Shape: exercise sheet

A set of exercises built as a sequence. Each new type of problem opens with a worked example, then a problem with its last steps left to do, then a few whole problems. A mixed review brings back the types of earlier sheets, and every problem offers a hint before its answer and its solution, all folded.

Skeleton

# <What the sheet trains, named by the skill>

> One or two sentences: the types of problem it holds and what it builds on.

**Shape:** [exercise sheet](<link to this file>).

## Types               each type of problem with an id
- **T1** <type>
- **T2** <type>

## <Type T1>
### Worked example (T1): <what solving it achieves>
    a solved problem in the core report's form: plan, steps named by their goal
### Complete (T1): <problem>
    the first steps given, the last ones left to the learner
### Problem (T1): <problem>
    <details><summary>Hint</summary> ... </details>
    <details><summary>Answer</summary> ... </details>
    <details><summary>Solution</summary> ... </details>

## Mixed review        problems of the types of earlier sheets, in mixed order

## See also            the lesson it practises, the sheet before and after

The blocks

Block	Rule	Grade	Evidence
Types	named, with ids; every item carries its type	convention, needed by the checks below	exercise sets
Worked example first	the first item of a new type is solved in full, in the core report's form	measured, for novices	exercise sets, strand 3; worked examples with subgoals
Complete	completion problems leave the last steps before whole problems	measured, for novices: removing steps one by one beats a jump to whole problems (Renkl 2004, a controlled trial, in the core); this review's fading studies were read only in summaries (secondary). Parsons problems are a completion format for code	worked examples with subgoals; exercise sets, strand 3
A short block	a few problems of the new type, not a long run	supported with caveat: one lab study, read in a summary (secondary), where nine problems did no better than three	exercise sets, strand 6
Mixed review	problems of earlier types, in mixed order	measured in classroom trials, mostly mathematics; a year-long trial found no gain on the final test	exercise sets, strand 2; against
Hint, answer, solution	in that order, each folded; the solution released after the attempt	convention on paper; hints are measured only in software	exercise sets, strand 5
Variation	within a type, one feature changed at a time for novices	supported with caveat: variability helps experienced learners and hurts novices	exercise sets, strand 4

The baseline this shape must beat is the blocked sheet: one run of same-type problems per lesson, answers at the end (baseline).
Not tested as a whole: no study tests a sheet built from all these parts, or the share of mixed review it should hold. The thresholds of the checks are tunables, marked PROVISIONAL in checks/config.toml.

How it derives from the topic

Topic block	Here	Change
Title	what the sheet trains	none
Summary	the quoted line under the title	none
Body	Types, one part per type, Mixed review	added
Contents	none: the worked examples are parts, not topics	left empty
See also	See also	none
(metadata)	the Shape line	names this shape

Checked by

Check	What it verifies	Kind
`sheet`	every item carries a declared type; the first item of each type is a worked example	exact
`sheet`	each type has a Complete item in its own part, between the example and the problems	heuristic, a warning: fading helps novices and reverses for experts
`sheet`	every problem folds its answer and its solution; a hint comes before them	exact
`sheet`	a Mixed review holds problems of at least `mixed_types` types	heuristic, PROVISIONAL threshold
`sheet`	no type has more than `max_block` problems in its own part	heuristic, PROVISIONAL threshold
standard	the headings Types, Mixed review, See also	exact

An example that passes every check: binary search.