There is an interesting discussion (take down?) of a set of these experiments as referenced in Lera Boroditsky's previous WSJ piece at language log:
http://languagelog.ldc.upenn.edu/nll/?p=2592
Of course Liberman seems to be firmly against this new spate of Sapir-Whorf ideas. Personally I think the distinction between languages is less relevant than distinction between metaphors used, which of course may vary by language. This manifests itself in cases even with single language speakers: if readers are given two different metaphors for the same topic, they will behave differently (crime as a virus vs. crime as a beast being their example).
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjourna...
Which the US government is interested in for terrorist understanding:
http://www.theatlantic.com/technology/archive/2011/05/why-ar...