Science is a method

Empirical         (measuring devices) 

Longitudinal    (time span)

Double-blind   (not allowing anybody in the study to know anything including the testers, no one is allowed to know anything so they can't mess with it. They don't know where what when why and why, and studies running all the time and they randomize them, keys and codes are put in safes that need 3 to 5 people to open in a public ceremony, lottery draw and then the encrypted keys are keep in multikey safes in public and monitored, so nobody has the slighest clue whats going on)

Meta-analysis (do the math on all the studies at once, what words they use and how many times, what sentence is most common, what sets them apart. How many letter t's are in it. Make graphs and charts of it all, time place people price materials labs )

Peer tested for 10 years (after ten years it can be classified as science if its still standing up to private investigators.)

Obviously almost nothing is. 

Scientists love saying "Its the most tested model in physics" because its barely to rarely ever being tested.  Anything tested with all those methods will never change , theres barely any science.

The word placebos meaning changing on the sly should be in meta study and that behavior as graph  , thats a  very common corruption.  What is the placebo substance, its not sugar no more and now got all these amazing wonder drugs  yet if they tested against sugar the'd be trash