Admitting to Entertainment Programming -- Mode Response

We know from loads of previous research that people answer survey questions differently if asked face-to-face as opposed to the phone or, especially, via the Internet. Let's have some fun and look at a few responses between national samples of the same questions asked F2F vs. over the Web, with the data weighted to account for demographic differences between the two samples.

Yes, I'm thinking of a study along these lines with these data. Consider this me thinking out loud.

Below I'm going to toss a few shows at you, with the F2F and Web percentages. My underlying assumption is people will more readily, in a F2F situation, to say yes to more prestigious answers, like news programming, and be less likely F2F to admit to entertainment programming. Let's see if my theory holds water. Below the table, I discuss briefly.

TV Program	F2F%	Web%
News Shows
20/20	24.5	18.6
60 Minutes	28.5	24.5
ABC News	18.9	14.4
Meet the Press	12.3	8.5
NBC News	29.0	22.4
Entertainment
The Mentalist	13.7	16.0
The Voice	15.3	2.3
American Idol	14.3	11.9
Big Bang Theory	24.8	22.0
Dancing w Stars	13.5	14.5

These are selected programs, not all of them available, so a more comprehensive analysis of the data might find something different. Still, you kinda get the idea. Look at the news programming. The F2F percentages are uniformly higher than the web-based responses. And now look at the entertainment-based programming. It's generally higher as well. Two exceptions: The Mentalist and Dancing with the Stars.

So all in all, not a lot of support for my thesis. I'd have to generate a more complete list, create an index of each type of television programming, and then see if it held up. But it doesn't look good.

I hate it when data gets in the way of a good theory.


Methodological Stuff

This is from the 2012 ANES, which used random samples of both face-to-face and web-based surveys. You can find the study page here.