Tag: research/Archives


CPB logoCPB’s Ted Coltman sent some pubcasting audience/contributer figures to the Pubradio maillist. It’s the first time I saw this data so concisely assembled. So, with his permission:

In FY 2009, TV and radio station grantees reported to CPB (on their Annual Financial Reports) a total of 5.6 million individual contributors (either to stations or to their associated “Friends” groups). There is undoubtedly some double-counting of individuals in that number (because some individuals give to more than one station), but CPB has no way of “de-duping” these data.

In validating a rule-of-thumb such as “only about one listener in 10 is a member/contributor,” it makes a big difference whom you count as a “listener” — anyone, for example, who listens at least once a month?… or only someone who listens at least once a week?

Public radio’s weekly cume audience is about 30.6 million persons, so the roughly 2.5 million contributors reported by public radio stations would be about 8 percent of the weekly cume listeners (roughly 1 in 12). Its monthly cume audience is estimated at about 64.7 million persons, so the reported contributors would represent about 4 percent of the monthly cume listeners (about 1 in 25).

Similarly, public television’s weekly cume audience is about 60.4 million persons, so the roughly 3.2 million contributors reported by public TV stations would be about 5 percent of the weekly cume viewers (1 in 20). And public TV’s monthly cume audience is about 121.9 million persons, so the reported contributors would constitute about 2.6 percent of the monthly cume audience (1 in 40).

Of course, the “unit of giving” is often a family or household, and the “units of listening or viewing” I just cited are individual persons (aged 12+ in the case of radio, or 2+ in the case of TV), so these ratios shouldn’t be considered very precise or reliable… but then neither are most “rules of thumb.”

Data sources: CPB’s ISIS database, RRC, Arbitron, PBS Research, Nielsen.
See 170 Million Americans for Public Broadcasting: The Numbers.


You just got back from checking RealClearPolitics polls, didn’t ya? What’s that, like the fifth time today? I know, bro, I’m there w/ ya.

That’s why I’ve resolved to drastically change my daily routine. No, not by mindlessly checking polls any less, but rather by adding another site to my hourly obsession list: Five Thirty Eight.

These guys are baseball stats folk who, for the last few months, have turned their mighty number-crunching powers from earned-run averages and stolen bases to Electoral Votes and Senate races.

Like RCP they average all the recent polls. But to 538 all polls are not reported equal: “we assign each poll a weighting based on that pollster’s historical track record, the poll’s sample size, and the recentness of the poll.”

Their result is not the typical whose-up by how-much number, but instead their more baseball like Win Percentage pie chart (mostly blueberry right now, with a small slice of cherry). 538’s Super Tracker graph plots trendlines, with points for daily averages. An Electoral Vote Distribution chart uses 10,000 voting simulations to outputs the probability for a range of possible outcomes, based on daily data. And their Senate Projections are likely the best in the biz.

Why the name 538? It’s the number of electors in the electoral college.

Real-time Web Monitor

ScreenshotAkamai has a Real-time Web Monitor tracking “global Internet conditions around the clock.” Areas w/ highest traffic are brightest. You can also color the map by areas with the slowest connections (latency) or the most recent “network attacks.”

International Podcastersurvey

Researchers at Jacobs Univerity at the University of Technology- Berlin have published a International Podcastersurvey (pdf) of 1000+ podcasters, with results such as:

    Quantify motivations

  • Podcasting is not mainly a techies‘ affair
  • It‘s more about identity and relationship management
  • But most of all, it‘s about information and expression of opinion

RealClearPolitics – Polls

RCP logoPolls of Presidential voters are remarkably accurate. The day before the 2000 Iowa Democratic Primary, the MSM News folk were still reporting Dean as the frontrunner he’d been for months, despite the several polls that showed a last-minute turn towards Kerry. The news was wrong, the polls right; Kerry won.

Then in November the Voter News Service was criticized for picking Al Gore the winner of the 2000 election. But it turned out (years later) Gore did win. Again, the polls proved correct, even when the margins were so slim.

Most media folk, including myself, are notoriously pitiful at prognostication (hell, we’re not even any good at evaluating the present), so you really shouldn’t pay us any mind when we mindlessly try to predict wuz gonna happen. But do pay attention to those polls we pay for, cuz when several say the same thing, it’s likely an accurate reflection of reality.

All a long way of saying, check: RealClearPolitics – Polls. It’s an up-to-date listing of the major Election 2008 polls, averaged and listed individually. Then click their National Head-to-Head Polls and see who’d beat who if the prez were picked today. (Answer: Hillary’s kickin’ ass.)

This American Life: 12-Year Survey

What we thot was a radio series turns out to be a research project: This American Life Completes Documentation Of Liberal, Upper-Middle-Class Existence.

Says senior producer Julie Snyder: “There is not a single existential crisis or self-congratulatory epiphany that has been or could be experienced by a left-leaning agnostic that we have not exhaustively documented and grouped by theme.”

So, where’s that leave pubradio when it’s hippest hour is now Onion fodder?

Grading GWB

Newsflash!!! Hold the presses. This just in: we now have incontrovertible evidence that Republicans approve of our Prez, Democrats don’t, and Independents aren’t sure. It took the resources of both Slate Magazine and researchers MediaCurves to divine this world-shaking new data; ie, here’s how researchers spend their and their focus group members’ time: