REGSTRTEST - General LM Test for STR
Posted: Wed Apr 17, 2013 1:51 pmThis implements a general LM test for linearity vs nonlinearity in the form of Smooth Transition (either logistic or exponential) based upon a threshold variable. The basic idea behind this comes from Luukkonen, Saikkonen and Terasvirta(1988), "Testing Linearity against smooth transition autoregressive models", Biometrika vol 75, 491-499; this is just applied to a regression which isn't necessarily an autoregression and in which the threshold variable isn't necessarily one of the regressors.
The first step is to run the base linear regression. Then do
@RegSTRTest(options)
immediately afterwards.
Options
THRESHOLD=series for threshold variable [dependent variable of regression]
D=delay on threshold variable [1]
TITLE=title for output ["Test for STR in regression"]
WEIGHTS=series of observation weights [equally weighted]
[PRINT]/NOPRINT
The WEIGHTS option can be used to adjust for outliers (viewtopic.php?f=8&t=1587).
The output is organized the same as @STARTest (viewtopic.php?f=7&t=1456) and the tests have the same interpretations.
RegSTRTest.src- Procedure file
- (5.04 KiB) Downloaded 23 times
The first step is to run the base linear regression. Then do
@RegSTRTest(options)
immediately afterwards.
Options
THRESHOLD=series for threshold variable [dependent variable of regression]
D=delay on threshold variable [1]
TITLE=title for output ["Test for STR in regression"]
WEIGHTS=series of observation weights [equally weighted]
[PRINT]/NOPRINT
The WEIGHTS option can be used to adjust for outliers (viewtopic.php?f=8&t=1587).
The output is organized the same as @STARTest (viewtopic.php?f=7&t=1456) and the tests have the same interpretations.