I love meta commentary–that's why it's so much fun to read George Will and Charles Krauthammer–that's what they do: they make (usally wrong) observations on the logic of argument.  Well, at least someone is doing it.  For this reason I was glad to run across a phrase close to our heart here, "nut picking."  It appears in a Dave Weigel column in the Washington Post concerning misconceptions about the "Tea Party Movement." 

The tea party is racist.

2. It's a phenomenon that some activists call "nutpicking" — send a cameraman into a protest and he'll focus on the craziest sign. Yes, there are racists in the tea party, and they make themselves known. But tea party activists, in most cases, root them out. Texas activist Dale Robertson, who held a sign comparing taxpayers to "niggars" at a 2009 rally, was drummed out of that event and pilloried by his peers. Mark Williams, formerly the bomb-throwing spokesman for the Tea Party Express (he once told me he wanted to send the liberal watchdog group Media Matters "a case of champagne" for calling him racist), was booted after penning a parody essay that had the NAACP pining for slavery.

Liberal critics of the tea party make the case that conservative opposition to social spending is often racially motivated. That's not new, though, and it's certainly not the basis for the tea party.

"Nut picking" has its origin in a 2006 Kevin Drum post of the Washington Monthly as far as I can tell.  To be precise, it refers to the all-too-common practice of trolling the comments of internet fora–what you humans call "blogs"–for the crazies.  One then alleges that the crazy commentor represents a typical view of the opposition.  Therefore, etc., as the medievals would say.  Real logicians call this practice "weak manning" or more technically, "the selectional form of the straw man."  

Now to be precise again (sorry, it's my job), the claim that the tea party is racist might be justified (badly, let's say) in some instances by nut picking, but it is not the same as nut picking.  Nut picking may be one of the many mechanisms used to produce an unrepresentative sample, upon which one then makes an inductive generalization.