[sanxiyn]

I agree the problem here is an incorrect model. Mask does not act on seroprevalence. Measuring mask's effect on seroprevalence is just wrong study design, although it may be easier to do.