I don't think the API is well suited to data analysis. If you want to do analysis of the SE dataset(s), you should download the latest [torrent file][1]. Depending on how which, and how much, data you need for your analysis, you *might* be able to combine the dump with some judicious use of the API to pull current metadata from the site. [1]: http://blog.stackoverflow.com/2009/06/stack-overflow-creative-commons-data-dump/