📚 Education
AP Statistics Free Response Item Writer: Original Multi Part Questions Aligned to Course Units and Skills, E P I Part Scoring, Four Point Holistic Conversion, and Sample Student Responses
Write original AP Statistics style free response practice questions for a class or tutoring set: a realistic context and data, parts aligned to a course unit and skill category, a scoring guide that grades each part Essentially correct, Partially correct, or Incorrect, a four point holistic conversion, and scored sample responses.
0Reviews
Prompt
Act as an AP Statistics teacher and experienced free response reader who writes practice items for classroom use and knows how readers apply E, P, and I scoring to each part before converting to a 0 to 4 score. Inputs: - Course unit and topics to assess, using the unit names from the current College Board Course and Exam Description: [UnitAndTopics] - Skill category to emphasize (Selecting Statistical Methods, Data Analysis, Using Probability and Simulation, Statistical Argumentation): [SkillFocus] - A context the students will find realistic (school, sports, local business, health): [ContextIdea] - Item type: standard multi part question or longer investigative task: [ItemType] - Data you want used, or let the item writer create plausible numbers: [DataProvided] - How many parts and how much time students get: [PartsAndTime] - Output format: [Format] Generate: 1. The item: a short stem with the context and data, then lettered parts that build on each other. Each part must ask for one clear task (identify, calculate, interpret, justify) and state what students must show. 2. A worked solution with every number computed and rounded consistently, so a teacher can check it quickly. 3. A scoring guide for each part: what is Essentially correct, what earns Partially correct, and common responses that are Incorrect. Include the components readers look for, such as naming the procedure, checking conditions with numbers, correct mechanics, and a conclusion in context with a linkage to the evidence. 4. A holistic conversion table for this item that turns the combination of part scores into 4, 3, 2, 1, or 0, written for this item's number of parts. 5. Three sample student responses (strong, middle, weak) with part scores, the holistic score, and one sentence explaining each decision. 6. Alignment notes: unit, topic, and skill each part targets, and which common errors the item is designed to surface. 7. A quick variant: the same structure with a new context and new numbers for a retake. Constraints: - Write original items only. Do not reproduce or lightly edit released College Board questions or scoring guidelines. - Do not claim the item or rubric is official. Confirm exam format, timing, and question counts in the current Course and Exam Description. - Numbers must be internally consistent and conditions must be checkable from the data given. No em dashes.
Instructions
Replace every [bracket] with your details before running. Works on ChatGPT, Claude, and Gemini.
Generated Output
This image was generated using the prompt above.
Examples
Example Input
UnitAndTopics: Unit 6 Inference for Categorical Data: Proportions; one sample z interval for p, conditions, interpreting the interval SkillFocus: Statistical Argumentation ContextIdea: student council survey about adding a later lunch period ItemType: standard multi part question DataProvided: random sample of 150 students from a school of 2,400; 57 favor the later lunch PartsAndTime: 3 parts, about 13 minutes Format: item, solution, scoring guide, holistic table, samples
Example Output
Item The student council at a high school with 2,400 students selected a random sample of 150 students and asked whether they favor adding a later lunch period. Of those sampled, 57 said yes. (a) Identify the appropriate procedure for estimating the proportion of all students at the school who favor a later lunch, and verify the conditions for it. (b) Construct and interpret a 95 percent confidence interval for that proportion. (c) The council president claims that fewer than half of all students favor a later lunch. Does your interval support this claim? Explain. Worked solution (a) One sample z interval for a population proportion. Random: random sample stated. 10 percent: 150 is less than 10 percent of 2,400 (240). Large counts: 57 yes and 93 no, both at least 10. (b) p hat = 57/150 = 0.38. SE = sqrt(0.38 x 0.62 / 150) = 0.0396. ME = 1.96 x 0.0396 = 0.0777. Interval (0.302, 0.458). We are 95 percent confident that the interval from 0.302 to 0.458 captures the true proportion of all students at this school who favor a later lunch. (c) Yes. Every value in the interval is below 0.5, so the interval gives convincing evidence that fewer than half favor the later lunch. Scoring guide Part (a) E: names a one sample z interval for p (by name or formula) and checks all three conditions with numbers (150 < 240; 57 and 93 at least 10). P: correct procedure with one condition missing or not tied to numbers. I: wrong procedure (t interval, two sample) or no conditions. Part (b) E: correct interval with work shown and an interpretation that names the parameter in context (all students at this school) and uses the word confident or an equivalent. P: correct interval with a weak interpretation (probability language, or about the sample), or interpretation correct with a minor arithmetic error. I: no interval, or interval and interpretation both wrong. Part (c) E: says yes and links the decision to all values below 0.5. P: says yes with vague reasoning (the center is below 0.5). I: says no, or no reasoning. Holistic conversion 4 Complete: EEE 3 Substantial: two E and one P 2 Developing: two E and one I, or one E and two P, or three P 1 Minimal: one E and one P, or one E alone, or two P 0: anything less Sample responses Strong: all conditions with numbers, interval (0.30, 0.46), 'we are 95% confident the true proportion of students at this school...', (c) cites upper bound below 0.5. Scores E, E, E = 4. Complete and linked throughout. Middle: names procedure, checks random and large counts only; correct interval; says '95% chance the true proportion is in the interval'; (c) correct. Scores P, P, E = 2. Missing 10 percent check and probability language each cost a part. Weak: uses a t interval; computes 0.38 plus or minus 0.05 with no work; (c) 'yes because 0.38 is less than 0.5.' Scores I, I, P = 0. Point estimate reasoning is not interval reasoning. Alignment (a) Unit 6, conditions, Selecting Statistical Methods. (b) Unit 6, interval and interpretation, Statistical Argumentation. (c) Unit 6, using an interval to judge a claim, Statistical Argumentation.