Annime Wong on Nostr: Doing that for every paper would be impractical. First off, there is a barrier ...

Doing that for every paper would be impractical. First off, there is a barrier between fields. A mathematician (or as it would be in the case of this paper, a statistician) may not understand the data the astronomer is working with (this can be a big deal because astronomers do a lot of tweaking to extract data from what is basically noise, which skews it's statistical nature). Conversely, Astronomers and physicists tend to have a very poor background in statistical analysis (this actually frustrates me because two of the biggest fields in physics, Quantum Mechanics and Thermodynamics, are really just applications of statistics). Ideally a peer-review process would catch these mistakes, but if no one understands the math behind it well enough, then the "experts" will inevitably make mistakes. But also, that's a lot of papers, it's unrealistic to expect everyone to hire a mathematician to check the most basic aspects of your work. In theory you should be competent enough to do it yourself.

The only way forward is to train astronomers so that they are more well versed with the math they are trying to apply. This means actually teaching them statistics, which they currently do not do to any significant degree.