Friday, April 29, 2016

Why a Journalism Major?

A study in the latest Journalism Educator (yes, there is such an academic journal) tests a scale designed to explain "the underlying reasons why undergraduate students seek a degree in journalism," what the authors call the Journalism Degree Motivations scale.

Keep in mind this scale is not a done deal. It's a first shot at validating what they think may be a good measure of the motivations behind signing up as a journalism major. Read the article if you're into methodological details. They look reasonable to me.

Essentially the authors use factor analysis to examine a bejillion questions to see which ones fall together as factors. This is a common statistical technique and eventually can lead to reducing the number of questions to measure the factors that seem to emerge from such validation testing.

Okay, let's get to the good stuff. Why do students decide to major in journalism? What makes 'em tick?  Factor 1 -- they couldn't get into business school (just kidding). The authors found eight factors, none of which include business school. They are, in order of explanatory power:

  • Social Responsibility -- made up of questions like "I want to hold public officials accountable," this one doesn't really surprise you now that it's presented. J-students should be driven by this.
  • Reporting Skills -- This has more to do with enjoying the chase of the story, interviewing, and the like. It doesn't mean they have those skills, it means they enjoy learning them and using them.
  • Social Prestige -- Sigh, students grossly overestimate this one.
  • Sports Media -- This is interesting. An interest in sports is a major factor in becoming a journalist. This bodes well for my school, UGA, where we have two faculty devoted to just teaching sports journalism and a popular certificate program in the discipline.
  • Visual -- Mostly photojournalism and video, this one has to do with folks interested in such visual skills.
  • Writing -- No surprise here, except that I would have expected it to be higher.
  • Varied Career -- This one is interesting enough to revisit on another post. That last thing millennials want is to be bored, and journalism can be boring, but it's definitely varied as well.
  • Science & Numbers Anxiety -- Another one that deserves its own post. Basically a few questions on this found students majored in journalism because they don't want to learn numbers or science. Sorry, darlings, but never have such skills been so necessary in journalism. You should have been j-majors in the 1950s.
For the statistically interested, I recommend going to the article and checking out the Eigenvalues and other methodological notes. 

Wednesday, April 27, 2016

Measuring Media Consumption

Ah, we return to the blog's original purpose -- discussing research. In this case, research on how to measure people's media consumption, a kinda important thing if you happen to be in the media business (academic, or otherwise).

This new study examines one technique. I provide the abstract below because, dammit, even on my university computer I can't access the entire article. Yeah, BS. But still, let's look at what we have and I'll try to break it down for you best I can.
How to measure exposure to information in the media is highly disputed, primarily due to the difficulties of obtaining accurate self-reports. The growing supply of outlets and proliferation of information sources have added an additional level of complexity to these problems. Reflecting on old and new approaches for measuring exposure to political information, it is argued that both the specific source and the frequency of exposure must be taken into account. The validity of this so-called “list-frequency technique” is tested using a two-wave panel survey as well as a split sample experiment from the survey pre-test to enable comparison with the “list technique.” The results support the list-frequency technique in being a good solution, since it provides the same aggregate estimates of media use as the already validated list technique, and may give more detailed effect estimates and increase the explained variance when predicting political knowledge.
Let's do it this way, by taking key concepts one at a time.

  • Self-reports. Most of our survey work depends on self-reports of news consumption. You know, questions like how many days a week do you watch television news. Or how often do you watch Fox News. Or something like that. But as House often tells us: "People lie." They lie about how often they vote, how often they attend religious services, and they lie about the media they consume and how often they consume it. Or if they don't lie, they misreport in a fashion that makes them look better. In other words, these measures are full of error.
  • List-Frequency is just what it sounds like, a technique in which people tell us what media they consume and then how often. Often this is more specific than our apriori questions that focus on a generic term, like television news, or even specific networks or programs.
  • The down side? It adds time and complexity to a survey instrument. But the method, according to the abstract, provides more explanatory power when predicting what people know.

For $41 I could read the whole article. Yeah, right.

Friday, April 22, 2016

Christian-like Correction

So there was an odd story I came across Thursday, one that kinda threw me because I'd heard nothing about it. See the screen grab below. Note the hed: University of Georgia. Same in the lede. But the dateline is all wrong, and later the story it suggests it's University of North Georgia.

Confusing, right? I commented below the story asking for a clarification because if this happened at UGA, the local news media missed it. I also sent an email to the editor. I never received a response to my email and, interestingly, my comment on the story was removed. BUT ... the story was corrected, with no notice of a correction being made. Here's what you see today. Note the change in hed and lede.

Apparently it's not Christian to acknowledge corrections and tell your audience you screwed up. C'mon, folks, it's Journalism 101 to put a correction notice. You know, confess.

Thursday, April 21, 2016

UGA Open Records Requests

I've written on this earlier, but I now have cleaner and more comprehensive data on the kinds of open records requests UGA receives and how many -- as seen by a new law in Georgia -- involve athletics. Below I discuss not only how many requests were made, and not only how many of those involve sports, but also who made the most requests from UGA.

As we all know, the governor signed into law a 90-day (versus the usual 3-day) time period for athletics departments in Georgia universities to respond to the public's right to know. Yes, the public can know ... eventually. In 90 days. Or longer, really, a law written mainly because sports journalists wanted to know how much money the coach was using to fly here and there recruiting.

What I'm presenting here are all open records requests made since August 1, 2015, to late last week. Below are the basic results. As you can see, sports-related requests make up over half of the requests (54.7 percent).

205 (54.7%)
170 (45.3%)

To be honest, the proportion of sports-related open records requests would be even higher if not for a controversy involving an employee on campus and how UGA bungled her case. There were quite a number of requests on that and, taken out of the mix, sports would dominate even more.

I didn't take the time to break down the non-athletics requests. A lot are routine stuff, like the number of bee keepers in Clinch County (yes, really). A lot are from lawyers niffing around for info for their clients (some interesting stories there, budding journalists, if you just take a closer look). There's even a helicopter mom asking why her kid didn't get accepted into UGA.

Top Requesters

The news media make far more requests than anyone else and especially in sports. Below you'll see the Top 10 Requesters from UGa in the last eight or so months.

# Requests
Marc Weiszer
Robbie Burton
Seth Emerson
Chip Towers
Lee Shearer
Jason Butt
Kevin Kelley
Kale Duffell
Rebecca Burns
Dan Eveloff

By the way, I'm #11 with four requests. Fear me.

Most of the folks above are sports or news, with Weiszer by far leading the pack. Indeed, 17.3 percent of all requests were made by him. Maybe they should have named the new 90-day law after him?

Monday, April 18, 2016

Learning from Shared News

We all realize by now that many people get their news via social networks, largely Facebook. This study looks at how sharing is related to two types of knowledge: factual and structural. First, the definitions. By factual knowledge, the authors mean exactly what you think they mean, the ability to correctly identify bits of information. "Structural knowledge, on the other hand, is conceptualized as being able to see the connections that exist between related concepts."

The first is easy to measure, the second a bit more of a challenge.

By knowledge structure density, they asked respondents how related they see two concepts on a seven-point scale until all possible combinations of the five items had been used -- Medicare, taxes, politics, economy, and debt.

Viewing news online, they found, is related to factual knowledge but not structural knowledge. But sharing is just the opposite, more related to structural as opposed to factual knowledge.

Interesting. Still trying to figure out exactly what this means in both a theoretical and practical sense.

Friday, April 15, 2016

UGA's PR Problem

I've noticed since the hiring of UGA's new vice president in charge of marketing and some such, all we see in news stories these days is an email response to journalists' questions.

What utter bullshit. I'll explain why below. First, and most recently, this story on faculty salaries is a good example. Near the bottom, we get this tripe (sorry for its length, this is how my local paper ran it online):
Michelle Garfield Cook, UGA’s associate provost for institutional diversity, defended the ERS study in an email.

“The university selected an independent consulting firm with more than 30 years of experience working with universities. Using a qualified, third-party consultant prevents potential conflicts of interest,” she wrote. “ERS Group applies an industry-standard methodology and has prepared pay equity studies at research institutions including Stanford University, Virginia Tech, the University of Alabama at Birmingham, Auburn University and the University of Kentucky. The study took into account factors that legitimately impact compensation, including academic discipline and rank as well as tenure status. Other factors that legitimately impact compensation include education level, work experience and administrative assignments, among other factors. The university is committed to gender equity and intends to review faculty salaries on a regular basis, at least every five years.”
Oh Hollander, you say, this is just one case, so stop making a big deal out of nothing. A couple of days earlier, more and similar tripe in this story. Sorry, again this is how the paper has it online. It doesn't say it's an email, but it certainly reads like one:
Karri Hobson-Pape, UGA’s vice president for markeint and communications, had this comment on the pay numbers:

“The trends you note are simply due to the recession period when our longer-term UGA faculty members could not receive merit raises,” she wrote in an email. “This is a challenge we are now addressing through three consecutive annual faculty raises generously permitted through the Board of Regents. In addition, we are also now in a position to hire incoming faculty at more competitive market rates. President Morehead and Provost Whitten share a strong commitment to increasing faculty salaries; it has been and will continue to be one of their top priorities.”
It's a basic journalism rule that you don't give sources the questions in advance. That's what an email does, gives academic bureaucrats the questions in advance, allows them to craft meaningless, stiff, bullshit responses, and there's then no opportunity for a follow-up question to point or quiz them on inconsistencies or to challenge the statements.

It's a PR flak's wet dream. It's a journalist's nightmare. And the public loses in the process.

So either UGA has decided its people cannot handle telephone or face-to-face conversations, or the local paper is not willing to challenge the administrators. Here's how I would handle this. I'd have a sit-down with the VP of Whatever and explain this is unacceptable, that either you respond to questions in a live setting or we'll simply run the stories without your carefully crafted crap, because we're not gonna put up with this bullshit email thing. And I'd put in every story that the university refused to respond to questions unless provided in advance. I'd put it in a friggin box. In bold face. Like this.

Yeah, that's how I'd handle it.

Thursday, April 14, 2016

Georgia's Least Drinking County

As my ones of regular readers know, I've been fiddling with health data of all counties in the U.S. One of the more interesting measures in the data is a favorite topic of mine -- drinking. In the whole country, the teetotaler winner is Plute County, Utah, coming in at #3,140 when it comes to excessive drinking.

But don't give up, Georgia fans.

Turns out Taliaferro County, Georgia, does pretty well with a tie at #3,120 (Quitman County, Georgia, is tied at #3,116).

So why Taliaferro? It's in the map, the red blip, and is the least populous county in Georgia and, according to the magic of Wikipedia, the second least populated county east of the Mississippi River. The 2010 Census had all of 759 households. Hell, there are neighborhoods with more homes.

What else stands out about this county that might explain this lack of adult beverage consumption? Oddly it scores high in number of mentally unhealthy days, with a rank of #84 about of over 3K counties in the U.S. But, it has a lot of poverty (49 percent of kids in poverty) and 43 percent of the households are single-parent. You might think those are good arguments for drinking to excess, but apparently not. It also ranks fairly high on obesity (#679 nationally, with over a third obese.

All in all, no easy answers to explain why.

Monday, April 11, 2016

Sports -- Now a Georgia State Secret (for 90 days)

Georgia Gov. Nathan Deal signed into law today a bill that makes college sports a state secret not for the normal three days for a public record, but 90 days.

Why? Who the hell knows.

As the AJC reports:
The legislation, Senate Bill 323, allows the athletic departments at UGA, Georgia Tech and other state colleges to wait 90 days before responding to Open Records Act requests. Athletic associations, like all state agencies, previously had three days to acknowledge the requests.
Yeah, it's a stupid law, but I decided to dig in my data to see the complaint about so many open records requests was a valid one. Keep in mind that 90 days doesn't mean you'll have fewer records requests, it just means you'll have more time to keep the public from knowing the public's business.

I have a dataset of UGA open records requests from the 2014-2015 academic year. There is no date on them, so I can't be sure if it's all the requests. Still, let's look at the 102 requests made of UGA during this time period. According to my count:

  • 66 of 102, or 64.7 percent, were about sports. Nearly two-thirds of the total.
  • 14 of them, or 13.7 percent, were about administration.
  • 10 of them, or 9.8 percent, were about research.
  • The rest are a scattering of stuff (business, law, and undetermined)
So what can we tell from this? Sports reporters may have created their own mess by, perhaps, overindulging in open records requests. Or maybe news reporters are not doing enough. Below are two examples of the sports requests. Apologies for the ugly cut-and-paste job. I'm rushed.

(1) All records mentioning or referring to Ketorolac or Toradol within the Athletic Department since January 1, 2005 (2) All emails to/from Ron Courson, Steve Bryant, Sean Boland, Mark Christensen, Fred Reifsteck, and Ron Elliott mentioning or referring to Ketorolac or Toradol since January 1, 2005.

(1) Any new football or men's basketball schedule contracts or agreements reached or past contracts changed in recent weeks. Please include any games scheduled for men's basketball in 2015-16 and beyond                    
(2) Any changes to the total compensation and/or contracts to head men's basketball coach Mark Fox or any football coaches in recent weeks. Please include any changes to contracts or new MOU                                      
(3) Any NCAA violations involving Georgia reported since Dec. 15. Please include any communications between Georgia and NCAA's enforcement division             
(4) The name and position of any new employee hired or employee who has left Georgia or moved into a new position since Dec. 15

You get the idea above. A lot of requests for contracts, of course, both for coaches and specific games (as in how much UGA is paying for some scrappy little team to visit the stadium and get slaughtered).

Do I really blame sports journalists for doing their jobs? Of course not, though if you read through some of the requests they're kinda broad. Still, a public university's sports program is a public business, and the public deserves to know, in a timely manner, what it's up to. Ninety days is not timely. Not even close.

Mentally Unhealthy Days

In playing with some county-level health data, I decided today to briefly talk about mentally unhealthy days. It's exactly what it sounds like, the number of days in a month you felt, well, mentally unhealthy -- at least according to surveys.

Yeah, for some of us that'd be pretty much every day.

The national average is 3.7 days a month. So who boasts the most nutso days? Sigh, you guessed it. It's the South.

Of the Top 10, five are in Alabama -- surprising no one. The county with the most mentally unhealthy days is Greene County, Alabama, with 5.6. Interesting factoid, that county is also #3 nationally in obesity. There is a correlation there, r = .44 in all U.S. counties between obesity and mental health, for you statistical nerds out there, so maybe they're related. Curiously, there's a negative correlation between excessive drinking and bad mental health days (r = -.64). I don't know what to make of that. Without booze, every day is a bad mental health day.

(Also keep in mind that respondents in the second place county, Sumter, reported 5.5 days, so we're not talking about a big difference and likely one within a statistical margin of error. Still, let's have fun with data).

You have to go all the way to #10 to find a non-southern county, that of Menominee, Wisconsin. By the way, the first Georgia county is Clay, coming in at #15.

Nine of the top 20 are in Alabama. Yes, I'll say it again -- surprising no one.

So who are the mentally healthy? Glad you asked. For some odd reason, South Dakota dominates with nine of the Top 10 (the other being a North Dakota county). Indeed, other than a Minnesota county slipping in, all of the top 25 or 30 or so come from the Dakotas. This makes me wonder if there's some measurement issue here in how the data were collected.

Below, the data aggregated by states

Most Bad Mental Health Days

1. Alabama (4.79)
2. West Virginia (4.70)
3. Tennessee (4.68)
4. Arkansas (4.36)
5. Oklahoma (4.34)

Fewest Bad Mental Health Days

1. North Dakota (2.68)
2. South Dakota (2.72)
3. Nebraska (2.78)
4. Minnesota (2.83)
5. Hawaii (2.90)

Thursday, April 7, 2016

Walking to Work

Not many of us live close enough to walk to work, but in Georgia I found an interesting factoid. Below are the Top Ten Georgia counties in terms of percent who walked to work (percent who did so in parentheses, data 2010-2012).
  1. Calhoun (19.5%)
  2. Towns (9.4%)
  3. Chattahoochee (6.9%)
  4. Turner (6.1%)
  5. Clarke (5.7%)
  6. Wilkes (5.7%)
  7. Lumpkin (4.7%)
  8. Pulaski (4.2%)
  9. Liberty (4.1%)
  10. Dade (3.9%)
What's striking is Calhoun County. I see no obvious reason why this speck in the southwest corner of the state should be so much higher than the rest of the state. Sure, there's only one red light in the county, or so says the source of all knowledge -- Wikipedia -- but still, that's a hell of a difference compared to all the rest of the state.

Clarke I expected to be high, perhaps even #1, but it's a measly #5. Oconee is #112 with 0.9 percent of commuters who hoof it. As you can tell from the map below, there's no geographic rhyme or reason behind the data, with darker colors meaning a greater percentage of walkers.

Data Source Here

Monday, April 4, 2016

Start Dates

I wrote yesterday about the nonsense early starts of UGA. In Fall 2016 (this coming fall) it'll be August 11th. As if that wasn't early enough, in Fall 2017 it's proposed we start on August 10th. Both are, stupidly, Thursdays. Both are records.

OK, what about other schools? Looking at just the 2016-2017 academic year, here are some comparisons of start dates:

University of Florida: 8/22 (a Monday, reasonable)
University of North Carolina: 8/23 (a friggin Tuesday?)
University of South Carolina (8/18, an ... ugh ... Thursday)
University of Tennessee (8/17, a Wednesday?)
University of Alabama (8/17, sigh, another Wednesday)
Auburn University (8/16, a Tuesday)
LSU (8/22 (another reasonable start, a Monday)

I could spend more time on this, but we can see there isn't a single school above that starts anywhere near our August 11th start for Fall 2016. A few have reasonable days as well (UF and LSU on Mondays). No one has a Thursday start. I could do a bigger search and look at UGA's aspirational schools, but some are not in the South and other regions have different traditions, so I used nearby universities to keep it more of an apples-to-apples, oranges-to-oranges comparison.

New UGA Record (sigh)

That's right, folks, UGA has proposed a new and unfortunate record -- the earliest Fall start date ever, or at least as far back as I can find. Important point, this is for Fall 2017, not this coming Fall.

August 10. That's insane.

No, that's criminally insane.

The University Council will vote on the Fall 2017 start as seen here (auto download PDF, so don't click on it unless ya want to). The 2016-2017 academic year (this coming Fall) will break the previous record by starting on the 11th, which is too friggin early. The 10th, even worse. I've written of this before, but I've recreated the table from the post to include the new (proposed) data. See below.

Year Start
2017-2018 10th (proposed)
2016-2017 11th
2015-2016 17th
2014-2015 18th
2013-2014 12th
2012-2013 13th
2011-2012 15th
2010-2011 16th
2009-2010 17th
2008-2009 18th
2007-2008 16th
2006-2007 16th
2005-2006 18th

At this rate, if the trend continues, we'll soon begin Fall Semester in July.

Oh ... this just hit me. This coming Fall we start on a Thursday (stupid) and Fall 2017 we also start on a Thursday (still stupid). That may explain the early times.

It's still stupid to start on a Thursday.