Ah, thanks for checking. There are a few other hypotheses then:
Testing volumes are low, meaning we fail to capture a sizeable chunk of cases.
The testing strategy is bad. We’re testing people that are less likely to be positive.
And possibly, the model is just wrong.
