Yes! Maybe a slight variant. In this case the overall effect is evident when you look at the whole, and not when you group by level. The important question is

Yes! Maybe a slight variant. In this case the overall effect is evident when you look at the whole, and not when you group by level. The important question is whether it seems causally reasonable to group or not group. In this case, grouping makes sense.