10k were interviewed. They extrapolate 10k to the population of Europe. C'mon that's just mad. Especially with something as complicated as poverty.
Europe also currently has a war ongoing and has huge areas of incredibly poor and wealthy. Can't really average that out. Wouldn't trust anything that comes out of this research institution
If that 10k is representative, that should give a very small uncertainty interval, less than 1%. You can get 95% confidence interval with only a few hundred samples depending on the standard deviation, so 10k is actually massive. It's pretty standard statistics, here's more info on how it's calculated.