Skill: course-build → write-mcqs → write-explanation @ df153ae; v2 (2 Oct 2026, the three reviews applied) @ b0fa825
Topic 2 — How to run a fair test — question types in the PP100 bank, one exemplar each
James reviews ONE exemplar per type before the rest of a bank is trusted. This sheet lists the ten types the Grade 4 Topic 2 bank runs on, prints one exemplar per type in full as it ships in content/G4T02_pp100_bank.py, and gives the design-move inventory (one line per item: the move, the scene kind, the nearest lesson-side instance) so a reviewer can check freshness in one look. Item ids are the bank's positions (PP-1 … PP-57). Cluster: the whole topic (six atoms), capstone FOR4-004.
Bank: 57 items — 24 easy · 20 medium · 13 hard; 52 four-option MCQs, 2 multi-selects (hard; the answer is a set of two), 2 text-entry, 1 numeric. Easy items lead the bank and are the bulk; every hard item carries a "hard because" line (in the bank file beside the item, and in the table below); the typed items are the term recall (PP-51, PP-52) and a number read and subtracted from a drawn table (PP-54) — 3 of the 13 hard items (v2: the plant-food item, whose stem lists the four things, became a four-option MCQ, PP-44, medium; the old PP-44 is now PP-45). The other hard items are judge-the-reason MCQs (the verdict in the stem, four single-statement reasons) because the move is a judgement, and a judgement typed is a phrase, which the typed-item rule refuses. Keys (v2): 0 ×13, 1 ×14, 2 ×15, 3 ×10 (no letter above 29 %; no run of four; no ABAB run). Difficulty rises with the atom's position: SR4-001 6 easy / 2 medium / 1 hard … FOR4-004 2 / 3 / 4.
Bar list (the type sheet's list of real things a generated bank must never name): no "best / nicest / cutest / most fun" question is ever a key for "can be tested"; no scene from the lessons (a toy car on floors, a ramp, two bean plants with water and light, paper towels dipped, ice cubes in sun and shade, paper planes with folds, a flashlight with new batteries, a ball on grass / sand / concrete, a paper-towel plan board, a cocoa-cup plan, spinning tops, a kite tail, a sled on snow and ice except PP-42 / 54 / 55 which own it) is reused with a noun changed; a named child is Felipe, Niall or Maeve only; "variable" and its kinds never appear.
The ten types
| # | Question type | Exemplar | Items (typed items starred) | Count | Hard end |
|---|---|---|---|---|---|
| 1 | Can the class test it? Four questions; pick the one that names something to try out and watch (or the one that does not). | PP-1 | PP-1, PP-2, PP-3, PP-4, PP-5, PP-6 | 6 | none — all easy |
| 2 | Say it so it can be tested: a vague question ("which is the best?") and four rewrites, one of which names something to try out and watch; the hard item asks what to do FIRST with a question nobody can test. | PP-25 · PP-45 (hard) | PP-25, PP-26, PP-45 | 3 | MCQ (PP-45: apply the rule to the next step) |
| 3 | The one thing changed: a described test (or its question) and four things in it; pick the one changed on purpose. | PP-7 | PP-7, PP-8, PP-9, PP-10, PP-23, PP-27, PP-28, PP-29, PP-30, PP-44 | 10 | none — easy and medium (v2: the typed hard end was an MCQ in disguise, review 2 Essential) |
| 4 | Keep everything else the same: which of four things must stay the same; which thing changed that should have stayed the same; the hard items SELECT THE TWO things that must stay the same (multi). | PP-11 · PP-46 (hard) | PP-11, PP-12, PP-13, PP-22, PP-31, PP-32, PP-33, PP-46, PP-47 | 9 | multi (PP-46, PP-47: a set of two) |
| 5 | What you measure: four measurements, one of which is the thing the question asks about after the change; the hard items give the reason a measurement fails, or chain result → tool. | PP-14 · PP-48 (hard) | PP-14, PP-15, PP-16, PP-24, PP-35, PP-36, PP-48, PP-49 | 8 | MCQ (PP-48 reason; PP-49 result → tool) |
| 6 | Fair test or not: four described tests, pick the one that changed only one thing (the stem never prints the term); or a test that is not fair and the two things that changed. | PP-17 | PP-17, PP-18, PP-38, PP-39, PP-40 | 5 | none — easy and medium |
| 7 | Name the test: the term by recognition among sister names (easy), then TYPED from a described case (hard). | PP-19 · PP-51* (hard) | PP-19, PP-51, PP-52 | 3 | text (PP-51, PP-52: "fair test"; accept lists carry "fair" alone and "it is a fair test") |
| 8 | The plan: which line goes under a heading; which question fits a plan; which line is missing or fails to name a measurable result (a drawn board); the hard item REPAIRS a plan. | PP-20 · PP-57 (hard) | PP-20, PP-21, PP-37, PP-41, PP-43, PP-57 | 6 | MCQ (PP-57: repair the 'We change' line) |
| 9 | Read the results: a drawn results table; which surface; the hard items TYPE the difference between two cells, or turn the table back into the plan's one-line answer. | PP-42 · PP-54* (hard) | PP-42, PP-54*, PP-55 | 3 | numeric (PP-54: 13 meters) |
| 10 | Judge the result or the claim: the verdict is in the stem; the options are four single-statement reasons. | PP-34 · PP-56 (hard) | PP-34, PP-50, PP-53, PP-56 | 4 | MCQ (PP-50, PP-53, PP-56) |
Every option is ONE complete statement; no option bundles a verdict with its reason (the verdict sits in the stem for type 10). No stem is negative ("Which sentence explains why …?", never "Why not?").
The exemplars, in full
Type 1 — PP-1 (easy, SR4-001)
Stem: Four questions are written on a class's board. Which one can the class test by trying it out?
Options: (a) Which board game is the most fun? (b) Which kind of mitten keeps a hand warmest in the snow? (c) Which puppy is the cutest? (d) Which movie is the best? — key (b)
Why: A question you can test names something you can try out and watch. / The class can wear each mitten in the snow and read a thermometer inside it. / 'Most fun', 'cutest' and 'the best' mean something different to each person, so nobody can try them out.
Type 2 — PP-25 (medium, SR4-001)
Stem: Felipe asks: which blanket is the best? Which question asks about the blankets in a way the class can test?
Options: (a) Which blanket is the best to sleep under? (b) Which blanket keeps a hot water bottle warm the longest? (c) Which blanket is the nicest color? (d) Which blanket feels the best? — key (b)
Why: A question you can test names something you can try out and watch. / The class can wrap a hot water bottle in each blanket and read its temperature later. / 'Best to sleep under', 'nicest color' and 'feels the best' mean something different to each person.
Hard end — PP-45 (hard because: the student applies the rule to decide the NEXT STEP, not the verdict): Maeve wants to find out which marker is the best. What should she do first? — (a) Try out all the markers and vote. (b) Ask Miss Alvarez which marker is the best. (c) Change the question to one the class can try out. (d) Test only her favorite marker. — key (c).
Type 3 — PP-7 (easy, SR4-002)
Stem: Maeve builds towers of 5, 10 and 15 of the same blocks on the same table, and blows on each tower with the same fan. What is the one thing Maeve changes on purpose?
Options: (a) the kind of block (b) the table (c) the fan (d) the number of blocks — key (d)
Why: The one thing you change on purpose is the thing your question asks about. / Maeve's towers have 5, 10 and 15 blocks; the blocks, the table and the fan are the same. / So the one thing Maeve changes is the number of blocks.
Medium end — PP-44 (v2; was the typed PP-45): Felipe plants the same kind of seed in three matching pots, with 1 spoon, 2 spoons and 3 spoons of plant food, and waters them the same way. Which thing does Felipe change on purpose? — (a) the amount of plant food (b) the kind of seed (c) the pot (d) the water — key (a). Typing stands for a recalled term or a produced number only; the stem names the four things, so the item is an MCQ.
Type 4 — PP-11 (easy, SR4-003)
Stem: Niall tests which of three sponges holds the most water by dipping each one in a bowl. Which of these must stay the same?
Options: (a) the kind of sponge (b) how long each sponge soaks in the bowl (c) how much water each sponge holds (d) the order the sponges go in — key (b)
Why: Everything except the one thing you change must stay the same. / Niall changes the kind of sponge on purpose, and how much water each holds is the result. / So the soaking time must stay the same for every sponge.
Hard end — PP-46 (multi; hard because: the answer is a SET — two kept-same things picked from a list that also holds the changed thing and the result): A class tests whether a longer string makes a metal ball on a string swing more slowly. Select the TWO things that must stay the same. — (a) the length of the string (b) the metal ball on the string (c) how far the ball is pulled back (d) how many seconds each swing takes — keys (b), (c).
Type 5 — PP-14 (easy, SR4-004)
Stem: A class tests which kind of soil lets water run through fastest. What should the class measure?
Options: (a) how dark each soil is (b) how many seconds the water takes to run through (c) how much each bag of soil costs (d) how heavy each pot is — key (b)
Why: What you measure to see the result is the thing your question asks about, after the change. / The question asks how fast the water runs through. / So the class times how many seconds the water takes to run through each soil.
Hard end — PP-48 (hard because: the verdict is given and the student picks the REASON a measurement fails): Maeve tests whether a heavier backpack makes her walk more slowly, and she writes down the mass of each backpack. What is wrong with what she measured? — (a) She should have weighed herself as well as the backpack. (b) She wrote down the thing she changed instead of the result. (c) The backpack's mass changes as she walks. (d) She should have used three backpacks instead of two. — key (b). PP-49 (hard because: two steps — decide what the result is, then the tool): A class puts matching ice cubes into salt water and into fresh water, to see which melts first. Which tool measures the result? — a thermometer / a digital scale / a tape measure / a stopwatch — key stopwatch.
Type 6 — PP-17 (easy, SR4-005)
Stem: In which of these tests did only one thing change?
Options: (a) Two pinwheels of the same size, held in the same wind, one made of paper and one of foil. (b) Two pinwheels of different sizes, held in different winds. (c) Two pinwheels of the same size, one made of paper held in the wind and one made of foil blown on. (d) Two pinwheels of different sizes, one made of paper and one of foil. — key (a)
Why: A fair test changes only one thing and keeps everything else the same. / In the first test only the material changed: paper or foil. / The other three tests each changed two things at once.
Type 7 — PP-19 (easy, SR4-005)
Stem: A test changes only one thing and keeps everything else the same. What is this kind of test called?
Options: (a) a true test (b) a fair test (c) a quick test (d) a safe test — key (b)
Why: A test where only one thing is changed, and everything else stays the same, is called a fair test. / This test changes only one thing. / So this test is a fair test.
Hard end — PP-51\ (text; hard because: the term is RECALLED from a described case, beside its near miss):* Maeve changes only the kind of ball and keeps everything else the same. Niall changes the kind of ball and the height too. What kind of test is Maeve's test? Type its name. — accept: fair test · a fair test · fair tests · the fair test · fairtest · fair.
Type 8 — PP-20 (easy, FOR4-004)
Stem: A class plans to find out which surface lets the same shoe slide farthest when it is pushed the same way. Which line goes under 'We measure'?
Options: (a) the kind of surface (b) how far the shoe slides, in centimeters (c) the same shoe and the same push (d) which surface lets the same shoe slide farthest — key (b)
Why: The 'We measure' line names the result: the thing the question asks about, after the change. / The question asks how far the shoe slides. / So the class measures how far the shoe slides, in centimeters.
Hard end — PP-57 (hard because: the student REPAIRS a plan, chaining the one-thing-changed rule and the keep-the-same rule into its lines): Niall's question is: which surface lets the same block slide farthest? His plan says: We change: the surface AND the size of the block. Which change makes his plan a fair test plan? — (a) Change only the surface, and use the same block every time. (b) Change the surface and the push, but keep the block the same. (c) Change the size of the block, and keep the surface the same. (d) Change the surface, the block and the push, so the test is big enough. — key (a).
Type 9 — PP-42 (medium, FOR4-004; figure results_pp_sled: Surface | How far the sled slid (meters): packed snow 12, loose snow 7, ice 20)
Stem: Here are a class's results for the same sled let go down the same slope onto three surfaces. On which surface did the sled slide farthest?
Options: (a) packed snow (b) loose snow (c) ice (d) the same distance on all three — key (c)
Why: Read the distance for each surface: packed snow 12 meters, loose snow 7 meters, ice 20 meters. / 20 is the longest distance. / So the sled slid farthest on the ice.
Hard end — PP-54\ (numeric; hard because: the student PRODUCES a number from two table cells and a subtraction):* Look at the results for the sled on three surfaces. How many meters farther did the sled slide on the ice than on the loose snow? Type the number. — accept 13 · 13 meters · 13 m · thirteen; why shows 20 − 7 = 13.
Type 10 — PP-56 (hard, FOR4-004; hard because: a claim is judged — the verdict is in the stem and the student picks the reason, chaining 'keep the same' into the plan)
Stem: A class tests which surface lets the same hockey puck slide farthest. On the rug, Felipe pushed the puck harder, and it went farthest. Miss Alvarez leaves the rug result out of the table. Which sentence explains why?
Options: (a) A puck cannot slide on a rug. (b) The rug result was too long to fit in the table. (c) Only the first two runs count in a test. (d) The push changed as well as the surface. — key (d)
Why: In a fair test, everything except the surface stays the same. / On the rug the push changed too, so two things changed on that run. / So the rug result tells the class nothing, and the class does that run again with the same push.
Design-move inventory (move · scene kind · nearest lesson-side instance)
Lesson-side instances the bank was read against: the car on three floors (L02–L06 headaches, L04 tape, L06 table); the push on the carpet (L03); the two bean plants with water and light (L05); the sort cards — balls bounce highest / most fun, paper planes farthest (L01); ramp height, plane folds, drum (L02 table); plane folds and plant water kept-same (L03 table); plant height, towel mass, ice-cube minutes (L04 table); towels, ice cubes sun + size, plane folds, flashlight batteries + distance (L05 run); the ball on grass / sand / concrete and the towel plan board (L06 quiz); every quiz, attached and practice scene (towels, cups of hot chocolate, shoes, soap, plant water amounts, kinds of ball from one height, soils, kite tails, sugar in water, ice-cube wrappings, bubble wands, a backpack on a swing, lemonade cups, toy boats with sails, freezing water, hot-chocolate lids, glues, spinning tops, sponges, cocoa cups, juice pops, sleds started higher, socks on a line, crayons, jars under cloth, bean seeds in two pots, mittens, a glass tapped, soup in a bowl or mug, coins in water, bean bags, rags on a board, warm or cold water on plants).
| Item | Move | Scene kind | Nearest earlier instance |
|---|---|---|---|
| PP-1 | pick the testable question of four | mittens in snow (new) | L01 q1–q8 (yes / no on one question) — the four-option pick is a new move |
| PP-2 | pick the question that asks for an opinion (positive stem, v2) | markers (new) | PP-1 |
| PP-3 | pick the testable question | thermos of soup (new) | PP-1 |
| PP-4 | pick the testable question | pencils (new) | PP-1 |
| PP-5 | pick the testable question | shirts drying (new; L05 q4 socks drying is a fair-test sort, a different move) | PP-1 |
| PP-6 | pick the testable question | bike tires (new) | PP-1 |
| PP-7 | pick the thing changed | block towers and a fan (new) | L02 q1–q8 (three options) — four options here |
| PP-8 | pick the thing changed | ice cubes in water of three temperatures (new scene kind: ice in water; L02 q6 wraps ice cubes) | PP-7 |
| PP-9 | pick the thing changed | balloon rubbed for three times (new) | PP-7 |
| PP-10 | pick the thing changed | paper helicopter with clips (new) | PP-7 |
| PP-11 | pick the kept-same thing of four | sponges dipped (L04 q5 measures sponges — a different move) | L03 q1–q3 (two options) |
| PP-12 | pick the kept-same thing | a bigger magnet and paper clips (v2; the heavier plane was L03's attached scene) | PP-11 |
| PP-13 | pick the kept-same thing | salt or fresh water freezing (new; L03 q3 is cold or warm water) | PP-11 |
| PP-14 | pick the measurement | soil draining (new) | L04 q2–q7 (three options) |
| PP-15 | pick the measurement | three balls pushed on a gym floor (new: different balls, one floor) | PP-14 |
| PP-16 | pick the measurement | batteries in a toy robot (new) | PP-14 |
| PP-17 | pick the one-change test of four (the stem never prints the term, v2) | pinwheels (new); option (c) now changes two things | L05 q1–q7 (yes / no on one test) — four described tests is a new move |
| PP-18 | pick the one-change test of four | tapes and posters (new; summary q3 is tapes in sun and shade — the fair-test SORT; here the fair case is side by side); option (b) now changes two things (v2) | PP-17 |
| PP-19 | recognise the term among sisters | — | L05 q9 (three options) |
| PP-20 | pick the 'We measure' line | a shoe pushed on surfaces (new) | L06 q4 (ball bounce measure) |
| PP-21 | pick the question that fits a plan | an eraser on surfaces (v2; its key had been PP-20's stem word for word) | PP-20 |
| PP-22 | pick the kept-same thing | wet or dry rag on a chalk line (PR-11's FRQ asks how to make it fair — a different move) | PP-11 |
| PP-23 | pick the thing changed | grass seed in three lights (new: light varied; L03 q5 is soil with light as the slip) | PP-7 |
| PP-24 | pick the measurement | raincoats on a doll under a tap (new) | PP-14 |
| PP-25 | pick the testable rewrite | blankets (new) | L01 q9–q10 (three options) |
| PP-26 | pick the testable rewrite | bread and mold (new) | PP-25 |
| PP-27 | read the thing changed off the question | toy boat sail width (L03 q0 is the sail size on a boat as a recall of L02 — same move, so PP-27 is marked DRAW-ONE with L03 q0; a reviewer may re-instance it) | L02 q8 (backpack on a swing) |
| PP-28 | read the thing changed off the question | seeds in a warm or cold room (new) | L02 q8 |
| PP-29 | read the thing changed off the question | rubber band thickness (new) | L02 q8 |
| PP-30 | pick the thing changed against the outcome distractor | foil or paper round a slice of bread (v2; the lemonade cups are L02 q9's) | L02 q9 |
| PP-31 | pick what should have stayed the same | glass and bowl losing water (new) | L03 q4–q6 |
| PP-32 | pick what should have stayed the same | heavy or light ball pushed harder (new: mass varied) | L03 q4–q6 |
| PP-33 | pick what should have stayed the same | a clay pot and a plastic pot watered differently (v2; the soils are L03 q5's) | L03 q5 |
| PP-34 | pick the reason a two-change result fails | paint and the number of brush strokes (v2; towels are L03's) | L03 q7 (glue reasons) |
| PP-35 | pick the measurement | bread and mold in a warm place (shares bread with PP-26; different move) | PP-14 |
| PP-36 | pick the measurement | thick sock in snow (new) | PP-14 |
| PP-37 | find the plan line that fails to name a measurable result | bean-seed plan board (figure plan_pp_seeds, new) | L06 q5 (missing line) |
| PP-38 | pick the one-change test of four | apple slices in lemon juice (new) | PP-17 |
| PP-39 | name the two things that changed | wet tennis ball and dry basketball (new) | L05 fb lines name the two things; a new ask |
| PP-40 | name the two things that changed | wool blanket in sun, cotton sheet in shade (new; v2 drops thick / thin so exactly two things change) | PP-39 |
| PP-41 | find the missing plan line | penny plan board (figure plan_pp_penny, new; the missing line is 'We measure', not L06 q5's 'We keep the same') | L06 q5 |
| PP-42 | read the table: which surface | sled on snow and ice (figure results_pp_sled, new) | L06 q6 |
| PP-43 | pick the 'We keep the same' line | block of wood on surfaces (new) | L06 q9 (juice pop) |
| PP-44 | pick the thing changed from a stem that lists the four things (medium MCQ, v2) | seeds with plant food (new) | PP-7 |
| PP-45 | decide the next step for an untestable question (hard) | markers (shares PP-2's scene; a different move) | L01 q9 |
| PP-46 | select the two kept-same things | a metal ball on a string (v2: weight is Earth's pull, never an object) | PP-11 |
| PP-47 | select the two kept-same things | toy parachute with a toy hanging under it (new) | PP-11 |
| PP-48 | pick the reason a measurement fails | backpack and walking (L02 q8's backpack is on a swing — a different scene kind) | L04 q9 (claim) |
| PP-49 | chain result → tool | matching ice cubes in salt or fresh water, which melts first (v2: the stem names the result so the stopwatch is the only tool) | L04 q8 (tool for a named measurement) |
| PP-50 | pick the reason the temperature is the result | lunchbox and juice box (new) | L04 q9 |
| PP-51* | type the term from a case beside its near miss | kinds of ball at one height (new pairing) | L05 q10 |
| PP-52* | type the term from a case | T-shirts in sun and shade (new) | L05 q10 |
| PP-53 | pick the reason a test IS fair against 'any change spoils it' | cotton ball through straws (new) | L05 q8 (claim) |
| PP-54* | type the difference between two table cells | sled table | L06 q7 (ball table, 25 centimeters) |
| PP-55 | turn the table into the plan's one-line answer | sled table | L06 q6 |
| PP-56 | pick the reason a spoiled go stays out of the table | hockey puck on a rug (new) | L06 q8 (claim) |
| PP-57 | repair a plan's 'We change' line | block on surfaces (shares PP-43's object; a different move) | L06 q9 |
⟂ (draw constraints for the test builder, v2): PP-20 ⟂ PP-21 (PP-20's stem is the shape of PP-21's key); PP-42 ⟂ PP-54 ⟂ PP-55 (one sled table; PP-55's options name PP-42's key — draw one per test); PP-43 ⟂ PP-57 (one block scene); PP-26 ⟂ PP-35 (bread); PP-2 ⟂ PP-45 (markers); PP-19 ⟂ PP-51 ⟂ PP-52 (the term by recognition hands the typed items their word — draw one per test); PP-15 ⟂ PP-32 (balls on the gym floor); PP-5 ⟂ PP-52 (shirts drying on a line). DRAW-ONE against lesson items (not a bank matter, noted for the reviewer): PP-27 / L03 q0 (PP-30 and PP-33 were re-scened in v2 and no longer share a lesson scene).