A four-stage flow for designing assignments where student thinking is visible, genuine, and difficult to offload.
📐
Stage 01
Assignment Design
Design the core task with Fink's dimensions in mind. Identify which dimensions the assignment already targets and which are absent. For each absent dimension, ask whether adding it would require minimal additional student time but reveal meaningful thinking.
Identify target Fink dimensionsCheck for Integration and LH2L gapsBuild in personal context AI cannot accessDefine what "excellent" process looks like
🎯 Key question: Which dimensions does this assignment genuinely require?
↓
Once the core task is designed, choose a thinking routine checkpoint that surfaces the process
🔄
Stage 02
Thinking Routine
Add a short, structured thinking routine before, during, or after the assignment. The routine is where the grade lives. It makes the student's reasoning process visible in a form that is idiosyncratic, personal, and difficult to fabricate without genuine engagement.
Connect-Extend-ChallengeThink-Puzzle-ExploreI Used to Think / Now I ThinkSee-Think-WonderDesign Decision LogProcess Annotation
⏱ Target: under 15 minutes of additional student time
↓
The routine generates a written record of the student's thinking process
🔍
Stage 03
Observable Evidence
The student's thinking process becomes visible in the routine output. Authentic thinking produces specific, idiosyncratic evidence: named connections to prior experience, genuine model revision, questions only this student would ask. Look for specificity as the primary signal.
Named prior course or projectDescribed moment of confusionIdentified genuine knowledge gapSpecific "I used to think / now I think"Idiosyncratic question or connection
🔑 Specificity is the signal. Generic fluency is the red flag.
↓
Evidence of thinking informs what and how you teach next
💡
Stage 04
Assessment Insight
The visible thinking gives you information the final product cannot: where students are genuinely confused, which connections they are making, whether they are building a coherent mental model or accumulating isolated facts. Use this to calibrate your next class session, not just your gradebook.
Reveal and correct misconceptionsIdentify students who need supportCalibrate next class discussionDocument learning for SoTLProvide targeted, low-effort feedback
📊 The thinking record is also your formative assessment data
Reading the Evidence: Authentic Thinking vs. AI Offloading
Authentic thinking looks like...
Names a specific prior course, project number, or system by name
Describes a moment of genuine surprise or confusion with a specific cause
Poses a question that only makes sense given this student's knowledge state
Shows model revision: "I assumed X, but now I see Y because..."
Connects material to a personal or professional experience the student actually had
Articulates a position they are willing to own and defend
Offloaded responses look like...
Generic connections ("like other algorithms I have seen")
Claims something was "interesting" or "new" without identifying what changed
Asks a question answerable by rereading the assignment
Describes growth in general terms with no named moment or experience
Uses correct vocabulary fluently but never situates it in a specific context
Could have been written by anyone who read the topic summary