Hi,

I am having some problems with the basics of regressions.
I have panel data, similar to what is shown below, with two variables (so far). If the value of the variable on the left (SPS) changes from a 0 to a 1, like in the data below. and then the variable on the right (NTM) changes a few periods after, is it possible that the regression will consider that this change in the SPS variable is correlated to the change in the NTM variable, even though the change in the NTM is not in the same time period?
Does a regression consider correlation between one value and another value in the future?

The data below is an example of an event occuring in SPS and events occurring in NTM in a few periods afterwards.

The time is in dd/mm/yyyy format but here it is displayed in a another format. each row represents a month. The second row along is the SPS and the next is NTM.
Code:
* Example generated by -dataex-. To install: ssc install dataex
clear
input int time byte(SPS NTM)
15706 0 1
15737 0 0
15765 0 0
15796 0 0
15826 0 0
15857 0 0
15887 0 0
15918 0 0
15949 1 0
15979 0 0
16010 0 0
16040 0 2
16071 0 1
16102 0 0
16131 0 0
16162 0 0
16192 0 0
16223 0 0
end
format %tdnn/dd/CCYY time

Please let me know if this is unclear.
Thanks,
Anthony O'Dowd