Cargo cult science.
As the good doctor said, "That's funny..." is exciting in science. That's when the data doesn't meet expectations. At which point you check the equipment, check the assumptions, check the calculations. The first and best guess is always that you made a mistake somewhere. When peer review works properly, that is what it does; get a whole bunch of other eyes looking at it to see if THEY can figure our where you messed up.
Or you can double down on the finding, argue the reviews, selectively search the literature for data that confirms (never data that disagrees). Not saying it doesn't happen. It does, a lot. We all do it. It even has some utility. But it is also characteristic of bad science.