Testing Flashcards
What is the Augmented Dickey-fuller test?
Used to determine is a time series contains a unit root
Process of ADF
Regress the time series using OLS to create estimators
Calculate the test statistic ß = ß/SE(ß)
Compare test statistic to ADF critical values under the dickey fuller distribution
Either reject H0: unit root (non-stationary) or H1: no unit root (stationary)
What is the Breusch-Pagen test
A test to determine the presences of homoscedasticity
Breusch-Pagen procedure
Regress the model to find the residuals
Regress the squared residuals
Use the R^2 from the auxiliary and compute to test statistic
Test statistic: t = R^2 x n
Compare t to the critical values of the chi-squared distribution
t > critical value reject H0
t <= critical value fail to reject H0
Breusch-Pagen Hypotheses
H0: Presence of Homoscedasticity
H1: Presence of Heteroscedasticity
Distribution of ADF
Dickey-Fuller Distribution
Distribution of Breusch-Pagen
Chi-Squared distribution
T-test
Tests whether the estimator follows a t-distributed.
Test statistic of a t-test
T = estimated coefficient - observed coefficient / Standard Error (SE) of the estimated coefficient
Distribution of t-test
T-distribution
Procedure of t-test
Generate the estimated coefficient
Calculate the test statistic
Compare t-statistic to critical values under the t-distribution
If t > t critical values: reject H0
Hypothesis of t-test
H0: test statistic follows a t-distribution
H1: t <= critical value fail H0