Statistics and probability cheat sheet

BOOKLET given in the examLEARN you must remember it · Each block links to its lesson.

Statistics and probability

IB MATHS AA · HL
pastpaperprep.com/revision-notes

4.1Sampling, data quality and outliers

IQR=Q3−Q1\mathrm{IQR} = Q_{3} - Q_{1}BOOKLET
lower fence=Q1−1.5×IQR,upper fence=Q3+1.5×IQR\text{lower fence} = Q_{1} - 1.5 \times \mathrm{IQR}, \qquad \text{upper fence} = Q_{3} + 1.5 \times \mathrm{IQR}LEARN
number from a stratum=size of stratumsize of population×sample size\text{number from a stratum} = \frac{\text{size of stratum}}{\text{size of population}} \times \text{sample size}LEARN
The population is the whole group; a sample is the part you collect data from. Counted data is discrete; measured data is continuous.LEARN
A sample is biased if the way it is chosen favours some members. Say who is favoured and why.LEARN
Leave missing values out, never record them as 00. Correct a recording error only if you know the true value.LEARN
outlier: x<Q1−1.5×IQRorx>Q3+1.5×IQR\text{outlier: } x < Q_{1} - 1.5 \times \mathrm{IQR} \quad \text{or} \quad x > Q_{3} + 1.5 \times \mathrm{IQR}LEARN
A value equal to a fence is not an outlier. Keep valid outliers; remove only errors.LEARN
Simple random: every sample equally likely. Systematic: every kkth, k=Nnk = \frac{N}{n}, random start. Stratified: random within groups, in proportion. Quota: fixed numbers, collector chooses. Convenience: easiest to reach.LEARN

4.2Data presentation

median at n2,Q1 at n4,Q3 at 3n4\text{median at } \tfrac{n}{2}, \qquad Q_{1} \text{ at } \tfrac{n}{4}, \qquad Q_{3} \text{ at } \tfrac{3n}{4}LEARN
IQR=Q3−Q1\mathrm{IQR} = Q_{3} - Q_{1}BOOKLET
Class intervals are inequalities with no gaps, such as 20≤t<2520 \le t < 25 and 25≤t<3025 \le t < 30. A boundary value goes in the class whose inequality it satisfies.LEARN
In a histogram with equal class widths, bar height is frequency and the bars touch.LEARN
Plot cumulative frequency at each upper class boundary, starting from 00 at the first lower boundary. It counts values at or below, so for “more than” subtract from nn.LEARN
median at n2,Q1 at n4,Q3 at 3n4,kth percentile at k100×n\text{median at } \tfrac{n}{2}, \qquad Q_{1} \text{ at } \tfrac{n}{4}, \qquad Q_{3} \text{ at } \tfrac{3n}{4}, \qquad k\text{th percentile at } \tfrac{k}{100} \times nLEARN
Box plot: box from Q1Q_{1} to Q3Q_{3}, line at the median, whiskers to the most extreme values that are not outliers, a cross for each outlier.LEARN
Compare medians for centre and IQRs for spread, in context. A symmetric box and whiskers means the data may be normal.LEARN

4.3Central tendency and dispersion

xˉ=∑i=1kfixin,where n=∑i=1kfi\bar{x} = \frac{\sum_{i=1}^{k} f_{i}x_{i}}{n}, \quad \text{where } n = \sum_{i=1}^{k} f_{i}BOOKLET
IQR=Q3−Q1\mathrm{IQR} = Q_{3} - Q_{1}BOOKLET
Median: the n+12\frac{n+1}{2}th value in order; for even nn, the mean of the middle two. Mode: the most frequent value.LEARN
Grouped data: estimate the mean with mid-interval values a+b2\frac{a+b}{2}. The modal class has the highest frequency.LEARN
Quartiles: medians of the lower and upper halves, median left out when nn is odd. Other methods may differ slightly.LEARN
Standard deviation: GDC only, using σx\sigma_{x}, not sxs_{x}. Variance =σ2= \sigma^{2}: square the unrounded value.LEARN
x→ax+bx \to ax + b with a>0a > 0: mean →axˉ+b\to a\bar{x} + b, σ→aσ\sigma \to a\sigma, variance →a2σ2\to a^{2}\sigma^{2}.LEARN

4.4Correlation and regression (y on x)

−1≤r≤1-1 \le r \le 1LEARN
Describe correlation by direction (positive, negative, zero) and strength (strong, weak, none).LEARN
rr near ±1\pm 1: strong linear correlation; near 00: none, though there may be a curve. ∣r∣\lvert r\rvert above the critical value: significant.LEARN
A line of best fit by eye, and the regression line, pass through the mean point (xˉ,yˉ)(\bar{x}, \bar{y}).LEARN
Correlation does not imply causation: look for a third variable.LEARN
In y=ax+by = ax + b, aa is the change in yy per 11 increase in xx; bb is yy when x=0x = 0.LEARN
Use the yy on xx line only to predict yy, and only inside the range of the data.LEARN

4.5Probability foundations and expected value

P(A)=n(A)n(U)\mathrm{P}(A) = \frac{n(A)}{n(U)}BOOKLET
relative frequency=number of times the event happensnumber of trials\text{relative frequency} = \frac{\text{number of times the event happens}}{\text{number of trials}}LEARN
P(A)+P(A′)=1\mathrm{P}(A) + \mathrm{P}(A') = 1BOOKLET
expected number of occurrences of A=n×P(A),where n is the number of trials\text{expected number of occurrences of } A = n \times \mathrm{P}(A), \quad \text{where } n \text{ is the number of trials}LEARN
A trial has outcomes; the sample space UU lists them all; an event is a set of outcomes. For two stages, list pairs or draw a table.LEARN
Counting works only for equally likely outcomes. Every probability is between 00 and 11.LEARN
Relative frequency =times the event happensnumber of trials= \frac{\text{times the event happens}}{\text{number of trials}} estimates a probability. More trials give a better estimate.LEARN
For "at least one", use 1−P(none)1 - \mathrm{P}(\text{none}).LEARN
Expected number of occurrences =n×P(A)= n \times \mathrm{P}(A) for nn trials. It need not be a whole number.LEARN

4.6Probability diagrams, rules and independence

P(A∪B)=P(A)+P(B)−P(A∩B)\mathrm{P}(A \cup B) = \mathrm{P}(A) + \mathrm{P}(B) - \mathrm{P}(A \cap B)BOOKLET
P(A∪B)=P(A)+P(B)\mathrm{P}(A \cup B) = \mathrm{P}(A) + \mathrm{P}(B)BOOKLET
P(A∣B)=P(A∩B)P(B)\mathrm{P}(A \mid B) = \dfrac{\mathrm{P}(A \cap B)}{\mathrm{P}(B)}BOOKLET
P(A∩B)=P(B) P(A∣B)\mathrm{P}(A \cap B) = \mathrm{P}(B)\,\mathrm{P}(A \mid B)LEARN
P(A∩B)=P(A) P(B)\mathrm{P}(A \cap B) = \mathrm{P}(A)\,\mathrm{P}(B)BOOKLET
Fill a Venn diagram from the middle out: the overlap, then the rest of each circle, then the outside. "Or" includes both: A∪BA \cup B means AA, BB or both.LEARN
Mutually exclusive events can't both happen: P(A∩B)=0\mathrm{P}(A \cap B) = 0.LEARN
Tree diagrams: multiply along a path, add across paths. Without replacement, the second-stage fractions change.LEARN

4.7Discrete random variables and expected value

∑P(X=x)=1\sum \mathrm{P}(X = x) = 1LEARN
E(X)=∑x P(X=x)\mathrm{E}(X) = \sum x\,\mathrm{P}(X = x)BOOKLET
fair game: E(X)=0,where X is the player’s gain\text{fair game: } \mathrm{E}(X) = 0, \quad \text{where } X \text{ is the player's gain}LEARN
A discrete random variable XX has a list of values xx. Its probability distribution gives P(X=x)\mathrm{P}(X = x) for each, as a table or a formula.LEARN
∑P(X=x)=1,0≤P(X=x)≤1\sum \mathrm{P}(X = x) = 1, \quad 0 \le \mathrm{P}(X = x) \le 1LEARN
To find kk, set the sum of the probabilities equal to 11. Reject any root that gives a probability below 00 or above 11.LEARN
E(X)\mathrm{E}(X) is the long-run average. It need not be a value XX can take. Over nn trials, expect a total of n×E(X)n \times \mathrm{E}(X).LEARN
If XX is a player's gain (prize minus cost), the game is fair when E(X)=0\mathrm{E}(X) = 0.LEARN

4.8Binomial distribution

P(X=r)=nCr pr(1−p)n−r,r=0,1,…,n\mathrm{P}(X = r) = {}^{n}C_{r}\, p^{r}(1-p)^{n-r}, \quad r = 0, 1, \ldots, nLEARN
at most kP(X≤k)fewer than kP(X≤k−1)at least kP(X≥k)=1−P(X≤k−1)more than kP(X>k)=1−P(X≤k)a to b inclusiveP(a≤X≤b)=P(X≤b)−P(X≤a−1)\begin{array}{ll} \text{at most } k & \mathrm{P}(X \le k) \\ \text{fewer than } k & \mathrm{P}(X \le k - 1) \\ \text{at least } k & \mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k - 1) \\ \text{more than } k & \mathrm{P}(X > k) = 1 - \mathrm{P}(X \le k) \\ a \text{ to } b \text{ inclusive} & \mathrm{P}(a \le X \le b) = \mathrm{P}(X \le b) - \mathrm{P}(X \le a - 1) \end{array}LEARN
E(X)=np\mathrm{E}(X) = npBOOKLET
Var(X)=np(1−p)\mathrm{Var}(X) = np(1-p)BOOKLET
X∼B(n,p)X \sim \mathrm{B}(n, p) counts successes in a fixed number nn of independent trials, each with two outcomes and the same success probability pp.LEARN
On your GDC, binomial pdf gives P(X=r)\mathrm{P}(X = r) and binomial cdf gives P(X≤k)\mathrm{P}(X \le k).LEARN
P(X≥k)=1−P(X≤k−1)\mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k - 1) and P(X>k)=1−P(X≤k)\mathrm{P}(X > k) = 1 - \mathrm{P}(X \le k).LEARN
npnp is the expected number of successes. The standard deviation is np(1−p)\sqrt{np(1-p)}.LEARN

4.9Normal distribution

P(X<μ)=P(X>μ)=0.5,P(X<μ−d)=P(X>μ+d)\mathrm{P}(X < \mu) = \mathrm{P}(X > \mu) = 0.5, \qquad \mathrm{P}(X < \mu - d) = \mathrm{P}(X > \mu + d)LEARN
P(μ−σ<X<μ+σ)≈0.68P(μ−2σ<X<μ+2σ)≈0.95P(μ−3σ<X<μ+3σ)≈0.997\begin{aligned} \mathrm{P}(\mu - \sigma < X < \mu + \sigma) &\approx 0.68 \\ \mathrm{P}(\mu - 2\sigma < X < \mu + 2\sigma) &\approx 0.95 \\ \mathrm{P}(\mu - 3\sigma < X < \mu + 3\sigma) &\approx 0.997 \end{aligned}LEARN
P(X>k)=p  ⟺  P(X<k)=1−p\mathrm{P}(X > k) = p \iff \mathrm{P}(X < k) = 1 - pLEARN
X∼N(μ,σ2)X \sim \mathrm{N}(\mu, \sigma^{2}): a bell-shaped curve, symmetric about μ\mu, with total area 11. Mean, median and mode all equal μ\mu. The second number is the variance.LEARN
For a continuous variable P(X=a)=0\mathrm{P}(X = a) = 0, so << and ≤\le give the same probability.LEARN
Normal cdf takes lower bound, upper bound, μ\mu and σ\sigma. Use −1E99-1\text{E}99 for less than and 1E991\text{E}99 for more than.LEARN
For a middle region of probability pp, each tail holds 1−p2\frac{1 - p}{2}.LEARN
Given that: P(A∣B)=P(A∩B)P(B)\mathrm{P}(A \mid B) = \frac{\mathrm{P}(A \cap B)}{\mathrm{P}(B)}, where A∩BA \cap B is the overlap of the two intervals.LEARN

4.10Regression line of x on y and prediction

The line of xx on yy, x=cy+dx = cy + d, minimises squared horizontal distances. Use it to predict xx from yy.LEARN
On your GDC, put the yy-values in the X list and the xx-values in the Y list.LEARN
A line of xx on yy can't reliably predict yy from xx: use the line of yy on xx. Never rearrange one line to predict the other variable.LEARN
Both lines go through (xˉ,yˉ)(\bar{x}, \bar{y}): solving the two equations together gives the means. They are one line only if r=±1r = \pm 1.LEARN
Trust a prediction only with strong correlation and a given value inside the range of the data. For a line of xx on yy, check the yy data.LEARN

4.11Conditional probability

P(A∣B)=P(A∩B)P(B)\mathrm{P}(A \mid B) = \dfrac{\mathrm{P}(A \cap B)}{\mathrm{P}(B)}BOOKLET
P(A∩B)=P(B) P(A∣B)\mathrm{P}(A \cap B) = \mathrm{P}(B)\,\mathrm{P}(A \mid B)LEARN
The given event, after the bar, goes on the bottom. P(A∣B)\mathrm{P}(A \mid B) and P(B∣A)\mathrm{P}(B \mid A) are usually different.LEARN
P(A′∣B)=1−P(A∣B)\mathrm{P}(A' \mid B) = 1 - \mathrm{P}(A \mid B), but P(A∣B′)\mathrm{P}(A \mid B') needs the definition again.LEARN
With a random variable, the top is the overlap of the two inequalities: P(X>7∣X≤9)=P(8≤X≤9)P(X≤9)\mathrm{P}(X > 7 \mid X \le 9) = \frac{\mathrm{P}(8 \le X \le 9)}{\mathrm{P}(X \le 9)}.LEARN
P(A∣B)=P(A)=P(A∣B′)\mathrm{P}(A \mid B) = \mathrm{P}(A) = \mathrm{P}(A \mid B')LEARN
P(A∩B)=P(A) P(B)\mathrm{P}(A \cap B) = \mathrm{P}(A)\,\mathrm{P}(B)BOOKLET
To test independence, compare two values that would be equal, write == or ≠\ne, then conclude.LEARN

4.12Standard normal variables

z=x−μσz = \dfrac{x - \mu}{\sigma}BOOKLET
x=μ+zσx = \mu + z\sigmaLEARN
P(X<x)=p  ⟹  x−μσ=z, where P(Z<z)=p\mathrm{P}(X < x) = p \implies \frac{x - \mu}{\sigma} = z, \text{ where } \mathrm{P}(Z < z) = pLEARN
zz counts standard deviations from the mean: positive above, negative below. Rearranged, x=μ+zσx = \mu + z\sigma.LEARN
Z∼N(0,1)Z \sim \mathrm{N}(0, 1) is the standard normal, and P(X<x)=P(Z<z)\mathrm{P}(X < x) = \mathrm{P}(Z < z).LEARN
Unknown μ\mu or σ\sigma: inverse normal with mean 00 and standard deviation 11 gives zz, then solve x−μσ=z\frac{x - \mu}{\sigma} = z.LEARN
Both unknown: two probabilities give two equations in μ\mu and σ\sigma. Solve them together.LEARN
Inverse normal wants the area to the left: for P(X>x)=p\mathrm{P}(X > x) = p, use 1−p1 - p.LEARN

4.13 ◆Bayes’ theorem

P(B∣A)=P(B) P(A∣B)P(B) P(A∣B)+P(B′) P(A∣B′)\mathrm{P}(B \mid A) = \dfrac{\mathrm{P}(B)\,\mathrm{P}(A \mid B)}{\mathrm{P}(B)\,\mathrm{P}(A \mid B) + \mathrm{P}(B')\,\mathrm{P}(A \mid B')}BOOKLET
P(Bi∣A)=P(Bi) P(A∣Bi)P(B1) P(A∣B1)+P(B2) P(A∣B2)+P(B3) P(A∣B3)\mathrm{P}(B_{i} \mid A) = \dfrac{\mathrm{P}(B_{i})\,\mathrm{P}(A \mid B_{i})}{\mathrm{P}(B_{1})\,\mathrm{P}(A \mid B_{1}) + \mathrm{P}(B_{2})\,\mathrm{P}(A \mid B_{2}) + \mathrm{P}(B_{3})\,\mathrm{P}(A \mid B_{3})}BOOKLET
Top: the one path through the event you want. Bottom: every path that ends in the given event, which adds up to P(A)\mathrm{P}(A).LEARN
P(B∣A)\mathrm{P}(B \mid A) and P(A∣B)\mathrm{P}(A \mid B) are usually different. A rare cause can stay unlikely even after strong evidence.LEARN
If P(A∣B)=P(A∣B′)\mathrm{P}(A \mid B) = \mathrm{P}(A \mid B'), then P(B∣A)=P(B)\mathrm{P}(B \mid A) = \mathrm{P}(B), and AA and BB are independent.LEARN

4.14 ◆Discrete and continuous random variables

Var(X)=E[(X−μ)2]=E(X2)−[E(X)]2\mathrm{Var}(X) = \mathrm{E}\left[(X - \mu)^{2}\right] = \mathrm{E}(X^{2}) - \left[\mathrm{E}(X)\right]^{2}BOOKLET
∫−∞∞f(x) dx=1\int_{-\infty}^{\infty} f(x)\,\mathrm{d}x = 1LEARN
P(a≤X≤b)=∫abf(x) dx\mathrm{P}(a \le X \le b) = \int_{a}^{b} f(x)\,\mathrm{d}xLEARN
∫−∞mf(x) dx=12\int_{-\infty}^{m} f(x)\,\mathrm{d}x = \frac{1}{2}LEARN
E(X)=μ=∫−∞∞xf(x) dx\mathrm{E}(X) = \mu = \int_{-\infty}^{\infty} x f(x)\,\mathrm{d}xBOOKLET
Var(X)=∫−∞∞(x−μ)2f(x) dx=∫−∞∞x2f(x) dx−μ2\mathrm{Var}(X) = \int_{-\infty}^{\infty} (x - \mu)^{2} f(x)\,\mathrm{d}x = \int_{-\infty}^{\infty} x^{2} f(x)\,\mathrm{d}x - \mu^{2}BOOKLET
E(aX+b)=aE(X)+b,Var(aX+b)=a2Var(X)\mathrm{E}(aX + b) = a\mathrm{E}(X) + b, \quad \mathrm{Var}(aX + b) = a^{2}\mathrm{Var}(X)BOOKLET
σaX+b=∣a∣ σX\sigma_{aX + b} = \lvert a\rvert\,\sigma_{X}LEARN
Var(X)=E(X2)−[E(X)]2,σ=Var(X)\mathrm{Var}(X) = \mathrm{E}(X^{2}) - \left[\mathrm{E}(X)\right]^{2}, \quad \sigma = \sqrt{\mathrm{Var}(X)}BOOKLET
A pdf has f(x)≥0f(x) \ge 0 and total area ∫f(x) dx=1\int f(x)\,\mathrm{d}x = 1. Probabilities are areas, so P(X=a)=0\mathrm{P}(X = a) = 0.LEARN
Split every integral of a piecewise density where the formula changes, and include every piece.LEARN
Mode: the xx where ff is largest (check the ends too). Median: ∫−∞mf(x) dx=12\int_{-\infty}^{m} f(x)\,\mathrm{d}x = \frac{1}{2}.LEARN
E(X)=∫xf(x) dx,E(X2)=∫x2f(x) dx\mathrm{E}(X) = \int x f(x)\,\mathrm{d}x, \quad \mathrm{E}(X^{2}) = \int x^{2} f(x)\,\mathrm{d}xLEARN
A game is fair when the player's expected gain is 00.LEARN

Which method when

Is a value an outlier? → compare it with Q1−1.5×IQRQ_{1} - 1.5 \times \mathrm{IQR} and Q3+1.5×IQRQ_{3} + 1.5 \times \mathrm{IQR} (4.1)LEARN
Predict yy from xx → regression line of yy on xx. Predict xx from yy → line of xx on yy. Only inside the range of the data (4.4, 4.10)LEARN
“At least one” → 1−P(none)1 - \mathrm{P}(\text{none}) (4.5)LEARN
Two stages, with or without replacement → tree diagram: multiply along a path, add across paths (4.6)LEARN
“Given that” → P(A∣B)=P(A∩B)P(B)\mathrm{P}(A \mid B) = \frac{\mathrm{P}(A \cap B)}{\mathrm{P}(B)}, with the given event on the bottom (4.6, 4.11)LEARN
Are AA and BB independent? → compare P(A∩B)\mathrm{P}(A \cap B) with P(A) P(B)\mathrm{P}(A)\,\mathrm{P}(B), then conclude (4.6, 4.11)LEARN
Unknown kk in a probability distribution → the probabilities add up to 11. Fair game → E(X)=0\mathrm{E}(X) = 0 (4.7)LEARN
Fixed number of independent trials, two outcomes, same pp → binomial, X∼B(n,p)X \sim \mathrm{B}(n, p) (4.8)LEARN
Normal, value to probability → normal cdf. Probability to value → inverse normal, using the area to the left (4.9)LEARN
Normal with μ\mu or σ\sigma unknown → get zz from the inverse normal of N(0,1)\mathrm{N}(0, 1), then solve x−μσ=z\frac{x - \mu}{\sigma} = z (4.12)LEARN

◆Which method when

Probability of a first-stage event, given what happened next → Bayes' theorem: that path over every path to the given event (4.13)LEARN
Continuous random variable with pdf ff → probabilities by ∫f(x) dx\int f(x)\,\mathrm{d}x; median mm from ∫−∞mf(x) dx=12\int_{-\infty}^{m} f(x)\,\mathrm{d}x = \frac{1}{2}; E(X)=∫xf(x) dx\mathrm{E}(X) = \int x f(x)\,\mathrm{d}x (4.14)LEARN
Follows the IB Maths AA guide (first assessment 2021). Not affiliated with the IB.