If we ask a question to 120 people and we get 80 yes, 20 no and 20 unanswered. The outcome is 80% because the 20 unanswered are ignored.
If we combine 2 x buildings and only one of them included that specific question it may look like we asked a question to 240 people and we get 80 yes, 20 no, 20 unanswered and 80 not asked. The result is we still need to arrive at 80%.
the survey used was the same master survey for both
one survey disabled the specific question and one survey kept it active
they are combined using the combine function
Can you please confirm that statistics will calculate as we want it per the above scenario/methodology?