Statistical glossary
Concise definitions, interpretation guidance, and direct links to practical articles and DataStatPro tools.
Alpha level
The probability threshold chosen before analysis for controlling the long-run rate of false-positive decisions under the null hypothesis.
Study DesignAttrition
The loss of participants or observations between recruitment and the final analysis.
ForecastingAutocorrelation
Correlation between values of the same variable observed at different time points or lags.
Statistical MethodsConfidence interval
A range produced by a repeated-sampling procedure that quantifies uncertainty around an estimated parameter.
Quality ImprovementControl limits
Statistically calculated boundaries used to identify unusual process variation on a control chart.
Statistical MethodsEffect size
A quantitative measure of the magnitude of a difference, association, or model effect.
Statistical MethodsEta squared
An ANOVA effect-size measure describing the proportion of total variation associated with an effect in a specified analysis.
Study DesignEstimand
The precise quantity that a statistical analysis is intended to estimate.
Statistical MethodsInteraction effect
A pattern in which the effect of one variable differs across levels or values of another variable.
Research ReportingP value
The probability, under a specified null model, of obtaining data at least as incompatible with that model as the observed data.
Statistical MethodsPartial eta squared
An ANOVA effect-size measure using variation for the target effect relative to that effect plus its associated error.
ForecastingPrediction interval
A range intended to contain a future observation or aggregate under a specified forecasting or regression model.
Study DesignSphericity
A repeated-measures assumption requiring equal variances for the differences between pairs of conditions.
Statistical MethodsStandard error
The estimated standard deviation of an estimator across repeated samples under a statistical model.
Study DesignStatistical power
The probability that a planned analysis will meet its decision criterion when a specified alternative is true.
ForecastingStationarity
A time-series property in which key distributional features remain stable over time under the chosen definition.
Statistical MethodsType I error
Rejecting a null hypothesis when that null model is true under the stated testing procedure.
Statistical MethodsWilson interval
A confidence interval for a binomial proportion based on inversion of the score test.