ETF1100 Cheatsheet · Business Statistics

Monash University · A4 Cheatsheet for the final exam. Monash Semester 2 final assessment period, 2–18 November 2026 (check your personal exam timetable). Full ETF1100 study guide →

A4 landscape · one note sheet · both sides · 6 columns per sideETF1100 · Monash University · S2 2026
ETF1100
Business Statistics
Monash University · Level 1 · Semester 2 2026
ETF1100 cheatsheet QR code – open the full ETF1100 study guide
Full guide + ask Sia about ETF1100
SIDE 1/2  METHODS · RULES · DECISIONSONE A4 NOTE SHEETFinal exam · 50% · two-hour supervised on-campus eExam
Tutorial 10%Quiz 5%Case 20%Group 15%Final exam 50%
PERMITTED MATERIALS: “One double sided A4 sheet of notes, on paper.”ETF1100 2026 S2 Week 12 Exam Revision page (links to s2-2026 resources) (MD_Week_12_Exam_Revision.md)

1 · Start With the RowW1

one rowvariabletypemethod
Guide redraw: define the observational unit before choosing a calculation.
Field roleUse / trap
Identifierduplicate checks; do not average
Nominallabels, no order → count/proportion
Ordinalordered levels; spacing not guaranteed
Discretecounts; numerical arithmetic
Continuousmeasurements; units essential
Missingnever silently zero

Audit sequence: unit → variable role → missing/impossible values → duplicate identifiers → target population. Preserve the untouched source; put derived fields in new columns and reconcile row counts after filters.

Exam bite A numeric code such as postcode can be nominal. Storage format does not determine statistical type.

2 · Choose the DisplayW2

What is the question?categorybarsdistributionhistogramtimeline2 numericscatterunequal groups → % + denominator
Guide redraw: purpose and variable type decide the chart, not Excel defaults.
QuestionReport
How many?count: workload / volume
What share?relative frequency = count ÷ relevant total
Unequal exposure?normalised rate, with exposure unit
Audittitle, axes, unit, order, missing groups, selected range

Bars start from a common baseline. Histograms use adjacent numerical bins. Lines preserve genuine order, especially time. A truncated axis may exaggerate a length difference; 3-D geometry obscures magnitude. Normalisation repairs the denominator, not uncontrolled product mix, hours or case complexity.

2b · Spreadsheet AuditW1–2

StageActionEvidence retained
landsave raw tab; one header rowuntouched source

3 · Categorical ProbabilityW3

OnlineStoreTotalNewReturn362460184260inside = joint; edge = denominator
Guide redraw: shade the conditioning total before dividing.

Core rulesP(Aᶜ)=1−P(A)
P(A∪B)=P(A)+P(B)−P(A∩B)
P(A|B)=P(A∩B)/P(B)
independence: P(A|B)=P(A), hence P(A∩B)=P(A)P(B)

AskedDenominatorCheck
joint A and Bgrand totalcell ÷ all
marginal Agrand totalA total ÷ all
A given BB totalB-conditioned shares sum 1

Association: compare outcome proportions across groups using the same conditioning logic. A large raw cell may merely reflect a large group. Report a difference between percentages as percentage points; relative percent change divides that gap by a named baseline.

Exam bite Reversing the words after “given” reverses the denominator. Write the fraction before the decimal. Association in recorded data is not causation.

3b · Denominator ChecklistW3

Before submitQuestion
Unitticket, visit, order or customer?
Eligibilitywho could experience “success”?
Missingexcluded, shown, or separate?
Scopesample description or population inference?

Repeated visits are not automatically independent customers. A pooled difference can also hide a lurking group pattern; investigate before choosing the preferred comparison.

3c · Probability Drill GridW3

From one two-way tableRequired working
jointinterior cell ÷ grand total; name both events
reversed conditionalssame cell, two different group totals

4 · Centre, Spread, ShapeW4

minQ1medianQ3maxIQR = Q3 − Q1; inspect tail/outlier
Guide redraw: the five-number summary shows centre, middle-half spread and tails.
Shape / purposeCentreSpread
symmetric, no strong outliermeanstandard deviation
skewed / influential outliermedianIQR
workload / total planningmean may still mattershow valid extremes

Standardise an observationz = (x − mean) / standard deviation
z > 0 above mean; z < 0 below; unit-free

Units: mean, median, range, IQR and standard deviation use the original unit; variance uses squared units; z has no unit. Standardising does not make skewed data normal. Use a normal model only after checking shape and whether the model is suitable.

μ−σμμ+σ
Guide redraw: interpret distance and sign before seeking a normal area.

Outlier rule Verify the record and business origin. Correct a confirmed error; retain and explain a valid exception; never delete merely because the answer changes.

4b · Distribution WorkflowW4

StepQuestionExam output
1 shapesymmetric, skewed, clusters, gaps?name visible form
2 centretypical case or resource average?median or mean, with unit
3 spreadmiddle half or mean-distance?IQR or SD, with unit
4 unusualerror, valid exception, second process?investigate; sensitivity
5 comparesame unit, period, inclusion rule?difference plus context

Resistance test: predict which summaries change if the largest value doubles. Mean, range and SD react; median and IQR are usually more resistant. Resistance is not permission to ignore operational extremes.

5 · Sampling & IntervalsW5

populationsampleestimateinterval
Guide redraw: the parameter is fixed; the estimate varies across random samples.

Interval architectureestimate ± critical value × SE
margin = critical value × SE
width = 2 × margin

Mean · large-sample practiceSE(x̄)=s/√n
95% CI: x̄ ± 1.96(s/√n)

Proportion · usual intervalp̂=x/n
SE(p̂)=√[p̂(1−p̂)/n]
CI: p̂ ± critical × SE(p̂)

ChangeInterval effect
n increasesSE falls; narrower (4×n ≈ ½SE)
confidence increasescritical value rises; wider
variability increasesSE rises; wider
selection biasnot repaired by a narrow interval

Interpretation: the confidence procedure captures the true population parameter at its stated long-run rate under assumptions. For one realised interval, state plausible population parameter values, not the percentage of individual observations and not a probability attached to a fixed parameter.

Endpoint checksum: midpoint = estimate; half-width = margin; lower < upper; both carry the parameter’s unit. Confirm whether the entry asks for an endpoint, margin, width or full interval.

Exam bite Random sampling supports generalisation. Independence, selection, nonresponse and measurement quality remain separate from arithmetic precision.

5b · CI Decision ChecksW5

fixed parametersamples vary → intervals vary
Guide redraw: confidence is a repeated-sampling property of the procedure.
Interval locationDecision reading
entirely above targetplausible means all exceed target
crosses targetvalues on both sides remain plausible

6 · Tests: Claim → DecisionW6–7

null reference modelobservedp-value
Guide redraw: the p-value is the tail at least as extreme under H₀.
  1. State population H₀ and H₁; direction comes from the claim before data.
  2. Check sampling/independence and approximation conditions.
  3. Compute SE and standardised statistic; use the matching tail(s).
  4. If p ≤ α, reject H₀; otherwise do not reject H₀.
  5. Conclude about the represented population; report effect size and practical importance.

Mean test · large-sample practiceH₀: μ=μ₀
z=(x̄−μ₀)/(s/√n)

One-proportion z testH₀: p=p₀
SE₀=√[p₀(1−p₀)/n]
z=(p̂−p₀)/SE₀

Claim wordingAlternative / tail
exceeds / above> · right
below / reduced< · left
differs / changed≠ · two-sided

Test SE uses p₀; usual proportion-interval SE uses p̂. For the normal approximation, check expected success and failure counts using the proportion appropriate to the method.

DecisionH₀ trueH₀ false
rejectType I errorcorrect
do not rejectcorrectType II error

Language A small p-value is not P(H₀ true). “Do not reject” is not “prove equal”. Statistical evidence is not automatically operational importance.

6b · Test Output DecoderW6–7

OutputMeaningCommon mix-up
estimateobserved sample effectpopulation truth
SEsampling fluctuation scalesample SD
znull distance in SE unitseffect magnitude
p-valuenull-tail probabilityP(H₀ true)
αpre-set Type I thresholdmove after seeing p
CIplausible parameter rangeindividual-value range

7 · Regression & ModelsW7–10

observedresidualpredictor
Guide redraw: inspect direction, form, strength and unusual points before r.

Simple regressionŷ=b₀+b₁x
residual e=y−ŷ
R² = proportion of sample y-variation explained

OutputInterpretation / check
b₁predicted y change per +1 x; name both units
b₀predicted y at x=0; may lack meaning
e > 0observed lies above fitted value
residual plotcurve, funnel or influential case warns
predictionavoid extrapolation outside observed x range

Correlation: unit-free linear association from −1 to 1; sensitive to outliers and blind to some curved relationships. Regression/correlation alone do not establish causation.

Multiple regression

ŷ=b₀+b₁x₁+⋯+bₖxₖ

Each slope compares predicted outcomes for a one-unit predictor change while holding the other included predictors fixed. For a 0/1 indicator, coefficient = predicted difference from the omitted reference category at the same included controls. Ordinary R² cannot fall when predictors are added; adjusted R² penalises weak additions. Neither validates causation or new-data prediction.

Optimisation

decisionvariablesconstraintsobjectiveverifysolution
Guide redraw: formulate in words and units before opening Solver.

Define variables, objective and every capacity/demand/policy constraint; add non-negativity and integrality where relevant. Then substitute the returned values into every constraint, compute slack and mark binding limits. Feasible satisfies constraints; optimal is best among feasible choices. Stress-test contribution, capacity and demand. A sensitive mix changes sharply after a plausible input change.

asksia.ai/cheatsheet/monash-etf1100 · Full ETF1100 study guide →
Forward this to your tutorial group · Shared by ________
AskSiav2 · 2026-10-25
Monash Semester 2 final assessment period, 2–18 November 2026
(check your personal exam timetable).
Your exam allows one A4 sheet of notes, both sides. Check on the LMS whether printed notes are accepted; if they must be handwritten, copy it.
ETF1100
Business Statistics · Worked Drills
Monash University · calculations to the conclusion
ETF1100 cheatsheet QR code – open the full ETF1100 study guide
Full guide + ask Sia about ETF1100
SIDE 2/2  WORKED EXAMPLES · TRAPS · FAST RECALLONE A4 NOTE SHEETKeep full precision → final requested format

8 · Counts vs RatesW2 worked

Given: metropolitan 48 late of 1,600; regional 30 of 600.

Metro rate = 48/1600 = 0.030 = 3.0%
Regional rate = 30/600 = 0.050 = 5.0%
Observed difference = 2.0 percentage points

late countsMetro 48Regional 30late rates3.0%5.0%
Count answers workload; rate answers relative performance.

Conclusion: metropolitan operations have more late cases to resolve, but regional operations have the worse late-delivery rate. Show both when the decision needs workload and performance.

TrapRepair
compare 48 with 30 onlydivide by exposure totals
say “2% higher”say 2 percentage points
claim causename uncontrolled mix / complexity

9 · Conditional RatesW3 worked

Given: chat resolves 72/90; email 77/110.

P(resolve|chat)=72/90=80.0%
P(resolve|email)=77/110=70.0%
gap = 10.0 percentage points

chat: 72 resolved / 90 = 80%email: 77 resolved / 110 = 70%
Each channel total supplies its own conditional denominator.

Conclusion: channel and first-contact resolution are associated in the recorded sample. Do not claim changing channel causes improvement: ticket complexity and staffing may differ.

9b · File Classification DrillW1 worked

File: one row per order; OrderID, Channel, LoyaltyTier, Items, Spend, OrderDate, DeliveryDate.

FieldRoleDefensible action
OrderIDidentifiercheck duplicates; never average
Channelnominalcount / conditional proportion
LoyaltyTierordinalpreserve order; spacing unknown

10 · Extreme ValueW4 worked

Data: 8, 9, 10, 10, 11, 12, 38 minutes.

sum = 98
mean = 98/7 = 14.0 min
median = 4th ordered value = 10 min
range = 38−8 = 30 min · IQR = 12−8 = 4 min

6 clustered cases38right tail: mean > median
Guide-derived drill: the valid extreme matters operationally even when median is typical.

Conclusion: right-skewed; 38 pulls the mean up, so median better describes a typical case. Investigate 38 rather than deleting it. If 38 is verified as a data-entry error for 13, corrected sum = 73, mean = 10.43, median remains 10.

Automatic-entry checkExpected
mean / medianoriginal unit
variancesquared unit
standard deviationnon-negative, original unit
z-scoresign agrees with above/below mean

11 · Mean Confidence IntervalW5 worked

Given: n=64 orders, x̄=31.5 h, s=8.0 h; use 1.96.

SE=8/√64=1.00 h
margin=1.96×1.00=1.96 h
95% CI=31.5±1.96=[29.54, 33.46] h

29.5431.533.46
Midpoint = estimate; equal half-widths = 1.96 hours.

Interpret for the population mean fulfilment time, assuming random sampling and approximation conditions—not individual orders.

11b · Interval RepairsW5 FAQ

Faulty claimRepair
“95% of orders lie from 29.54 to 33.46 h.”the interval concerns the population mean, not individual orders
“There is 95% probability μ is inside.”the procedure has 95% long-run coverage under assumptions
“Narrow means unbiased.”width covers sampling fluctuation, not selection or measurement bias
“Lower endpoint is 33.46.”order endpoints and recover midpoint 31.5

12 · Mean TestW6 worked

Given: n=49 calls, x̄=6.8 min, s=2.1 min; test at α=0.05 whether μ exceeds 6.

H₀: μ=6 · H₁: μ>6
SE=2.1/√49=0.30 min
z=(6.8−6.0)/0.30=2.67
right-tail p≈0.0038 < 0.05

H₀ distributionz=2.67p=.0038
Right tail is fixed by “exceeds”, not by the observed sign.

Decision: reject H₀. Conclusion: the sample supports a population mean wait above 6 minutes for the represented call process, under the sampling conditions.

WrongCorrect
“H₀ has probability .0038”result this extreme is rare under H₀
“accept H₁ because z is big”compare p with pre-set α
switch to two tailsuse claim-set direction

13 · Proportion TestW7 worked

Given: 94 conversions among 200 eligible visits; test p>0.40 at 5%.

H₀:p=.40 · H₁:p>.40
p̂=94/200=.470
SE₀=√(.40×.60/200)=.03464
z=(.47−.40)/.03464=2.02
p≈.0217 → reject H₀

p₀=.40p̂=.47+7 pp
Observed lift = 7 percentage points; the test SE still uses p₀.

Conclusion: evidence that the represented population conversion proportion exceeds 40%. Keep the visit-level denominator distinct from a customer-level claim if people can repeat.

13b · Test Completion GridW6–7 drills

Marking stepMean exampleProportion example
parameterpopulation mean wait μpopulation conversion p
benchmark6 minutes0.40
directionexceeds → rightexceeds → right
SE inputs/√np₀ under H₀
effectx̄−μ₀ = 0.8 minp̂−p₀ = 0.07 = 7 pp
conclusionpopulation mean claimpopulation proportion claim

14 · Simple RegressionW7–8 worked

Model: Timê = 18.4 + 0.72 Distance, minutes and kilometres. At x=50 km, observed y=59 min.

ŷ=18.4+0.72(50)=54.4 min
e=y−ŷ=59−54.4=+4.6 min

59 observed54.4 fittede=+4.650 km
Positive residual: the trip took longer than the fitted line predicted.

Slope sentence: within the observed range, each additional kilometre is associated with 0.72 predicted minute more travel time, on average. The intercept is meaningful only if zero kilometres is relevant. With R²=0.64, 64% of sample time variation is explained by this fitted distance model—not 64% prediction accuracy or causality.

Residual patternConcern
curvelinear form misses structure
funnelspread changes with fitted value
isolated leverage pointverify; compare fits transparently

15 · Indicator CoefficientW9 worked

Saleŝ=72−1.8Price+9.5Rating+14Premium
Price=12, Rating=4, Premium=1:
72−21.6+38+14=102.4 thousand
Premium=0: 88.4 thousand

standard 88.4premium 102.4+14.0
Same price and rating; Premium changes 0→1 against standard reference.

Premium sites are associated with $14 thousand higher predicted weekly sales at the same included price and rating. Do not say “causes”. Omitted site characteristics remain uncontrolled.

15b · Regression Output AuditW8–9 FAQ

Before interpretingWrite
outcomename and unit of y
predictorname, unit, meaningful increment
coefficientdirection + predicted change
indicatorcoded-1 group vs reference-0 group
controlsonly predictors actually included
scopeobserved range; sample / represented population
diagnosticresidual form, spread, influence

16 · Product-Mix ModelW10 worked

Variables: Standard kits x; Premium kits y.

maximise Z=30x+45y
labour: 2x+3y≤120
materials: x+2y≤70
demand: y≤30
x,y≥0 [integer if whole kits required]

Verify candidate x=30, y=20:

CheckSubstituteStatus
labour2(30)+3(20)=120binding
materials30+2(20)=70binding
premium demand20≤30slack 10
objective30(30)+45(20)$1,800
candidate cornerfeasiblexy
Guide redraw: feasibility comes before comparing objective values at corners.

The candidate is feasible and uses all labour and materials. It is optimal only after comparison with all feasible corners or a verified Solver run. Solver success validates the entered mathematics, not a missing business restriction.

WordsSymbolMeaning
at most≤capacity / upper limit
at least≥minimum requirement
exactly=fixed balance
whole unitsintegerdo not round after solve

Exam bite Recompute the objective and each constraint. Report quantities, objective value, bottleneck, spare capacity and the assumption to monitor.

17 · Method ChooserW1–10

What is required?describesampleestimateparametertestbenchmarkpredict /optimisetable/plotctr + spreadconfidenceintervalH₀/H₁, z,p-valueregression /SolverFinish: unit + scope + limit
New exam route: identify the task before selecting an output or formula.

18 · Fast Recall Matrixall weeks

SignalUseDo not say / do
category comparisoncount + within-group %ignore denominator
numerical distributionshape + centre + spreadmean by default
parameter rangeestimate ± critical×SE% of individuals
claim vs benchmarkH₀/H₁, statistic, pP(H₀ true)
binary population sharep̂; interval or test SEmix p̂ and p₀
two numerical variablesscatter → r / regressioncausation from fit
several predictorscontrolled coefficientcontrol omitted factors
best choice under limitsobjective + constraintstrust Solver blindly

Precision routine

read unit+ formatchoosecalculate fullprecisionround +enter
Guide redraw: automatic marking rewards exact reading and final-format discipline.
  1. Underline requested quantity: count, probability, percentage, percentage points, SE, statistic, p-value, endpoint, coefficient or objective.
  2. Write denominator / benchmark / reference category and units before calculating.
  3. Keep full precision through intermediate steps. Estimate sign and rough magnitude.
  4. Convert decimal ↔ percentage only when requested; round once at the final line.
  5. Check sign, unit, plausible range and whether the answer is sample description or population conclusion.

19 · Last-Minute TrapsFAQ + drills

If you see…Think…
postcode / IDlabel, not measurement
unequal group sizesrate plus count
given / conditionalrestrict denominator
skew / outliermedian + IQR; investigate
95% confidenceprocedure coverage; parameter
p>αinsufficient evidence; not equality
significantalso report effect size
residualy−ŷ, not ŷ−y
indicatorcoded 1 vs reference 0
feasiblenot necessarily optimal

Final sweep Method fits variable type? Assumptions named? Denominator correct? Tail chosen before result? Units and reference group stated? No causal overclaim? Requested precision and entry convention followed?

asksia.ai/cheatsheet/monash-etf1100 · Full ETF1100 study guide →
Forward this to your tutorial group · Shared by ________
AskSiav2 · 2026-10-25
Monash Semester 2 final assessment period, 2–18 November 2026
(check your personal exam timetable).
Your exam allows one A4 sheet of notes, both sides. Check on the LMS whether printed notes are accepted; if they must be handwritten, copy it.

Want one for YOUR exact syllabus?

Sia is your free desktop study agent. Drop your Monash University ETF1100 slides — Sia builds a sheet tailored to YOUR exam. Better than this library because it knows YOUR materials.

↓ Download Sia · Free