Calculates Durbin's h-statistic for autoregressive models.
The model being assessed
The variable in the model that represents the lag of the y-term
Using the Durbin-Watson (DW) test for autoregressive models is inappropriate because the DW test itself tests for first order autocorrelation. This doesn't apply to an ECM model, for which the DW test is still valid, but the durbinH function in included here in case an autoregressive model has been built. If Durbin's h-statistic is greater than 1.96, it is likely that autocorrelation exists.
Numeric Durbin's h statistic
1 2 3 4 5 6 7 8 9 10 11
##Not run #Build a simple AR1 model to predict performance of the Wilshire 5000 Index data(Wilshire) Wilshire$Wilshire5000Lag1 <- c(NA, Wilshire$Wilshire5000[1:(nrow(Wilshire)-1)]) Wilshire <- Wilshire[complete.cases(Wilshire),] AR1model <- lm(Wilshire5000 ~ Wilshire5000Lag1, data=Wilshire) #Check Durbin's h-statistic on AR1model durbinH(AR1model, "Wilshire5000Lag1") #The h-statistic is 4.23, which means there is likely autocorrelation in the data.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.