Inferential Statistics in 97 Flashcards: Regression, the Normal Curve and Hypothesis Testing
Deck EP8443 covers regression, the normal distribution and hypothesis testing in 97 cards. What it teaches well, and the six cards you should add first.
Studying. But it's also a video game. 🎮
Flashcards, quizzes and games that make you actually remember.
Posted by
FLIP IT TeamRelated reading
Probability Study Guide: 102 Flashcards on Rules, Distributions and Paradoxes
Learn probability rules and distributions: sample space, Bayes theorem, binomial, Poisson, the normal curve, z-scores and classic paradoxes in 102 flashcards.
How to Master Descriptive Statistics: 108 Flashcards on Mean, Median, Mode and Standard Deviation
Learn descriptive statistics with 108 flashcards: mean, median, mode, variance, standard deviation, z-scores, IQR, skewness and Pearson correlation.
Deck EP8443 — Inferential Statistics: Correlation, Regression & Hypothesis Testing holds 97 cards in three clearly separate blocks: 1–38 on correlation and least squares, 39–48 on the normal distribution, 49–97 on hypothesis testing and estimation theory. Study it block by block and it works; read it front to back and the seams will trip you.
Study this deck: Inferential Statistics: Correlation, Regression & Hypothesis Testing
1 / 10Correlation in statistics
Tap the card for the answer · swipe right if you knew it, left if not
What is on the cards
Coverage is wide: ordinary least squares, residual sum of squares, nonlinear regression, coefficient of determination, heteroscedasticity, confounding variables, extrapolation vs interpolation, the 68-95-99.7 rule, z-scores, the central limit theorem, Type I and Type II errors, statistical power, p-values, confidence intervals, one- and two-tailed tests, sampling distributions, standard error, Bessel's correction, Student's t-distribution, chi-squared, skewness, kurtosis, Q-Q plots, the law of large numbers, maximum likelihood estimation, the method of moments, Cochran's theorem, the probit function.
| Front | Back (abridged) |
|---|---|
| P-value | The probability of observing a test statistic at least as extreme as the one obtained, assuming the null hypothesis is true |
| Type II error | Failing to reject the null hypothesis when it is actually false (false negative); probability denoted β |
| Statistical power | The probability of correctly rejecting a false null hypothesis, equal to 1 − β |
| Standard error | The standard deviation of a sampling distribution. For sample mean: SE = σ/√n |
| Sample size vs standard error | SE ∝ 1/√n. To halve the SE, you need to quadruple the sample size |
| Coefficient of determination (R²) | How well the regression model fits the data; the proportion of variance explained |
| Heteroscedasticity | When the variability of residuals is not constant across levels of the independent variable |
| Bessel's correction | Using (n−1) instead of n in sample variance, to make it an unbiased estimator |
The p-value card states the definition correctly, which is rarer in flashcard decks than it should be, and card 72 adds that a statistically significant result can still carry a trivial effect size in a large sample.
The hole in the correlation half
The word Pearson appears nowhere in the deck. Correlation gets a prose definition on card 1, Spearman's rank coefficient gets card 18 — and the coefficient every course actually computes has no card. Nothing tells you r runs from −1 to +1, or that r² is the R² of card 29.
The regression block has the same shape. Slope and intercept are defined in words (cards 34–35), but the equation ŷ = b₀ + b₁x is never written down and no card carries a worked calculation. Across 97 cards there are four formulas: the normal PDF, z = (X − μ)/σ, SE = σ/√n, and s² = Σ(xᵢ − x̄)²/(n−1). No t-statistic, no ANOVA, no test carried out end to end, and degrees of freedom is used inside the chi-squared back without ever getting a front of its own.
Tip: write six cards before reviewing a single one — Pearson's r range and formula, r vs R², when to switch to Spearman, the regression equation, degrees of freedom, the t-statistic. All guaranteed exam items, all skipped.
Tip: turn the testing block into a decision table. Cards 49–58 and 69–71 teach what α, β, power and tails are; an exam asks which test to reach for. One row per scenario — one mean with σ known, one mean with σ unknown, two means, a variance, a goodness of fit — beats twelve isolated definitions.
Duplicates at the seams
Card 6 (residual in regression analysis) and card 89 (residual in statistics) are the same card. Card 19 defines descriptive vs inferential statistics; cards 59 and 60 then define each separately. Cards 63 and 88 both explain standard error. Bessel's correction appears three times (66, 67, 97), mean = median = mode three times (41, 47, 76). Keep the version carrying the formula, suspend the rest — about ten cards off your daily load.
Related reading
Ready to learn faster? 🚀
Join FLIP IT for free and turn studying into a game.