From describing data to inference — and knowing what a p-value does not mean.
Statistics is the most widely used and most widely misused mathematics in the world. Almost every adult will be shown statistical claims by people who want something from them, and almost none has been taught to interrogate one.
This programme runs from describing data through probability to genuine inference. The emphasis throughout is on what a result licenses you to say — which is a narrower thing than most people assume, and knowing exactly how narrow is the whole skill.
What you need before starting: Light prior knowledge, not concept-zero. Every programme is built to be entered without prior exposure to the specific subject.
Every rule is derived rather than issued. The programme is designed to be entered without prior exposure to the subject — light prior knowledge, not concept-zero.
TopicsData types (qualitative/quantitative), population vs sample, sampling methods, bias, experimental design
TopicsFrequency distributions, histograms, bar charts, stem-and-leaf plots, box plots
TopicsMean, median, mode, range, IQR, standard deviation, variance, outliers and resistant measures
TopicsZ-scores, data transformation, distribution shape, scatter plots, Pearson correlation
TopicsAssociation between variables, positive/negative correlation, correlation vs causation, interpreting values
TopicsData analysis case studies, choosing appropriate statistics, summarising and graphing data
TopicsSample spaces, counting principle, permutations and combinations, basic probability rules
TopicsConditional probability, independence, multiplication rule, addition rule, Bayes' Theorem
TopicsDiscrete vs continuous, PMF, PDF, expected value and variance of random variables
TopicsBernoulli trials, binomial distribution, Poisson distribution, when to use each
TopicsNormal distribution, standard normal, Z-scores, probability tables
TopicsNormal approximation to binomial, continuity correction, uniform, exponential, t, chi-square, F distributions
TopicsDistribution of sample mean, CLT, standard error, sampling distribution of proportion
TopicsPoint estimators, bias and efficiency, CI for means (σ known and unknown)
TopicsCI for proportions, CI for variances, sample size determination, margin of error
TopicsNull/alternative hypotheses, Type I/II errors, significance level, p-values, one vs two-tailed tests
TopicsOne-sample t-test, critical value and p-value methods, assumptions, power analysis
TopicsOne-sample test for proportions, chi-square for variance, goodness-of-fit, test of independence
TopicsTwo-sample t-test, independent samples, equal vs unequal variances, practical significance
TopicsPaired t-tests, comparing two proportions, F-test for variances, effect size
TopicsOne-way ANOVA, F-test, between/within variation, sum of squares decomposition, assumptions
TopicsInterpreting ANOVA, post-hoc comparisons, Tukey HSD, Bonferroni correction, two-way ANOVA
TopicsNon-parametric alternatives, Mann-Whitney, Kruskal-Wallis, bootstrap, permutation tests
TopicsTest selection flowchart, assumptions review, power analysis, effect size, practical considerations
TopicsComprehensive testing review, real data, mixed problems, connecting tests to research questions
TopicsScatter plots, least squares method, regression equation, slope and intercept interpretation, residuals
TopicsStandard error, hypothesis tests for slope, CIs for slope, R-squared, ANOVA for regression
TopicsResidual plots, checking assumptions, outliers, Cook's distance, leverage
TopicsPrediction intervals, CIs for mean response, multiple regression, multicollinearity
TopicsComprehensive review, real-world analysis, variable transformations, categorical variables
Every session includes a problem set with full written solutions. Progress tracking and individual feedback throughout, and a conversational programme textbook written for the course.
A group opens when four students are ready.