>Remove of alter some data records (observations) to better support the research hypothesis
There are limited (repeat LIMITED) instances where this is appropriate. But you had better damn well explain why those were removed, other than just saying they're "outliers" and going about your day.