If someone wants to supply the data, I would be happy to do the T Test and report back. That is the correct test for distinguishing real change from random noise.

Last edited by denton; 01/28/21.

Be not weary in well doing.