A short literature search will show you that all the drivers account. This will include the crop factor, soil, and other components. Basically, you need to determine your plant 100% water conditions and calculate irrigation accordingly.
The best way I think to create a variable drought experiment is to use tensionometer reading for the application of water. I would keep the water amount at a constant rate of application but apply at different triggering reading levels.
Giving the pots 10, 20 or 30% less water than a control would not necessarily result in any drought stress in the plant. Many factors such as the soil properties, the pot size, the crop that is grown and the humidity in the growth chamber could influence the rate of drying as well as the level of stress. It is important to consider plant water content when quantifying the stress.