I also got this one the first time I did it so 2 things helped me during the blind review.

In the first question asked, I interpreted the word “with” as an “and” statement so that we know that 40% of the 200 from the first group experienced both an awakening where they felt seemingly paralyzed AND a strange presence in the room at the same time

Some key word are that both groups are randomly selected and have different subjects, therefore there’s no overlap between groups and no known group shared characteristics. Therefore, we only know that 40% of one group of 200 people experienced both paralysis AND a presence while a completely different group of 200 people have only 14% have woken up feeling paralyzed.

Considering that “and statements” theoretically have a lower probability of being true (ie. The chances of rolling a 4 on a four sided die one time (0.25 chance) is greater than rolling a 4 on the die the first roll AND a 2 no the die a second time (0.25x0.25=0.063) it and that these groups of people should both represent the general population then there’s no way that the first group had so many more people experience both simultaneously as opposed to just one experience.

That leads us to E, since we don’t know whether one of event causes the other as in A. But we do know that the question with both experiences had higher reports than the question with one.

This was my more thoughtful line of logic hopefully it helps! I think you were right to think it was a weird relationship. Maybe looking into correlation v causation may be helpful for future problems!