Trial chats/activities as a Student

One of the great things about Flint is that it doesn’t give students too much support; however, I’m finding it a challenge to understand how much support they will receive.

For example, if I ask Flint to provide a research summary on a topic, such as Montessori approaches to education, and offer me links to explore further or verify the accuracy of ideas, it does so without hesitation. If a student asks the same question in a chat, it does not provide links and is reluctant to do so. The same thing has happened in activities.

I would appreciate the possibility of trying something out as a student so I can see what responses they are likely to get, so I can adapt my recommendations and approaches to students accordingly.

I have had a number of instances where I think Flint is going to provide students with the same sort of information it does for me when I try a task, and that leads to some frustration in the classroom.

Responses

Itzel Madero Hernandez

Thank you for sharing this. One thing that might be helpful, if you haven’t already tried it, is using the manual simulation to step fully into the student experience. That’s where the differences you’re noticing tend to show up most clearly, especially around things like links, depth of explanation, and how much Flint is willing to offer unprompted.

Something else that’s been useful for me is being really explicit with tutor guardrails during setup. Those guardrails seem to play a big role in preventing over-support and keeping student responses intentionally lighter than what we see as teachers.

For example: “Avoid providing external links or full summaries. Ask guiding questions first and prompt the student to identify key ideas before adding detail.”

With that in place, a teacher asking for a research summary with sources will get a very different response than a student asking a general question, which helps set expectations in the classroom.

It might not solve everything, but seeing the student-side behavior through simulation and tightening guardrails has helped me anticipate what students will actually receive and adjust my guidance accordingly.

Jacob Edington

Hi Keri-Lee Beasley! Great question! Please feel free to add a personal email to your workspace and set it up as a student. Once you have done that, you can log-in and fully test interactions based on being a student.

Matt Lyons

I know the folks at Flint read these, so hopefully they can mine this for ideas: Can we solicit prompts from our users that help shape the way Flint guides students through activities? It would be great to create an online document with lists of prompts for different situations. For my part, I would definitely use something like this as part of my activity prompt: “After asking students what part of the activity they need help with, provide two or three possible related topics where Flint can provide assistance.” (Sometimes students don’t know what they need or have a hard time being specific. This prompt guardrail can help move the conversation along.)

Emmett Gallagher

Hi Matt Lyons, I like what you’re proposing, and I imagine these types of prompts could be built into Flint’s activity builder, e.g., if a teacher/user requests that students revise using better word choices, Flint could offer the prompt to ask students to use a thesaurus and provide possible links. Not sure what the build-out for something like this would be, but I would think that a lot of the basics could be covered.

Karen Mihata

I’m growing comfortable with the manual simulations and appreciate the workarounds offered—super helpful. Even so, I would definitely appreciate a test student in my course or a ‘student view’ screen so I can gather more information before I walk into my classroom.

Keri-Lee Beasley

This is exactly what I mean. A “Student view”/practice is different from the feature where I can test the activity or chat as a teacher. A button to toggle between trialing as a teacher and as a student would be very helpful in future development!