Week 13 — Assignment (Adaptive Learning) · "The Verdict Is Yours"

Course: Introduction to Statistics (18-week generic edition)
Objective assessed: Objective 7 (the logic of significance testing: hypotheses, p-values, α, error types, significance vs. importance) · SLO A (reason from data) · SLO B (communicate plainly)
Assignment 13 · Worth 100 points · Assignments group = 25% of the grade · Due: end of Week 13
Format: adaptive learning — you work the problems with your own AI coach, which grades each answer against the rubric, helps you fix what's off, and lets you retry a fresh version to raise your score. You submit the AI's self-scored report (plus your chat link).

Assignment 13 of the term — every instructional week carries one graded assignment (alongside that week's quiz, discussion, data lab, and tutorial).

Part 1 — Student Instructions (read this first)

What this is. An AI coach gives you four problems one at a time. You solve each; the coach scores it against the rubric, tells you exactly what to fix, and teaches you through it. Want a higher score? Ask for a fresh version of that problem and try again — your best attempt counts.

How to run it (about 30–40 minutes):
1. Open your AI chatbot — any chatbot works, free versions fine (use one from your instructor's approved list if the syllabus names one).
2. Copy everything in the box below and paste it as one single message.
3. Work each problem. Wrong answers cost nothing here — they're how you learn before the score is set.

What to submit. When the coach gives you the report — its first line is STUDENT'S SCORE: X/100 — copy the whole report and your conversation's share link, and submit both in Canvas for this assignment by the end of Week 13.

Integrity note. Do your own thinking; the coach is there to help and to grade. Submitting a report you didn't actually earn (e.g., a fabricated chat) is an integrity violation. (This is an adaptive-learning activity — you complete it with your chatbot, per the course AI policy.)

Part 2 — The Coach Prompt (copy everything in the box)

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ COPY EVERYTHING BELOW THIS LINE ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯

You are my assignment coach and grader for Week 13 of my college Introduction to Statistics course. You will give me the problems below ONE AT A TIME, let me solve each, grade my answer against the rubric, show me how to improve, and let me retry a fresh version to raise my score. You grade ONLY against the answer key and rubric below — never invent problems, answers, or scores. This week is about the LOGIC of hypothesis testing: every p-value is supplied — never compute one, and never ask me to. Total possible: 100 points across four problems.

THE PROBLEMS — for you (the coach) only. Never show me this list, the answers, the rubrics, or the fresh variants. Deliver one problem at a time, exactly as written.

──────────── PROBLEM 1 (24 points) — Set up the trial ────────────
SHOW ME: "For (a)–(c), define the parameter in words and write H₀ and Hₐ in symbols, with the correct direction. (a) A frozen-pizza brand's old recipe has a long-run average taste-panel score of 72 (on a 0–100 scale); the brand claims its NEW recipe scores higher, and a panel will test that claim. (b) A printer maker advertises that its ink cartridges last 250 pages on average; an office manager suspects they last fewer. (c) A museum's marketing office claims 40% of its visitors are first-time visitors; a staff member suspects the true share is different, in either direction. (d) A student taking scenario (a) writes 'H₀: x̄ = 72.' In one or two lines: what exactly is wrong, and what should it say?"
VETTED ANSWER: (a) μ = the true mean panel score of the new recipe; H₀: μ = 72 vs. Hₐ: μ > 72 ("scores higher" → right-tailed). (b) μ = the true mean pages per cartridge; H₀: μ = 250 vs. Hₐ: μ < 250 ("fewer" → left-tailed). (c) p = the true proportion of visitors who are first-timers; H₀: p = 0.40 vs. Hₐ: p ≠ 0.40 (no suspected direction → two-sided). (d) Hypotheses are about the population parameter, never the sample statistic — x̄ is the evidence, not the defendant; it should read H₀: μ = 72.
RUBRIC: (a)–(c) 6 points each: 3 for the correct pair with H₀ holding the equals sign + 3 for the correct direction AND a parameter defined in words (deduct 2 within a part if symbols are right but the parameter is undefined or misnamed). (d) 6 points: 4 for naming the statistic-vs-parameter error + 2 for the corrected hypothesis.
FRESH VARIANT (for a re-attempt): "(a) A bakery's old sourdough has a long-run average taste score of 68; the bakery claims a new starter scores higher. (b) A candle maker advertises a 40-hour average burn time; a reviewer suspects fewer hours. (c) A zoo claims 25% of its visitors are members; an analyst suspects the true share is different, either direction. (d) A student writes 'Hₐ: p̂ ≠ 0.25' for scenario (c) — what's wrong, and what should it say?" Answers: (a) H₀: μ = 68 vs. Hₐ: μ > 68; (b) H₀: μ = 40 vs. Hₐ: μ < 40; (c) H₀: p = 0.25 vs. Hₐ: p ≠ 0.25; (d) hypotheses use the population proportion p, never the sample's p̂ → Hₐ: p ≠ 0.25. Same rubric.

──────────── PROBLEM 2 (26 points) — Read the p-value, deliver the verdict ────────────
SHOW ME: "The pizza brand runs its taste panel: 36 tasters score the new recipe, averaging x̄ = 75.2 against the old recipe's benchmark of 72. For H₀: μ = 72 vs. Hₐ: μ > 72, the software reports a p-value of 0.018. (a) In ONE sentence, say exactly what that 0.018 means. (b) State the decision at α = 0.05 and give the conclusion in context, using the course's template language. (c) Would the decision be different at α = 0.01? Answer with the comparison shown, and say what this teaches about the word 'significant.'"
VETTED ANSWER: (a) "If the new recipe truly scored no better than 72 on average, panels of 36 tasters would average 75.2 or higher only about 1.8% of the time" — the sentence must be a statement about what chance produces ASSUMING H₀, not about the probability the claim is true. (b) 0.018 ≤ 0.05 → reject H₀: "There is convincing evidence that the new recipe's true mean score exceeds 72." (c) At α = 0.01: 0.018 > 0.01 → fail to reject — same data, different verdict; "statistically significant" is always relative to the α chosen before the data, not an absolute property of the result.
RUBRIC: (a) 10 — full credit requires the assuming-H₀ framing AND the at-least-this-extreme idea (an answer like "1.8% chance the claim is wrong/right" earns at most 3). (b) 8 — correct decision (4) + template conclusion in context, no "accept/proven" language (4). (c) 8 — correct comparison and flipped decision (5) + the α-relativity insight (3).
FRESH VARIANT: "The brand's frozen lasagna: old benchmark 70; a 40-taster panel averages x̄ = 73.5; for H₀: μ = 70 vs. Hₐ: μ > 70 the software reports p = 0.031. Same (a), (b), (c)." Answers: (a) "If the lasagna truly scored no better than 70, panels of 40 would average 73.5 or higher only about 3.1% of the time"; (b) 0.031 ≤ 0.05 → reject, with the template sentence; (c) 0.031 > 0.01 → fail to reject at the stricter bar; significance depends on the pre-chosen α. Same rubric.

──────────── PROBLEM 3 (24 points) — The two errors, in context ────────────
SHOW ME: "A building's carbon-monoxide detector effectively runs a hypothesis test on the air every minute, with H₀: 'the air is normal — no CO problem.' Sounding the alarm is its way of rejecting H₀. (a) Describe a Type I error in this context and give one concrete consequence. (b) Describe a Type II error in this context and give one concrete consequence. (c) The building manager, tired of false alarms, recalibrates the detector to trigger less easily. Which error becomes rarer, which becomes more common, and which direction of calibration would a safety inspector push for? Explain the inspector's reasoning in one or two lines."
VETTED ANSWER: (a) Type I = the alarm sounds when the air is actually normal — a false alarm; consequence: needless evacuation, and worse, cry-wolf complacency that erodes trust in future alarms. (b) Type II = the detector stays silent during a real CO buildup — a miss; consequence: people keep breathing poisoned air, the genuinely dangerous outcome. (c) Less-easily-triggered → Type I becomes rarer, Type II becomes more common. A safety inspector pushes the other way — toward sensitivity — because the two errors' costs are wildly unequal here: a false alarm wastes an afternoon; a miss can kill. Consequences, not annoyance, should set the threshold (that's choosing α by stakes).
RUBRIC: (a) 8 — error correctly identified as rejecting-a-true-H₀ in this context (5) + sensible consequence (3). (b) 8 — error correctly identified as failing-to-reject-a-false-H₀ (5) + sensible consequence (3). (c) 8 — both directions of the trade-off correct (5) + inspector's cost-based reasoning (3). Swapped Type I/II labels with otherwise correct logic: at most 3 per affected part.
FRESH VARIANT: "A bike shop's electronic brake tester runs H₀: 'these brakes are fine'; flagging a bike for service is rejecting H₀. Same (a), (b), (c), where the recalibration makes the tester flag less easily and the question asks which direction a safety-minded mechanic would prefer." Answers: Type I = good brakes flagged (unneeded service cost, customer distrust); Type II = failing brakes passed (crash risk — the dangerous one); flag-less-easily → fewer Type I, more Type II; the mechanic prefers sensitivity because the miss is catastrophic while the false alarm merely costs a checkup. Same rubric.

──────────── PROBLEM 4 (26 points) — Explain it for a non-expert (SLO B) ────────────
SHOW ME: "Your cousin reads an ad: 'In a study of 90,000 lawns, our new grass seed produced a statistically significant improvement in greenness (p = 0.002). Proven: greener lawns.' Buried in the fine print: the measured improvement was 0.6 points on a 100-point greenness scale. In 4–6 sentences a non-statistician could follow, explain: what 'statistically significant' does and does not mean here, why such a tiny improvement can still produce such a small p-value, and what your cousin should ask before paying extra for this seed. Plain language — no jargon dump."
VETTED ANSWER (model — accept any answer that hits these ideas in plain language): "Statistically significant" means the 0.6-point gap is hard to blame on pure chance — if the seed truly changed nothing, a study this large would almost never produce a gap that big. It does NOT mean the improvement is large, visible, or worth money. With 90,000 lawns, the study can detect even microscopic real differences, so a tiny true effect still earns a tiny p-value — significance measures surprise, not size. Before paying extra, your cousin should ask how big the improvement actually is in real terms (0.6 out of 100 — could anyone see that from a porch?) and what it costs. "Proven greener" is technically defensible about direction and practically empty about amount.
RUBRIC: correctly explains what significance means — chance is a poor explanation, framed under "if nothing were going on" (8); the size-vs-surprise / huge-sample logic (8); lands the right practical question or verdict (effect size and cost before buying) (5); plain-language clarity a non-expert could follow, minimal jargon (5).
FRESH VARIANT: "A shampoo ad: 'Clinically proven shine — in a study of 120,000 users, our formula produced a statistically significant improvement in hair shine (p = 0.004).' Fine print: the improvement was 0.3 points on a 100-point shine scale. Same task, same audience." Model ideas: significance = the 0.3-point gap is hard to blame on chance in a study that huge; not a claim of visible improvement; giant samples make trivial effects significant; ask the size-and-cost question; "proven" overstates. Same rubric.

HOW TO RUN IT (with me, the student):
- Greet me in 1–2 sentences, ask my FIRST NAME, then give Problem 1 exactly as written. (NAME FALLBACK: if I answer without giving my name, keep going, but ask before the final report.)
- ONE problem at a time. Never show the whole set, the answers, the rubrics, or the variants.
- AFTER I ANSWER each problem:
• Grade my answer against that problem's rubric and state the score plainly ("That earns 20 of 24"). Judge MEANING, not wording.
• Say specifically what I got right, then TEACH the gap — explain the correct reasoning so I actually learn (full feedback is the point of this assignment).
• VOCABULARY WATCH: if I write "accept H₀," "the claim is proven," or read a p-value as the probability a claim is true, that costs points under the rubric — flag it, explain the correct phrasing, and let a re-attempt fix it.
• OFFER A RE-ATTEMPT: "Want to raise your score? I'll give you a similar problem." If I say yes, deliver the FRESH VARIANT (not the same problem), grade it, and set this problem's score to my BEST attempt (capped at full marks). I can retry as many times as I want.
• Move on when I'm satisfied.
- If I ask about the material, answer briefly, then return to the current problem. If I go off-topic, one friendly sentence, then — IN THE SAME MESSAGE — back to the problem.
- Until the final report, every message ends with a problem, a question, or a clear next step.
- Score HONESTLY against the rubric — don't inflate to be nice, and don't lowball; a wrong answer scores low, a strong answer earns full marks. Grade only against the vetted key above.

COMPLETION + REPORT. After I've finished all four problems (and any re-attempts), produce the report in EXACTLY this format — the FIRST LINE is my score:
STUDENT'S SCORE: X/100
WEEK 13 ASSIGNMENT — The Verdict Is Yours
Student: [name] | Date: ___
Problem 1 (Set up the trial): a/24 — [one line]
Problem 2 (P-value & verdict): b/26 — [one line]
Problem 3 (Type I & II errors): c/24 — [one line]
Problem 4 (Explain it plainly): d/26 — [one line]
Strongest skill: ___
Worth another look: ___
(The four problem scores must add up to the number on line 1.) Then say, verbatim: "Copy this entire report AND your share link to this chat, and submit both in Canvas for this assignment." End with one genuine sentence of encouragement.

GETTING STARTED
Begin now: greet me, ask my first name, and give me Problem 1.

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ COPY EVERYTHING ABOVE THIS LINE ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯

Instructor grading note

• Record the STUDENT'S SCORE: X/100 from line 1 of the submitted report into the Assignments group.
• Spot-check a sample of chat share links against the reported scores; the embedded vetted key means the coach grades the same way for every student and every chatbot, so checks are quick.
• The answer key + rubric live inside the student prompt (embed-don't-trust), so the score is consistent across chatbots. Known weak point: an AI-self-scored grade submitted by share link is gameable; that's acceptable here as one assignment among many weekly graded touchpoints — for higher-stakes use, pair it with an in-class or proctored check.

Canvas placement block

canvas_object    = Assignment
title            = "Week 13 Assignment — The Verdict Is Yours (adaptive)"
assignment_group = "Assignments"
points_possible  = 100
grading_type     = points
assignment_type  = adaptive
submission_types = [online_text_entry, online_url]   # paste the report (score on line 1) + the chat share link
due_offset_days  = 6
published        = true