Statistical parameter

Common types and notation

Typical parameters include the population mean, variance, proportions, regression coefficients, and parameters of parametric families (such as the rate of an exponential distribution or the mean and variance of a normal distribution). Parameters are often denoted by symbols such as θ, μ, σ2; while the article cannot render every mathematical symbol, a common schematic notation is shown near this description: . Parameters are treated as fixed but unknown quantities in frequentist settings and may be given probability distributions in Bayesian analysis.

Parameters versus estimators

Parameters differ from estimators. A parameter is a constant attribute of the population or model, while an estimator is a rule—often a statistic computed from sample data—used to infer the parameter. An estimator varies from sample to sample and has its own sampling distribution; its typical notation uses a hat, for example θ versus θ^{^}, illustrated here: . For more on estimator construction and properties, see materials on estimators and inference.

Role in practice and examples

Parameters are the targets of estimation, testing, and prediction. In applied work they guide hypothesis tests, confidence intervals, model selection, and forecasting. Examples: a public-health study may estimate a population proportion of vaccinated individuals; an econometric model estimates coefficients that quantify relationships between variables; a reliability engineer models component lifetimes with a rate parameter.

Important distinctions and notes

• Fixed vs. random: In classical inference parameters are fixed unknowns; in Bayesian inference they are random variables with prior distributions.
• Identifiability: A parameter must be distinguishable from data and other parameters; nonidentifiable parameters cannot be reliably estimated.
• Parametric vs. nonparametric: Parametric models specify a finite set of parameters, whereas nonparametric approaches avoid strong parametric assumptions.

Brief history and further reading

The formal idea of a parameter grew with the development of probability and mathematical statistics in the 19th and early 20th centuries, as statisticians formalized models for populations and sampling. Contemporary texts treat parameters as central objects of inference and modeling; introductory resources and advanced treatments are widely available for those who wish to explore estimation theory, Bayesian methods, and identifiability in greater depth.