I think your survey should have more questions to eliminate some variables in your survey, especially since you are trying to prove or disprove a statement.
How long have you had chickens? -New chicken keepers may never have lost one.
What steps did you go to to insure the safety of your chickens? (Chicken wire run v.s. welded wire/hardware cloth run, predator proof latches, automatic door v.s. manual pop door, etc.) -this question will make a big difference in your results. Not having this question will skew the data you collect because you can't show that more people who use bad safety choices loose more birds, or visa versa.
What predators are common in your area? -Raptors can be a big problem in uncovered runs, whil a net up top wont stop a raccoon. Cyotes can't climb, but they do dig, so an apron would be nessacary. You can compare the local predators with the security steps somone has instilled to find the most likely cause of a loss.
Where do you live: city, suburbs, or country? -Cities will have less cyote and bobcat problems, but racoons will be more prevelant.
You should take these things into account before trying to prove if lack of security leads to losses in the flock. Your data will be flawed otherwise.
Good luck!
