studies are expensive so they end up with small sample sizes. Add in the push to only publish exciting findings and you start to see why so much research is later invalidated.