Probability is the measure of how likely an event is, expressed from 0 (impossible) to 1 (certain); it is the foundation of statistical analysis because all conclusions from data involve uncertainty.
Definitions
- Classical: \(P(A) = \dfrac{\text{favourable outcomes}}{\text{total equally likely outcomes}}\). Drawing a king from a deck: \(4/52 = 1/13\).
- Empirical (relative frequency): \(P(A) \approx\) (number of times \(A\) occurred)/(number of trials). If 45 of 500 products are defective, \(P(\text{defective}) \approx 0.09\).
- Axiomatic: \(P(A)\ge 0\), \(P(S) = 1\), and \(P(A\cup B) = P(A)+P(B)\) for mutually exclusive events.
Key rules: addition rule \(P(A\cup B) = P(A)+P(B)-P(A\cap B)\); multiplication rule \(P(A\cap B) = P(A)P(B\mid A)\); Bayes' theorem for updating probabilities with new evidence.
Significance in statistical analysis
- Modelling random variation: probability distributions (binomial, Poisson, normal) describe data such as number of defects or measurement errors.
- Sampling and inference: random sampling relies on probability; it lets us estimate population values from a sample with a known margin of error.
- Hypothesis testing: p-values and significance levels are probabilities that tell us whether an observed difference could be due to chance.
- Confidence intervals: "95% confidence" is a probability statement about the method.
- Decision making under risk: expected value = \(\sum x\,P(x)\) is used in insurance, finance and business planning.
- Prediction and machine learning: classifiers such as Naive Bayes and logistic regression output probabilities; risk scores and weather forecasts are probability statements.