Data reuse

I quite enjoyed reading about post-empirical research in this week's Knight reading, and noticed that it overlaps with the ideas of data curation and sharing discussed in the data librarianship course that I am currently in. It seems that the idea of data reuse and sharing is relatively new as it stands, and small-scale researchers have generally focused  on producing and disseminating articles/papers or other textual accounts rather than on the data itself as a product of research. Unfortunately this means that while small-scale researchers have collected huge amounts of data, it is largely unavailable for anything beyond the original research project even though it could be useful for further conceptual analysis.

Of course this raises quite some issues. One from the reading that raises a red flag for me is the claim that statistically insignificant results could be combined to reach more powerful findings. While I like the idea, it seems like it would be tough to do this while avoiding disputes that the results are invalid because even minor differences in methodology are thought to make data sets incomparable.