OP's title changes are messy
I think these stats are fair, the sample size is huge, the margin of error is below 5 (increasing the confidence interval), they acknowledged confounding variables, and im sure they did random sampling as well.
Artemisia does make a good point, but then again im sure the parents/adults have heavy influence in what the kids think.
i think a less messy title would be something like " on
average 14% of Nigerians are ISIS Supporters" - of course you want to add the " ages 18+" but for the sake of title limit i think thats the least messy title.