Wednesday, November 22, 2017

Boosting My Ragged Self Esteem

Sometimes to boost my ragged self esteem I like to go online and look at who's citing my work. It requires some effort, mind you, but if I did hard enough via the magic of Google Scholar I can often find a few suckers scholars who pointed to my work while building their own theoretical arguments.

Examples, you ask?

This 2017 study includes a cite of one of my older studies from, wow, 1995. The study here is about need for cognition and political knowledge.

And this 2017 study is about knowledge of nanotechnology and cites my 1995 and 1996 papers.

I could go on, but I'm bored and you're bored and it's a holiday week. I will take a second to point out my most cited works, which when compared to colleagues the numbers are not that great, but I'm the top cited person in my household. So there.

  1. By far #1 cited is my study of young people learning from late-night comedies. It's from 2005 and has triple the cites of my #2.
  2. So, #2 is that 1995 "new news" and knowledge piece discussed above.
  3. Close behind with only two fewer cites than the one above is another piece with a good title and data over time, which helped get it attention.

Fourth and fifth places are talk radio studies. I did a lot of those and they got me tenure. Sixth is maybe my favorite piece, a study of why people believe Obama is Muslim and one of the first of its kind. If I'd published it in a bigger journal is would have gotten a lot more attention, but that's the way it goes.

Monday, November 6, 2017

Slow Going

Due to health issues, I've fallen behind in posting here. Apologies to my tens of readers worldwide. Hopefully I will pick up the pace soon.

I do point to this study that says basically the greater an individual's news literacy, the less likely they are to believe in conspiracy theories. It's main flaw is it doesn't cite me.

Also there's this brilliant piece of research that everyone should read, memorize, and recite to random strangers on the street.

Wednesday, October 18, 2017

About that Alabama Poll

A new poll on the Alabama U.S. Senate race has it a tie between Roy Moore the Republican and Doug Jones the Democratic, both at 42 percent.

Don't believe it.

I'm not raising doubt because it's a Fox News poll. Their shop does good work. It's just that previous polls had Moore with a steady lead. Now maybe that lead has faded. Maybe news about some questionable charity monies or a bejillion Russian bots following him on Twitter have eroded Moore's lead, or maybe it dissipated because some folks realize he's kinda nuts.

Or maybe this poll is an outlier. These happen. Statistically speaking every poll has a 5 percent (1-in-20) chance of just being wrong. That doesn't mean it takes 20 polls for one to be wrong. The first could be outlier. I think it may be an outlier because all the other polls show Moore with a lead. For example, one earlier this month had Moore with a solid 8 percentage point lead.  Every public poll previous to this latest one had Moore ahead by 6 to 8 percentage points.

Unless another poll validates the latest result by showing it a lead or within the margin of error, I wouldn't make much of it. I get why Alabama Dems like this poll result and how they'll probably use it to raise money and energize their base, but as an observer I just don't think it's right. I've lived and worked in Alabama and I know you'll never go broke betting on the Republican candidate, even the crazy Republican candidate.

As an aside I did take a few minutes to dig into the crosstabs to see if anything odd shows up in this latest poll to explain the tie. Sometimes you'll find something there, such as more respondents of a particular socio-demographic group included than you'd normally see, thus skewing the results one way or the other. For example, more educated respondents than normal would favor Jones and more white evangelical respondents than normal would favor Moore.  Nothing jumps out at me, but then again I don't have baseline numbers to work with. We'll just have to wait for another poll to see if Jones has indeed clawed into a tie with Moore.

Friday, September 22, 2017

Journalism 101 -- We are NOT the story

I'm a journalism guy, but I firmly believe we should go out of our way to not be the story. It was pounded into me in j-school, it was pounded into me by my first editor at a small but really good Mississippi daily newspaper. I still believe it, I still preach it. Yes, there are exceptions, but our audience wants to know about the story, not about us. Plus it makes us look like snowflakes.

So we have this Atlanta TV station story about the bonding out of three students who protested the Georgia Tech police shooting of a troubled student. Watch the video yourself. I don't think the hed on the web site seen to the left about protester curses the media is really the story here. It's a basic making bail story, the kind we do all the time, but him tossing a few expletives in a journalist's direction is hardly news. Jeez, if I made a story about every time I'd been cursed at as a reporter I'd never have time for real news.

Watch the story, decide for yourself if this is the hed that really deserves to be on there. I suppose it is clickbaity because, face it, we all love seeing those damn journalists getting cursed at. But I'm not a clickbaity guy.


Monday, September 18, 2017

Strange Alabama. Tell me Moore

As political junkies know, there's a fascinating U.S. Senate race going on next door in Alabama between Roy "Ten Commandments Everywhere" Moore and Luther "Trump Likes Me Best" Moore. The election itself is Sept. 26, but there's been a spate of polls showing either Moore winning big or the race a statistical tie. No polls seem to show Strange ahead. This bodes well for Moore and, of course, poorly for Strange.

A poll out today suggests the race is tightening. Here's the opening grafs. Keep in mind that tightening in this case still has Moore thumping Strange.
U.S. Sen. Luther Strange has made significant cuts into Roy Moore's lead in a Republican primary runoff poll but still has a lot of ground to make up over the next eight days. 
In a poll released Monday by Louisiana-based JMC Analytics and Polling, Alabama's former chief justice holds an eight-point lead over Strange. 
The same polling firm last month had Moore with a 19-point lead.
It's helpful when the same firm, using the same method, shows change. It's more difficult to compare across polls, especially when they use very different sampling or weighting methods, and that assumes they report any of these details at all.

A few days earlier a different poll essentially said it's too close to call. Keep in mind this particular poll has ties to those supporting Strange, but I'm certain it's a mere coincidence they call it a tie.

A summary of polls by Real Clear Politics paints a pretty picture for Moore.

President Donald Trump and the GOP establishment have all endorsed Strange, as Moore infamous for his on-again, off-again career on the state supreme court. In this case, Trump's endorsement doesn't seem to matter.

Maybe it will tighten some before the election. That's what usually happens in most races as undecideds, well, decide. Those may break more for Strange but it's hard to imagine enough of them remain out there to change the outcome.


Tuesday, September 5, 2017

More First Names

I wrote yesterday about the incoming UGA freshman class and the most popular first names. This is Part II in which I provide the data of the top five names since 2008. See the table below. A few years have ties for a spot.


08
09
10
11
12
13
14
15
16
17
1
Sarah
Sarah
Emily
Emily
Sarah
Emily
Emily
Emily
& John
William
William
2
Emily
William
Elizabeth
John
Emily
Sarah
Sarah

Emily
Sarah
3
Lauren
Emily
Michael
Sarah
John
John
John
Hannah
Sarah
Madison
4
William
Elizabeth
Taylor
Andrew
William
Anna
William
William
Caroline
Emily
5
Michael
John
William
Michael
Rachel
Hannah
Rachel
Anna &
Caroline
John
Caroline



As you can see, for a long time Emily and Sarah dominated and even in 2017 they're in the top five, but William stepped up. Why is that? Good question, as UGA tends to have more female than male students. Keep in mind the differences in raw numbers are slight. See below for the top five in 2017 and the numbers of students who share that first name.
  1. William (67)
  2. Sarah (64)
  3. Madison (59)
  4. Emily (57)
  5. Caroline (56)
Think about those numbers above, as they're out of 5,482 or so freshman in this year's class. So all it takes is a statistical quirk of a few more students who share a name to bump it up the list. You really can't make too much out of this, though that doesn't stop me from trying every year, just for fun.












Most Popular First Names

Every year since 2008 I've collected the first names of UGA's freshman class and have then written about it. Why? I have too much time on my hands, that's one. Two, it's a fun exercise to show students, to demonstrate you can find story ideas in odd places.

Well, it's time for the 2017 update.

First, a bit of history. Since 2008, the most popular UGA freshman name has been either Sarah or Emily, but in 2016 we saw a shift to a new leader -- William. And William is also the most popular freshman name in 2017, at least at UGA, at least among freshmen. Here's a quick wordcloud of the 2017 names:


Here are the Top 10 for 2017, with the 2008 rank in parentheses

1. William (3)
2. Sarah (1)
3.  Madison (235, wow)
4. Emily (2)
5. Caroline (23)
6. Hannah (27)
7. (tie) John (15) and Anna (33)
10. (tie) Matthew (6) Jacob (39) and Andrew (21)

The big move up is obvious -- Madison. Ranked only #235 in 2008, now it's #3. That said, you have to keep in mind the actual numbers here are relatively small between names. Number 1 in 2017 was William, with 67 of them. Sarah had 64, Madison 59. Back in 2008, there were only three named Madison, so that's something. 

To really have fun with this, I'd need to dip into the Social Security baby name data for the year these students were likely born, see if there was a burst nationally or even in Georgia of kids named Madison compared to earlier years. I can say, based on a cursory look, William was the most popular boy name in Georgia in 1999, roughly when these students were born, and Madison was the 4th most popular name. If we go back 10 years, Madison is the #1 Georgia girl's name, so I can't really explain the boomlet in Madison at UGA. It will take some more data digging.