Not my definition, as the scientific method is not up for interpretations as it's absolute. I suggest not using bastardised resources like wordnik.

Next you will be telling us gender is up for interpretation because it said so on some website.
How people interpret data is a problem, a BIG one.
I'll use what I know as an example - food.
Say we need to develop a product with a targeted 'overall liking' score of >70/100, which we will know will do well in market based on comparative benchmarking.
Easy example- say our beef pies have a liking score of 75 then the marketing team alerts us that we need to make sure our 'cheese and bacon beef pie' hits 70+.
Now... does 'science' care if we give these pies to the following groups for evaluation:
- Elderly people with no taste buds
- Infants with no recognition of what a pie is
- Europeans or Americans who associate 'pie' with desserts?
- Vegans who associate beef pie with murder
A lot of people focus on the 'sample size' when talking about food scores. Like, 10 people is great but 1000 people is 'way more scientific'.
Sooo... if you got 1000 eldery, infant, Euro, Yank or vegan to taste these pies then it's just going to tell you something (more 'scientifically' powerful, yes) but in the wrong direction. You want to find the motherfuckers who are going to the MCG and paying $8 and gagging for it with or after a pint.
This is where science goes wrong. I don't even need to detail the rest of it. But this is where science can be manipulated. Not a flaw with science itself but how it is deployed, and the potential deceptiveness.
This is why I think a degree in general science is helpful for youth. If you can learn how to screen information then you have an upper hand in a world where science is so manipulated the average punter doesn't have a chance. Hint: It's all targeted to the scientifically illiterate.