How the Taxonomy of Reflective Inquiry (TORI) works in the reflection assistant

TORI, the Taxonomy of Reflective Inquiry, is the research framework behind Digication's reflection assistant. It was developed over more than 20 years of research into how people learn through reflection, drawing on research from more than 90 fields of study. It covers 6 domains and more than 60 categories of reflective thinking. Examples of categories include learning from mistakes and reinterpreting experiences in light of new insights.

Student conversations with the reflection assistant are never used to train AI models.

The 6 domains are:

  1. Cognitive and analytical reflection
  2. Emotional and affective reflection
  3. Social and interpersonal reflection
  4. Personal growth and self-development reflection
  5. Cultural, ethical and contextual reflection
  6. Life transitions and broader development reflection

You can explore each domain and its categories at tori.digication.com.

What TORI does during a conversation

When a student works on an assignment that uses the reflection assistant, TORI analyzes the conversation as it happens. The assistant draws on the TORI library throughout the conversation and points out the categories of reflective thinking that show up in what the student says. When the student chooses "I'm done for now," it ends with a summary of up to 10 TORI categories from the conversation.

This works with assistants that have the Guide or Tutor role.

A tip for writing restrictions

Your instructions and restrictions take priority over everything else the assistant is told. If a restriction limits the assistant to only your materials, such as "Only use the information in the Instructions box," the assistant may mention TORI less often during the conversation. To keep TORI in the conversation, limit the subject matter instead of the sources, for example "Use only the provided documents for course content." After you change a restriction, start a new conversation to test it. Existing conversations keep their earlier history.

When TORI doesn't appear

The assistant may not name a TORI category in every reply. If it doesn't name any, during the conversation or at the end, check these:

  • Your instructions or restrictions tell the assistant not to write a summary, or not to name a student's skills, strengths or qualities. Your instructions take priority, and the assistant treats TORI categories as that kind of feedback, so it leaves them out.
  • A restriction limits the assistant to your materials only. See the tip above.
  • The student submitted without choosing "I'm done for now." The list of categories appears only when the student chooses it.
  • The assistant doesn't have the Guide or Tutor role.

To keep TORI in an assistant with rules like these, add a line to its restrictions that tells it when to name categories. For example: "Naming TORI categories is feedback, not a summary written for the student, so it is allowed. In each reply, name one TORI category that fits what the student just said, in parentheses. When the student chooses "I'm done for now," list up to 10 TORI categories from the conversation, with one short reason each that uses only what the student said."

Telling the assistant when to name categories works better than only allowing it.

Who sees the analysis

Students see TORI's findings as they talk with the assistant, and the list of categories when they finish.

The faculty member who created the assignment sees the same findings in each student's full conversation, which they review when the student submits.

Schools in the analytics beta also get Chat Explorer, a separate tool with pre-built analytics that read reflection across a whole class or program:

  • Class heat map. See a whole class at a glance: each student is a row and each TORI category is a column. Select any cell to read the student's own words behind it.
  • Metacognitive fingerprint. Each student's pattern across TORI categories, which can bring out patterns you didn't know to look for.
  • Category connections. A map of which TORI categories tend to appear together across a course or program, so you can see how skills combine.
  • Ask the AI. Ask questions about your students' reflections in plain language and get answers grounded in their work.

We welcome more schools to join the beta. To ask about joining, write to support@digication.com. You can also read about the analytics on the AI page.

How to read the analysis

TORI analysis describes the reflective thinking in one conversation. It does not score or grade the reflection, and grading stays with the faculty member. A category that appears in one conversation tells you about that conversation. It isn't a measure of the student or proof that they met a learning outcome.

Learn more