Thanks for details, helps to see clearer.
So yes I agree, the deviation (> 10) comes from the bubble.

Any chance you could use a thinner tube (e.g. ID = 2mm)? This will still fit on the doser fitting.

Still try to get rid of the air in the reagent - make it a bit warm (e.g. 30 - 40°C), shake it carefully