Thoughts on what I do:
1) I was taught the layman's version of the scientific method in school. When I'm wearing my technical support hat (daily) I use the layman's model to try to collect as much relevant data from real life as I can to cut branches off the troubleshooting tree to narrow down the potential hypotheses' to test. This is how I was trained.
2) Sometimes third-parties form their own narratives and demand compliance from the org I'm paid to represent without providing data to back up their claims. Almost always, the narratives are biased in favor of the organizations which provide the third-parties with the means of survival via employment, paychecks, shareholder value, etc.
I'm still learning how to navigate #2 from more experienced engineers.