Found this through search years later and it is still the best discussion of the question, so adding one thing. If you are going to change something, change one thing and give it long enough to express itself. Four weeks is the usual minimum for anything pharmacological on this board, and two weeks of data has told you almost nothing.

Clinical perspective, offered as context rather than as advice. If two explanations both fit, the useful question is which one predicts something the other does not. That is answerable; arguing about which sounds more plausible is not.

julia.endo said:

If two explanations both fit, the useful question is which one predicts something the other does not.

This is my experience too, for whatever a second data point is worth. I had assumed I was the exception until I read this.

julia.endo said:

If two explanations both fit, the useful question is which one predicts something the other does not.

Adding the part of the answer the thread has not reached. Order of operations matters more than any single choice here: establish a baseline, change one thing, wait long enough for it to express itself, then measure again under the same conditions.

One concrete data point for the thread. One habit that pays for itself: post the method alongside the number. A figure without its method cannot be checked, and an unchecked figure is how this community accumulates folklore.