- Step 1OpenCounting, Factorials, Permutations & CombinationsCounting outcomes starts with one question: does changing the order create a genuinely different result?
- Step 2OpenProbability Language, Sample Spaces & Set RepresentationA probability question is first a language question: what is the experiment, what is the event, and what is the whole universe of outcomes?
- Step 3Full accessAddition & Multiplication RulesAdd when you want one event or another; multiply when you want one event and then another. The words decide the operation.
- Step 4Full accessConditional Probability & ReplacementGiven narrows the universe. Once you know B has happened, only the outcomes inside B are still possible.
- Step 5Full accessExpected Value & Fair GamesExpected value is the long-run average per attempt: each outcome weighted by how often it happens.
- Step 6Full accessBernoulli Trials & Repeated SuccessesCheck the four Bernoulli conditions first; drawing without replacement fails them. Then P(exactly r successes) = number of orders × probability of one order, where a success is whatever the question counts.
- Step 7Full accessFirst and kth SuccessName the success as the result being waited for, write the sequence the event describes, then multiply. For the kth success on trial n, fix trial n and count orders only in the trials before it.
- Step 8Full accessNormal Distribution & z-ScoresA z-score says how many standard deviations a value is from the mean. The table gives left areas, and right tails, negative z-scores and between-values areas are all built from them.
- Step 9Full accessPopulation, Sample, Variables & Statistical QuestionsName the population the question is about and the sample actually observed, classify each variable as counted, measured or a label, and state any sample result as an estimate for the population.
- Step 10Full accessSampling Methods, Bias & RepresentativenessJudge a sample by how it was chosen: name the method by what is picked, say who could be left out or not reply and which way that pushes the result, and give each group its share in a stratified sample.
- Step 11Full accessStudy Design, Randomisation, Control & EthicsOnly a designed experiment with randomised allocation supports a cause-and-effect claim; an observational study can only report association.
- Step 12Full accessDisplaying & Reading Univariate DataThe display must match the data type and the question, and every display should be read for shape, centre and spread.
- Step 13Full accessCentre, Spread, Quartiles, Percentiles & OutliersA single number for centre is only half the story; spread says how much the data vary around it.
- Step 14Full accessBivariate Data, Scatterplots & CorrelationA scatterplot shows whether two numerical variables move together; the correlation coefficient measures how closely, and nothing more.
- Step 15Full accessSampling Variability, Sampling Distributions & the Empirical RuleDifferent samples give different results. The pattern of those results is itself predictable, and that pattern is what inference uses.
- Step 16Full accessMargin of Error & Confidence IntervalsA confidence interval reports an estimate together with how far off it could reasonably be.
- Step 17Full accessHypothesis Tests for a Population ProportionState what would be true if nothing had changed, then ask whether the sample is too far from that to be explained by sampling variation.
- Step 18Full accessHigher Level Mean z-Tests & p-ValuesThe test statistic says how many standard errors the sample mean is from the claimed mean; the p-value says how often chance alone would go that far.