To ignore aspects that don't fit the hypothesis is definitely not science. (realclimate.org)
You have to «explain» facts that don't fit the hypothesis. (drtimball.com)
There is a continuum from corrections based on known problems (essential), to adjustments based on probable errors in the data (good practice as long as all assumptions are made clear), to adjustments done solely to make the data fit a hypothesis (distinctly dodgy). (newscientist.com)