For a large N, the study is only less likely to produce a false positive. It is more likely to produce a false negative though. There is an optimal N where type

For a large N, the study is only less likely to produce a false positive. It is more likely to produce a false negative though. There is an optimal N where type 1 and 2 error are minimized which is not infinity.