Monday, January 10, 2011

Curing Myopic Voting

It was a snow day, so I had time to scour the net for -- to me -- obscure research.  This draft of a study examines how people consider the economy in their voting.  The research itself is a bit complicated, but basically the author argues people over place too much weight on recent economic information when voting and fail to consider the cumulative economic data.  No surprise there.  This "end bias" makes sense, a recency effect of sorts, but it is a bias -- a failure to consider everything and an over-reliance on what's most recent.

Or, as the author notes:
Using a three-year panel survey, I showed that citizens' memories of the past economy are inconsistent with their actual experience of the economy as they reported it in earlier panel waves. They fail to remember the past correctly in part because the present shapes their perceptions of the past.
How do you fix this?  Not easily.  News reports, obviously, could focus on the big picture and not the most recent data, but you can be damn sure partisans will focus on whatever numbers make their candidate look best, or the other candidate look worst. This seems to be the cure, says the author.
Presenting participants with cumulative information on performance (e.g., total income growth) cures myopia. On one hand, these results are troubling for democracy because they confirm citizens’ incompetence at retrospection. On the other hand, they point to a remedy, one that candidates and the news media could adopt.
Using this study, imagine the 2012 election.  Let's say unemployment is down a bit, but is still too high.  The economy is doing okay, but not great.  If journalists focus on the long-term picture, the improvement, that's a more reasonable economic approach and one less "myopic" -- to use his term.  That kinda helps Obama.  But I don't see Fox News playing it that way.  And you can certainly argue that it should be better than it is, even if there is a slight improvement.  So in this hyper-partisan times, I'm not sure the "cure" for myopic, short-term voting, will find its way to the public.

Friday, January 7, 2011

Scary Twitter Research

As the one or two people who actually check out this blog know, I spend a lot of time scouring the Net and academic journals for appropriate stuff to talk about.  Well, I hit the jackpot on this one -- the scariest Twitter study ever imagined.  Why scary?  Because it goes so far in compiling data even I, a numbers geek who loves data analysis, finds this one a bit too much to contemplate.

Lemme hit you with some of the abstract below.  Warning ... it's dense.  Just skim it and you instantly see we're talking a whole different world of data gathering.

We have crawled the entire Twitter site and obtained 41.7 million user profiles, 1.47 billion social relations, 4,262 trending topics, and 106 million tweets. In its follower-following topology analysis we have found a non-power-law follower distribution, a short effective diameter, and low reciprocity, which all mark a deviation from known characteristics of human social networks [28]. In order to identify influentials on Twitter, we have ranked users by the number of followers and by PageRank and found two rankings to be similar. Ranking by retweets differs from the previous two rankings, indicating a gap in influence inferred from the number of followers and that from the popularity of one's tweets. We have analyzed the tweets of top trending topics and reported on their temporal behavior and user participation. We have classified the trending topics based on the active period and the tweets and show that the majority (over 85%) of topics are headline news or persistent news in nature. A closer look at retweets reveals that any retweeted tweet is to reach an average of 1,000 users no matter what the number of followers is of the original tweet. Once retweeted, a tweet gets retweeted almost instantly on next hops, signifying fast diffusion of information after the 1st retweet.

To the best of our knowledge this work is the first quantitative study on the entire Twittersphere and information diffusion on it.

Holy quantitative hell, Batman!  This scares even me, an SPSS guy who never met a multiple regression he didn't enjoy running twice, just because he can.  But these South Korea guys?  I am humbled.

Okay, but what does it all mean?  I dunno, beyond the sheer amount of information fed into Twitter and then spread, byte by byte, out into the Twittersphere.  For those who specialize in Twitter, this study not only reinforces your argument that the social networking site matters, but gives a sense of how much it matters. But I do find it fascinating that 85 percent of Tweets are news of some kind. 

Wow.  I'll never make fun of Twitter again.

Okay, yes I will.

The News, Cancer, and a Counter-Intuitive Result

It's a given -- we learn a lot about disease and illness from the media, either in the news, or from those annoying pharmaceutical commercials, or in entertainment programs.  Me, I get all my medical knowledge (and behavioral modeling) from House.

Cancer in particular gets lots of media attention, as it should given the number of people who suffer and die from the disease every year.  But is it accurate media attention?

This study suggests not, at least when it comes to the news.  Only the abstract is available (I bet Gregory House gets access for free).  But what's really interesting is not that the news media isn't as accurate as it could be -- surprise -- but the direction of that inaccuracy.  Stay with me here.  It's kinda sorta fascinating.

The content analysis of cancer coverage by eight newspapers and five magazines reaches a counter-intuitive conclusion.  News reports often discuss aggressive treatment and survival, according to the study, but rarely touch on the bad news.  All that emphasis is mine because I love doing it.  Think about it.  How's this result for turning the traditional criticism of journalism on its ear?  We (the royal we, as in journalists) focus too much on the bad, or so goes the criticism by just about everyone on the planet who isn't a journalist, but according to this study, the news "may give patients an inappropriately optimistic view of cancer treatment, outcomes, and prognosis."

Wow.  Journalists optimistic?  Something's deeply wrong here.

Unfortunately we can't judge the quality of the research, given all we have a measly abstract.  And when non-media scholars attempt a content analysis, they often screw the pooch -- as in no inter-coder reliabilities, no cleanly drawn categories, no sense of the difficulties of analyzing a text or constructing units of measurement.  Sure, it's The Archives of Internal Medicine, a pretty damn good medical journal, but too often medical folks conduct really bad behavioral research.  For them, a media study can seem like slumming it.  That attitude often shows in the work.

But if we assume for the moment the study was done halfway well, then this result does surprise. Clearly someone in journalism is asleep at the switch.  Get back to that bad news!

Thursday, January 6, 2011

The Constitution and the Nutjobs

As the U.S. Constitution was read today in the House, during the part that says "no person, except a natural-born citizen, or a citizen of the United States" is eligible for the presidency, a woman yelled out that it didn't apply to President Barack Obama.  Yes, the wingnuts and nutjobs are out.  Story here.

Move Over TV News

A new Pew study shows that, among college grads, the Internet has begun to rival television news as a source of information.

First off, let's establish that a historical point -- the greater the education, the less the reliance on TV for news and information.  It's that whole delusions of adequacy thing.  The audience for TV news, local especially, tends to skew to the less educated.  So the Pew report is not all that surprising, but it does document quite nicely the growing power of the Net when it comes to how the educated keep informed.

Blogger is in one of its moods, so I'm having trouble dropping in a key graphic from the report.  Here's a few key numbers.  For those with a high school education or less, 75 percent said they rely on TV news and 29 percent said they rely on the Internet.  For those with a college degree, the same numbers are 54 and 51 percent.  That's huge, and telling.  And problematic for the TV folks. 

What I find interesting is the lack of a partisan effect.  Dems, GOPers, Independents, they all break down more or less the same in the aggregate media use columns.  Now if we get into the nitty gritty, not provided here, we'd find the usual partisan migration to Fox or MSNBC, but it's interesting as hell to see that at the aggregate media use level, your partisan leanings don't seem to correlate with a general preference for TV news versus that of the Net or other sources.

In terms of what people know, we can make a few assumptions based on these results.  First, TV news appeals to the less educated not because TV news is dumb, but because it's more easily consumed, provides less detail, but makes the world's events more understandable given the way TV tells a story.  We know from other studies that TV news seems to inform only those with little or no information in the first place -- which is a fine and useful mission, though it might raise certain issues when it comes to selling advertising that appeals to a prime audience.  Print, especially, seems to require a lot more background knowledge and certainly motivation in order to grasp the content. 

What we may see, as this migration among media continues, is a growing knowledge gap between those of less and more education. 

Wednesday, January 5, 2011

Reading the Constitution

Consonants and vowels alike have tragically died as people write about the U.S. House of Representatives and its reading aloud of the U.S. Constitution -- for apparently, as best we can tell, the first time in history.  That alone is kinda interesting.

But whether this is a cynical gimmick or meaningful gesture, it probably depends in part on your political predisposition.  I'd side with gimmick, if only because the document itself is dense, disconnected from everyday policy, and will have little or no informative or educational benefit.  In other words, people won't suddenly learn anything by hearing it read aloud.  There will be no "Oh my God" moments, assuming anyone can stay awake long enough to listen to the thing.  It's a brilliant piece of writing and political thinking, but in terms of suddenly having people think a certain (or any) way about democracy and Congress and policy and laws?  No, no effect at all.

Monday, January 3, 2011

Increasing Trust Through Knowledge

Natural experiments are terrific -- if you can plan ahead or have the resources to quickly take advantage of one as it occurs.  This study does just that, finding that people who received information from the Social Security Administration gained more knowledge and had more confidence in Social Security than people who did not receive the information.  And this includes statistical controls for such predictors as age, education, race, and gender. 

The study, published in a recent The Journal of Politics, includes a neat structural equation model for those who like pretty pictures and smoke-and-mirrors statistical analyses to arrive at an interesting conclusion, that knowledge leads to confidence.  And not the other way around. 

Oh, and Happy New Year.