Huge sample studies with high stat significance can have very low effect sizes, though. If an effect is shown on a small sample, it means it is a really big effect, and will likely be reproduced in a bigger study.