Suppose that 3000 drivers in Wakanda were randomly breath-tested on 21 April 2019 and 116 were above the limit of 0.05 blood- alcohol level. On 15 May
2019, 4000 drivers were tested and 98 were above this level.

What additional information would you require before trying to draw conclusion
from these data?

What factors, other than a real change in driver behaviours, could cause such a
drop in the proportion of above-the-limit drivers.