That "much worse" at 3% was represented by a single reply.

When i read about the study my biggest surprise was how poorly it was. A poll with 40 people is so pointless that I don't understand why they didn't hire professionals to do a real survey.