Hilarity ensues when Ann Althouse has difficulty understanding the meaning of “granular” when applied to interpreting voter file data. Specifically, she was trying to understand the following:

“If there’s one conclusion that’s going to come out of this process, it’s that we have to be much more granular in our approach to partners in the community like African-Americans, Hispanics and Asians,” said Reince Priebus, the Republican National Committee chairman, who is overseeing one of the most ambitious review efforts.

Here is how the U.S. News described “granularity” in a March 2012 article:

Once a key advantage to Republican campaign strategy, the GOP has doubled down on analyzing voter data in 2012 after Democrats have spend years beating them at their own game.

The Republican National Committee signed an agreement with Data Trust, a firm that maintains voter files and matches them with relevant commercial data.

Data Trust has spent months working out agreements with conservative groups who will work with the company to use and update the voter files.

The move signals the GOP’s renewed commitment to “microtargeting,” a technique that uses granular information on voters to reach out and build relationships with specific demographics.

You can use the data to create an overall message to certain demographic groups, but the more common usage relates to individual voters. How can the Republicans identify African-Americans who might be willing to vote for them? The answer is to look at things like the magazines they subscribe to, the things they like on Facebook, and other kinds of commercial or public data. There are a couple of ways to go about this. You can build a profile of the black voters who the Republicans know support them and use the profile to identify and register more voters like them. Or, you can figure out why black voters are not supporting you and what might make them change their mind, and make either real policy changes, or (at least) better messaging. The same things can be done for Asian-Americans, Latinos, single white women, etc.

The granularity refers to the richness of the data. If you know enough about who your voters are, you can identify people who will vote for you if you register them or contact them, and you can make your turnout effort much more efficient and effective. If you are in the mood to make policy changes, you can identify the policy changes most likely to appeal to specific types of individuals.

A lot of commenters in Ann Althouse’s thread attempted to figure out what ‘granular’ meant in this context, but I haven’t seen so many swings-and-misses since the the last time I watched the Bad News Bears.