Mistakes that data science students make

I admire the author's effort in creating research like this, but the truth is that data science has been taught using the current teachings of "first try the code and then learn" approach, and some people manage to evolve and motivate themselves like this, but others don't. The truth is that to do well in the world of technology, you need to use many skills that are not taught in these courses, like Coursera for example. things are very "chewed", at the first bug due to different versions of a library, the student has already given up. There's no point in having everything ready, pressing play on some Jupyter notebooks and getting a certificate. Data science students need to get comfortable of doing hard things and being wrong 90% of the time

What value do you see in doing this? I think that in this cases we know how to teach people how to code so they can improve. In my experience mentoring Data Scientists and Data Science students I always felt like there's much bigger blockers to value than the technical skills itself.