Can anyone explain where the percentage error is coming from?

Hi Duval:

In table 3, "Error from Hand based on change in water level" is based on the change in water level height. If you calculate the error at each specific water level height, it just gives us an indication of how parasitic capacitance or "hand interference" affects the measurements per water level.

A better way to represent error is based on the dynamic range of the system, so that is why it is based on change in water level which tells us how the system changes as your dynamic range increases. As your water level increases, your error should decrease at higher water levels even though the change in capacitance from baseline may be similar at all water levels. Water level 0 is our "baseline", this is used in the percentage calculation since the "capacitance" is not necessarily 0.

If you just calculate percent error based on a specific water level, if you have similar change in capacitance behaviors, then your error would be worse as your water level decreases.

Hope that makes sense.