Forget the quality of the data, you can give the same data to different researchers and get wildly different conclusions. [1, from 2, from 3]

(Disclaimer: not my area of expertise.)

[1] N. Breznau et al., “Observing many researchers using the same data and hypothesis reveals a hidden universe of uncertainty,” PNAS, October 2022, URL: https://www.pnas.org/doi/10.1073/pnas.2203150119

[2] B. Klaas, “The Crisis of Zombie Social Science.” The Garden of Forking Paths. https://www.forkingpaths.co/p/the-crisis-of-zombie-social-sc...

[3] The crisis of zombie social science (20 points, 2 days ago) – https://news.ycombinator.com/item?id=44272057