Not great stats
10% of 400 respondents is ~40 developers used Ruby at least some of the time last year. Now there are 40% more; i.e. 56 developers. Is an increase of 16 respondents in the sampled population of developers statistical significant?
Without looking at the base data properly, it could just be that within the 400 respondents, perhaps 100 change language as projects come and go in any given year, meaning that this is just a blip one way or another.
Oh.. and "commercial SQL databases are [tow] and a half times more likely to be used as a primary database than open source SQL databases"? I find that a little strange as most commercial database vendors also offer free versions; so I would suspect that many developers (such as myself) use open-source RDB' like MySQL for many small projects, but might still consider a commercial SQL database like Oracle or MSSQL as a 'primary' for larger production sites.