cities?" In other words, "do people in the 6 cities wear the different mask types with similar likelihood?"
Let's say you have 4 types of masks: N95, pleated blue polyester, cloth, neck gaiter. Then you'll have a data table that is 4x6 with how many degrees of freedom? With this table you only have one overall H0: the dist'n of mask types is the same for all 6 cities. The corresponding H1 is "at least one city is not like the others."
What happens if you have a violation of Cochran's Rule among the 24 expected values? What will you do?