I’m reading about it for my paper, but I can’t understand most of it because of the super complex jargons. Hoping anyone can help. Thanks!
In: Mathematics
So we’re talking about two ‘signals’, which just means two sets of data, and we want to see if one of them ‘controls’ or ’causes’ the other.
Let’s say we have a signal – it could be the population of Norway, or Illinois employment numbers, or the Dow Jones average, anything like that. We’ll call that X2. I’ll give you all the X2 information I have, from January to November of this year, and then I’ll ask you to use that information to predict what X2 will do in December.
Now I bring out another signal, called X1. Again, this could be COVID numbers, my weight, anything. You get to keep all the X2 information I gave you before, and also you now get the X1 information from January to November, and *again* I ask you to predict what X2 will do in December.
If your second prediction is better, that means the X1 signal helped you, and that means X1 “passes the Granger test”; X1, in some sense, controls or causes X2; it contains information about how X2 will behave. If that happens, we say that X1 *Granger-causes* X2, or there is a *Granger-causality* relationship between X1 and X2.
Latest Answers