The great thing about science evidence is that it can be tested.
If someone fakes up a study or conceals unfavourable results, this can’t stop other experimenters from analysing the data statistically and/or producing new data by repeating the experiment.
Thereby, the flaws can be revealed.