ETX5900 S2 2026 quiz bank: adaptation for the ETX1100/ETX5900 eight-week intensive

The source bank has 11 original weeks, usually 20 questions each. S3 Q7 below means Question 7 in source Week 3, not intensive Week 3. It contains marked answers and should remain an instructor source file. Keep means the statistical task is suitable; all retained questions still need answers and solution cues removed before student release. Adapt means retain the skill but replace data, context, or wording. Exclude means omit from the core eight-week quizzes.

The original Elecmart, Hytex, PlanFinan, Health care, Hospital Quality, and Data for quiz 11 workbooks were not found in this project. The embedded tables and image references in the Markdown file do supply some questions without those workbooks. Current intensive workbooks can supply replacement data, but the old numerical answers must be recalculated.

This is an instructor selection review, not a released quiz or answer key. Follow AGENTS.md, course_info.R, and the current weekly slides and workbooks when adapting items. The eight weekly quizzes are individual and together carry 15% of the ETX1100 assessment. Weeks 1–4 provide the scope for the Week 4 written assignment; regression and business modelling begin later. The outcome tags below use the ETX1100 outcomes as planning labels. Confirm the official ETX5900 outcome wording before using them as its formal assessment map.

Current Week 5 decision: teach and assess the chi-square test of independence as a categorical-relationship application of LO4. Quiz 5 may assess hypotheses, expected counts, cell contributions, degrees of freedom, CHISQ.TEST, conditions, and a contextual decision. Quiz 4 and the written assignment remain limited to the material taught through Week 4. The source-bank S6 label is its old week number, not the intensive week.

Excel boundary: the Analysis ToolPak is the only analysis add-in in the current course. Ordinary worksheet formulas and probability functions, PivotTables, charts, filters, and manual scenario comparisons remain appropriate. Where an item requires a fitted regression, align its output and instructions with Data > Data Analysis > Regression. Do not require Solver, Goal Seek, What-If Data Tables, or programming-based optimisation. A ToolPak regression question alone assesses LO5; an LO6 item must also test how the fitted relationship enters a business decision model.

Recommended question map

Intensive week	Source questions and decision	Skill to assess	Outcome	Adaptation
1: Data story	Keep S1 Q1–2	Distinguish population, sample, parameter, and statistic	LO1; LO3 foundation	Refresh the context; check that the stated population is unambiguous.
1	Adapt S1 Q3–10, Q19–20	Mean, standard deviation, range, quartiles, coefficient of variation, and interpretation	LO1	Replace missing `Health care` data with `week1/week1-data.xlsx` or a small self-contained table. Choose a few complementary items rather than eight near duplicates.
1	Adapt S1 Q11–18	Filters, pivot tables, counts, conditional averages, relative frequencies	LO1	Replace missing `Elecmart` data; ask for the category and value where useful. State the denominator and unit of observation.
1	Keep/adapt S2 Q1–2, Q4–6	Interpret skewness, symmetry, and a boxplot	LO1	Preserve Q4’s figure as an ordinary image. Reword Q5–6 to avoid implying equal whisker lengths in every symmetric sample.
1	Adapt S2 Q7–10	Select a histogram and compare numerical variability	LO1	Rebuild charts and answers from the current Week 1 workbook. Do not use S2 Q3 as written: no option correctly identifies what can be read directly from a boxplot.
1	Adapt S2 Q11–20	Compare categorical proportions and use grouped pivot-table counts	LO1; LO4 foundation	Replace `Elecmart`/`Health care` data. Reduce repeated filter-only questions and ask for a one-sentence interpretation.
2: Probability and distributions	Keep S3 Q1–8, Q10; optional Q9	Mutually exclusive versus independent events, complements, union, conditional probability, and total probability; Q9 adds Bayes’ rule	LO2	Convert image-only probability notation to editable math; verify all numerical keys. Use Q9 only if inverse conditional probability is taught.
2	Adapt S3 Q11–14	Conditional probabilities from a contingency table; comparison of categories	LO2; LO4 foundation	Inline hospital tables may be reusable. Make the requested numerator and denominator explicit.
2	Adapt S3 Q15–20	Pivot-table joint and conditional percentages	LO2; LO4 foundation	Replace missing `Hytex`/`Elecmart` data with `week2/week2-categorical-data.xlsx`; regenerate keys.
2	Keep/adapt S4 Q3	Properties of the normal distribution	LO2	Move this item ahead of the sampling-distribution material; add a business probability or percentile calculation using `week2/week2-distributions-data.xlsx`.
3: Sampling and estimation	Keep S4 Q1–2, Q4–10	Sampling distribution, central limit theorem, standard error, and effect of sample size	LO3	Use a few varied items. Qualify the blanket `n ≥ 30` wording in S4 Q5; sufficient sample size depends on the population shape.
3	Adapt S4 Q11–20	Probabilities and percentiles for sample means	LO3	Repair ambiguous normal notation such as `N(16, 22)` and the incomplete answer in S4 Q13. Prefer Excel probability functions over old table-lookup instructions. S4 Q16 is present but has no Q16 label.
3	Keep S5 Q1–17	Choose t versus normal methods; calculate and interpret intervals for means and proportions	LO3	Keep one or two each for method choice, mean CI, proportion CI, and the effect of confidence level. Replace answer-only items with contextual interpretation.
3	Adapt S5 Q18–20	Confidence intervals for a proportion using spreadsheet data	LO3	Recalculate with `week3/week3-proportions-data.xlsx`; original `Elecmart` data are missing.
4: Testing claims	Keep/adapt S7 Q1–7, Q9–18	State hypotheses for a mean; one- and two-sided t tests; p-values; Type I/II errors	LO3	Convert image-only hypotheses to text. Keep a small set of distinct decision and interpretation tasks. Verify all stated test statistics before reuse.
4	Adapt S7 Q8, Q20	Mean tests using Excel data	LO3	Replace `Elecmart` with `week4/week4-testing-data.xlsx` and regenerate answers. Include a distinct one-sample, one-sided decision-threshold item with a calculated t statistic and `T.DIST` p-value to support A1 without copying its dataset or answer.
4	Keep/adapt S8 Q1–12, Q14–20	One-sample proportion tests, p-values, and errors	LO3	Select different alternatives and conclusions. Replace dated/medical contexts where they distract from the statistics. Verify normal-approximation conditions and keys.
4	Exclude as written S8 Q13	Intended Type I error for a proportion	LO3	`π ≤ 32` and `π > 32` are impossible for a proportion. Rewrite with values in `[0,1]` if needed.
5: Chi-square and simple regression	Adapt S6 Q1–2	Frame independence and association hypotheses for two categorical variables	LO4	Replace the TV/storybook context with the current `week5/week5-chi-square-data.xlsx` cases or a new self-contained table. Remove marked answers.
5	Adapt S6 Q3–6	Calculate an expected count and one cell’s contribution to \chi^2	LO4	The embedded defect-rate table is self-contained, but use fresh data for the released quiz. Show row, column and grand totals; recalculate the key from the final table.
5	Adapt S6 Q7–8, Q13–14	Calculate the test statistic or interpret a p-value and conclusion	LO4	Prefer `CHISQ.TEST` and a contextual conclusion over several critical-value lookups. The two source spokesperson tables differ; do not mix their keys. State the significance level and avoid causal wording.
5	Adapt S6 Q9–12	Recognise degrees of freedom and the right-tail decision rule	LO4	Select at most one distinct item if needed. Rewrite around an Excel p-value; check expected counts and independent observations.
5	Keep/adapt S9 Q4–7, Q10–15	Correlation, slope, least-squares prediction, and limits of association	LO4; LO5	Most use embedded Excel-output images; extract legible figures or replace them with a small output table. S9 Q12–13 are self-contained concept checks.
5	Adapt S9 Q1–3	Build/interpret simple regression from data	LO4; LO5	Replace missing `Hytex`, `PlanFinan`, and `Elecmart` files with `week5/week5-regression-data.xlsx`; regenerate ToolPak Regression output and keys.
5	Keep/adapt S10 Q3–4, Q7–8, Q13–18	Interpret R-squared, slope tests, and least-squares fit	LO4; LO5	Retain the best distinct items and readable output. Check one-sided versus two-sided p-value wording.
6: Multiple regression	Keep/adapt S9 Q8–9, Q20; S10 Q5–6, Q9–10, Q19–20	Interpret a predictor while holding others constant; model fit and coefficient tests	LO4; LO5	Use the embedded outputs only after checking legibility and numerical keys; or rebuild from `week6/week6-realestate-data.xlsx`.
6	Adapt S9 Q16–17; S10 Q1–2, Q11–12	Fit a multiple model, predict, interpret R-squared and partial coefficients	LO5	Original `PlanFinan` workbook is missing. Rebuild ToolPak Regression output with the current Week 6 workbook and validate all answer values.
6	Keep/adapt S9 Q18	Predict from a supplied multiple-regression equation	LO5	This is self-contained and a good case where an indicator/other predictor takes zero.
6	Exclude as written S9 Q19	Intended multiple-regression prediction	LO5	With zero personal-training sessions, predicted personal-training expenditure is positive. Change the outcome to total fitness-centre expenditure or write a coherent new case.
7: ToolPak-supported business decision	No original LO6 decision-model questions; adapt a source regression-output skill only as a component of a new case	Read and check ToolPak price–demand coefficients; identify data-derived and business inputs, listed-price decision, formulas, objective and rules; apply whole-unit and capacity limits; compare predicted profit and one-at-a-time sensitivity	LO6; LO5 application	Write new items tied to `week7/week7-modelling-data.xlsx`. Ask for the best listed feasible price, a nearby alternative, and a limit on the recommendation. Do not claim a global optimum.
8: Integrated case and revision	Reuse selected revised questions from intensive Weeks 1–7 and write linked items	Choose a method, read Excel output, explain uncertainty and prediction, and justify a business decision	LO1–LO6	Use `week8/week8-case-data.xlsx` for the supermarket case and a short Week 7 pricing-model revisit. Include LO6; the old bank has no complete decision-model item.

Small starter set for each weekly quiz

These are selection priorities, not a final student quiz. They favour distinct skills over repeated arithmetic; items marked adapt need new data or a clean output figure.

Intensive quiz	First questions to select	What the set covers
1	S1 Q1–2; S2 Q1–2, Q4; adapt S1 Q13	Population versus sample, skewness, visual reading, pivot-table count (LO1; LO3 foundation)
2	S3 Q1–2, Q7, Q10; adapt S3 Q11; S4 Q3	Event rules, conditional and total probability, contingency table, normal distribution (LO2; LO4 foundation)
3	S4 Q2, Q6, Q9; S5 Q1, Q10, Q15	CLT, standard error, choosing a mean CI, mean and proportion intervals (LO3)
4	S7 Q1, Q4, Q12, Q14; S8 Q1, Q10, Q14	Hypotheses, mean/proportion tests, Type I/II errors, p-value decisions (LO3)
5	Adapt S6 Q1 and Q3 or Q5; write a `CHISQ.TEST` decision item from the Week 5 categorical workbook; S9 Q12–13; adapt S9 Q4 or Q10	Independence hypotheses, expected counts, p-value and interpretation, then correlation, slope or prediction (LO4–LO5). Keep both categorical testing and simple regression in the final set.
6	S9 Q18; adapt S9 Q20 and S10 Q5, Q9, Q19–20	Multiple-model prediction, partial coefficients, R-squared, coefficient tests (LO4–LO5)
7	Write new linked items from the Week 7 pilot and decision-model sheets	ToolPak coefficient interpretation, formula audit, capacity rule, best listed feasible price, baseline comparison, one-at-a-time sensitivity and limitations (LO6; LO5 application)
8	Write an integrated case; draw revised items from 1–7	Method selection and communication across LO1–LO6, including a ToolPak-supported pricing decision

Material outside the core sequence

S6 Q15–20: These refer to Hytex, Elecmart and PlanFinan workbooks absent from this project. The skill is now in scope for Quiz 5, but these numerical items need new data and independently calculated keys. Avoid letting the source bank’s 20 chi-square items crowd out simple regression in Quiz 5.
S11 Q1–20: The old Week 11 is seasonal indices, trend forecasts, MAD/MSFE, and spreadsheet forecasting. Exclude from core quizzes because full time-series forecasting is not in the intensive content spine or named 2026 outcomes. A new Week 1 chart-choice question could use a time-series plot, but the original questions mostly ask forecasting methods. Its final Q9 appears after Q20 and should be moved if this bank is ever republished.
Legacy Excel decision tools: The old ETF business-modelling source uses Solver, Goal Seek and What-If Data Tables; these are outside the implemented Week 7 plan. Do not import their output-reading or procedure questions into Quiz 7 or Quiz 8. Use ToolPak regression and ordinary Excel formulas to compare only the listed feasible scenarios.

Before any questions go to students

Remove answer leakage. This bank has [x] selections, bold numerical answers, and worked conclusions in stems. Draft clean student versions and keep a separate answer key.
Replace missing-data items deliberately. No old named quiz workbooks were found here. Current weekly files are useful replacements, but their variables and answers are not interchangeable with Elecmart or PlanFinan. Recompute every key from the final released dataset.
Repair known defects. S2 Q3 has no defensible choice; S4 Q13 has an incomplete answer and S4 Q11–14 have ambiguous distribution notation; S4 Q16 lacks its number; S7 Q19 states a test statistic of 0.077 although (795−775)/(75/√30) ≈ 1.461; S8 Q13 uses an impossible proportion; S9 Q19 has an incoherent outcome.
Use questions to assess interpretation. Prefer one calculation plus a denominator, assumption, or business conclusion to several near-identical number-entry items. Keep a balanced mix of conceptual choice, Excel use, and contextual reasoning in each weekly quiz.
Audit figures. Many output screenshots are embedded as base64 images in the source Markdown. Extract and relabel only the figures selected for reuse; replace tiny formula images with editable equations and provide accessible text or tables.
Check week and tool alignment. Week 4 items must stay within description, probability, sampling, intervals and mean/proportion tests. Introduce chi-square and simple regression in Week 5, multiple regression in Week 6, and the decision model in Week 7. For chi-square, name the observed and expected count ranges, exclude totals, check expected counts, and verify that CHISQ.TEST agrees with the summed contributions and right-tail probability. For any required Excel regression, state ToolPak X and Y ranges clearly and check the resulting coefficients and units.
Audit LO6 directly. A Quiz 7 item should make students distinguish ToolPak estimates from cost and capacity assumptions, apply the non-negative whole-unit capacity rule, calculate revenue, cost and profit, and choose among listed prices. Add a sensitivity or limitation item that resets each trial to the same baseline. No source-bank question by itself covers this workflow.