the data set is too small to draw any real conclusions
There are *so* many examples like this where there is a sound investigative idea and not a shred of methodological rigor.
Just a refresher for many of you:
1. Choose your outcome
2. Choose a threshold that will count as an important effect size
3. Power your study (choose how many units will be tested) based on #2
4. Document 1 - 3 before you do the study so we know you didn't go fishing for spurious correlations after the fact
5. Run the study
6. Inferential statistics!
7. Profit! Except, not really.
Those who think #6 is more about obscuring things than revealing them don't know much about #6.
Why on earth would you expect their sentences to be comparable?
Some shared commonsense notion of justice?
I jest.
And before we judge if that seems too harsh a punishment, I would ask if anyone knows what the Chinese government would do to an American engineer who did the same thing to a Chinese company.
It seems to me that what some other country would do to a similar criminal is irrelevant. If North Korea were involved, would that justify a still harsher punishment?
How about this: the punishment is fair because the guy is a crook and the crime wasn't petty (though as an automotive engineer in R&D I would agree that it was less consequential than advertised).
UNIX was not designed to stop you from doing stupid things, because that would also stop you from doing clever things. -- Doug Gwyn