Couple of questions about Wald structural break test- (running latest Stata in Win 10)
When running -sbknown- with two given dates, does it identify whether there is a structural break in the trend BETWEEN the dates? For example I'm identifying breaks in 17 years of monthly population counts - if I feed two separate dates, and it gives a test statistic with a p value of <.05, does that mean there is a structural break in the series between the two dates which I use in the syntax?
Also, is there any reason why one would get different results when using -sbsingle- (searches a time series for the first structural break from a given point in time) versus the aforementioned -sbknown-? As far as I can see from searching the help file and PDF documentation, the underlying math is the same - Wald test. Or it should be.
When running -sbknown- with two given dates, does it identify whether there is a structural break in the trend BETWEEN the dates? For example I'm identifying breaks in 17 years of monthly population counts - if I feed two separate dates, and it gives a test statistic with a p value of <.05, does that mean there is a structural break in the series between the two dates which I use in the syntax?
Also, is there any reason why one would get different results when using -sbsingle- (searches a time series for the first structural break from a given point in time) versus the aforementioned -sbknown-? As far as I can see from searching the help file and PDF documentation, the underlying math is the same - Wald test. Or it should be.
