I don't intend to put a fine point on it but do you mean by "scientific", the proper investigation such as a cohort or cross-sectional study into the phenomenon with conclusions peer-reviewed and published, etc, etc?
If so, I know of none. That said, there are plenty of anecdotal works around but what I see is almost always informal.
The difficulty with citing an accident like Colgan is, we have Turkish at Amsterdam, and the difficulty in citing Spanair at Madrid is we have the FedEx MD11 landing accident at Tokyo. By this I mean, it is possible to attribute handling issues in some but not all cases so anecdotally, the premise is inconclusive even though it is a felt characteristic at the present time.
To me this is where data programs come into their own but privacy, protection and sharing issues as well as an increasing propensity to prosecute rather than learn, hobbles any such formal study. But the data is there - I know it.