Students arrange scrambled code blocks into correct programs. Stuck? The problem simplifies — fewer distractors, merged blocks, reduced choices. Research-backed and faster than writing from scratch.
A proven scaffolding technique from computing education research.
Students can identify correct code before they can write it from scratch. Parsons problems use this — arrange given blocks instead of facing a blank editor. Research shows it's faster with comparable learning gains.
Fixed problems leave some students stuck and others unchallenged. Adaptive Parsons track each learner's interaction — then the student chooses how to get help: reveal a subgoal, remove a distractor, merge blocks, or request a hint.
Scaffolding works best when learners have a say. Students pick the support that fits their struggle.
Show the label for one section of the problem. Clarifies purpose without giving away the answer.
Eliminate one wrong option. Reduces choices while keeping the core decisions intact.
Combine related lines into one chunk. Fewer pieces to arrange, same concept to learn.
Receive targeted feedback on what's wrong in the current arrangement without changing the problem.
Three research-backed strategies that respond to each learner's struggle — applied when the student asks for help.
Each correct block is paired with a plausible-but-wrong alternative. Research shows this boosts post-test scores ~11 points — students pay closer attention to the details that matter. When cognitive load is too high, distractors are removed adaptively.
When a student keeps reordering without progress, the system merges related lines into larger chunks. Fewer arrangement decisions, same underlying concept. Scaffolding research calls this "reducing degrees of freedom."
Each problem is decomposed into labeled subgoals — "initialize loop," "update accumulator," "return result." Research shows students given subgoal labels outperform those without. Labels make the purpose of each section visible, helping learners recognize patterns that transfer to new problems.
Every feature maps to a finding from computing education and learning science.
Every drag, drop, and retry is logged. See which distractors students pick, where they backtrack, and what misconceptions persist.
Problems simplify automatically based on each learner's interaction. No difficulty levels to set or hint sequences to author.
Parsons problems are faster than write-code tasks. More reps on the same concept in a single class period.
See which subgoals each student masters and where they stall. Pinpoint structural gaps, not just right-or-wrong scores.
Distractors map to real errors. Repeated wrong picks reveal exactly what a student misunderstands.
Parsons sit alongside textbook content, write-code exercises, and Jupyter notebooks.
See how adaptive Parsons compare to static versions and traditional code writing.
| Feature | Alps Adaptive Parsons | Static Parsons | Write-Code Tasks |
|---|---|---|---|
| Difficulty adjustment | ✓ Real-time, per-student | ✗ Fixed for all students | ✗ None |
| Distractor handling | ✓ Paired + adaptively removed | ✗ Fixed or absent | ✗ N/A |
| Block merging | ✓ Available when student struggles | ✗ Not supported | ✗ N/A |
| Subgoal labels | ✓ Built in; revealable as scaffold | ✗ Not supported | ✗ N/A |
| Student agency | ✓ Student chooses scaffold type | ✗ No choice | ✗ N/A |
| Misconception detection | ✓ From distractor selection patterns | ~ Limited | ✗ Manual grading |
| Time efficiency | ✓ Fastest (constrained + adaptive) | ✓ Fast (constrained) | ✗ Slowest (open-ended) |
| Teacher visibility | ✓ Full interaction traces | ~ Completion only | ✗ Final submission only |
Start with the free Educator Basic tier. Your students will experience research-backed code scaffolding from day one.