[scalable]

Not even a large sample will help against systematic errors.

[phren0logy]

Good point; upvoted. However I was just trying to address the idea that there are ways of minimizing problems in study design. No study is ever perfect, but many of them are sufficient.