Econometrics Glossary

Correlation between the error terms at different time periods in a regression, common in time series data and violating the assumption of independent errors.

A statistical relationship in which two or more non-stationary series share a common stochastic trend, so a linear combination of them is stationary.

A quasi-experimental method that estimates causal effects by comparing outcome changes over time between treatment and control groups.

The branch of economics that applies statistical and mathematical models to economic data to test theories, estimate relationships, and make forecasts.

The condition where an independent variable is correlated with the regression error term, leading to biased and inconsistent parameter estimates.

A test of joint significance used to determine whether a group of coefficients are simultaneously equal to zero.

A panel data estimation approach that controls for all time-invariant unobserved heterogeneity by demeaning or differencing the data within each entity.

Generalized Autoregressive Conditional Heteroscedasticity, a class of models for time-varying volatility in financial time series.

Non-constant variance of the error term across observations in a regression model.

Constant variance of the error term across all observations, a key assumption for valid OLS inference.

A variable used to obtain consistent estimates when endogeneity is present; it must be correlated with the endogenous regressor and uncorrelated with the error term.

A parameter estimation technique that finds the values maximizing the probability of observing the data under the specified model.

High linear correlation among two or more independent variables in a regression model, which inflates the variance of coefficient estimates.

Bias in coefficient estimates caused by excluding a relevant variable that is correlated with both the dependent variable and an included regressor.

A regression estimation method that minimizes the sum of squared residuals to find the best linear fit between dependent and independent variables.

A dataset that tracks multiple cross-sectional entities (individuals, firms, countries) across multiple time periods.

The proportion of variance in the dependent variable explained by the regression model, ranging from 0 to 1.

A panel data estimation approach that treats unobserved entity-specific effects as random variables uncorrelated with the regressors, using generalized least squares.

A causal inference approach that exploits a discontinuity in treatment assignment at a threshold in a continuous running variable.

Standard error estimates that remain valid in the presence of heteroscedasticity or clustering, allowing correct inference without assuming a specific variance structure.

A regression that yields statistically significant results between unrelated non-stationary variables due to shared stochastic trends.

A property of a time series whose statistical characteristics (mean, variance, autocovariance) do not change over time.

A test statistic for individual coefficient significance, calculated as the estimated coefficient divided by its standard error.

A feature of a non-stationary time series process in which shocks have permanent effects, causing the series to drift without reverting to a mean.

A multivariate time series model in which each variable is regressed on its own lagged values and the lagged values of all other variables in the system.