors and weak to moderate correlations in teacher and student data, hinting at the value of nonverbal immediacy. In co-design studies with teachers, I then investigate how to integrate teacher and student data while supporting reflection-for-action. Findings showed instructors value relationships between teacher and student data, want to see data in a spatial and temporal form, and are interested in activity information combined with student engagement. Support for self-efficacy and value helped them assess performance, set concrete goals and provide actionable suggestions on behaviors to change.

This thesis contributes to research at the intersection of the Learning Sciences and Technologies and Human-Computer Interaction. I create a better and deeper understanding of teachers' data needs, their behavior patterns, and of how to support and influence their reflection-for-action. I create dashboard prototypes based on those needs and provide evidence on how these dashboards affect teachers' reflection, goal setting and behavior change. I also provide a theoretical framework and design guidelines for designers of technologies that share data with teachers to support reflection-for-action and long term behavior change. Finally, I present a concrete example of how data can support professionals in their workplace.

350 pages

Thesis Committee:
Amy Ogan (Chair)
John Zimmerman
Geoff Kaufman
Marsha Lovett (Eberly Center/Psychology)

Jodi Forlizzi, Head, Human-Computer Interaction Institute
Martial Hebert, Dean, School of Computer Science

Return to: SCS Technical Report Collection
School of Computer Science homepage

This page maintained by reports@cs.cmu.edu