Hypothesis annotation tool — dissertation research

Context

My doctoral research at Indiana University Bloomington focused on social annotation as a mechanism for learning — specifically, the Hypothesis tool used in university online and hybrid courses. The question I cared about: when students annotate shared texts together, what kinds of learning conversations actually happen, and what conditions help generative ones emerge?

The work sat at the intersection of learning sciences, educational psychology, and human-computer interaction.

What I did

Across the dissertation arc I worked with university instructors using Hypothesis in their courses, analyzed thousands of student annotations across dozens of course documents, and conducted multiple focus groups with both instructors and students.

The mixed-methods design paired:

• Qualitative content analysis of student annotations, using a coding scheme built specifically for studying collaborative learning in annotation contexts.
• Focus groups with instructors and students to ground the patterns I was seeing in the data against the lived experience of the courses.
• Iterative refinement of the coding scheme itself, which became its own contribution — a lens through which other researchers could study annotation conversations.

The body of work generated several peer-reviewed publications on expansive framing, social annotation, and strategies for instructors implementing annotation in online learning.

Outcome

The dissertation contributed three things I’m proud of:

1. A coding scheme for analyzing collaborative learning in annotation environments, usable by other researchers studying similar phenomena.
2. Empirical evidence about what kinds of annotation activities generate substantive learning conversations versus surface-level engagement.
3. Practical guidance for instructors designing annotation activities, grounded in the patterns the research surfaced.

The work translated directly into my industry UX research practice — the same “start with the user, ground claims in evidence, design for the activity not just the tool” instincts apply equally to product research.

Reflection

A dissertation is a long-form research training. What I took into industry research afterward wasn’t any specific finding from this work — it was the discipline of long-horizon study design, iterative coding under uncertainty, and the muscle of producing claims I could actually defend.

The other thing that stuck: most of what makes a learning environment work or not work is invisible at first glance. Reading dense annotation transcripts taught me to trust patterns that build slowly across data rather than conclusions that feel obvious early. That habit transferred directly into UX.