Skip to ContentSkip to Radioplayer

Our members keep us going.


Data Blog

Know the score: How schools rate in the Louisiana Scholarship Program

We compiled data on all Louisiana schools that participated in the state’s scholarship program since the 2015-2016 school year.

Many schools don’t have letter grades because Louisiana only reports test scores for schools with at least 10 scholarship students per grade level. But most students in the program are concentrated in schools that did report. We calculated letter grades based on each school’s SCI score, which the Louisiana Department of Education says is comparable to the scores used to calculate public school letter grades.

#school-database iframe { width: 100%; height: 550px;} var id = ‘school-database’; var pymParent = new pym.Parent(id, ‘’); function sendSearchUrl() { var message = JSON.stringify({‘url’: window.location.href, ‘id’: id}); pymParent.sendMessage(‘receiveUrl’, message); }; pymParent.onMessage(‘requestUrl’, sendSearchUrl);

Source: Louisiana Department of Education Graphic: Ted Han .article-layout__fullwidth {display:none;} .article-layout__fullwidth .pull-outside-layout { grid-column:2/5;} p {max-width: 700px;} @media screen and (max-width:1023px) {.article-layout__fullwidth .pull-outside-layout { grid-column:1/4;}}

These police officers were members of extremist groups on Facebook

We wanted to know whether police officers nationwide were members of extremist groups.

That’s not a simple task, but we found one way to do it. We learned that on Facebook – at the time – you could download the membership rolls of private groups. So, we downloaded the names of members of two different kinds of groups we identified on Facebook: extremist groups and police groups.

We loaded those roughly 1 million names into a database and asked a simple question: How many people were members of at least one extremist group and at least one police group?

It would have taken years to deeply examine all 14,000 hits that surfaced, so we selected a fraction of the list to vet. We worked to verify whether users were actually police officers by reviewing their biographies or photos or matching their names in police records. Sometimes that was easy. Other times, identities were cloaked behind aliases. Sometimes people turned out to not be police officers at all – they’d just joined the police groups as supporters.

We started with a list of more than 1,200 extremist groups compiled by Megan Squire, a computer science professor at Elon University and leading expert on extremism on Facebook. Squire, who has been scouring Facebook for extremists for a couple of years, methodically works through the platform, searching for groups with names that denigrate people by race, religion, gender or sexuality or that contain code words used by extremist movements. And because Facebook recommends new groups based on... Read More >

Concussion laws: How does your state stack up?

Nearly a decade ago, lawmakers in Oregon and Washington state enacted the first laws on how to handle student-athlete concussions, in reaction to severe brain injuries on the football field. Since then, similar laws have swept the country.

By 2014, all 50 states and Washington, D.C., had passed youth concussion laws, focused on letting young athletes heal after a concussion. Most require that athletes be pulled from games or practice until they’re cleared by a health care professional. But important details differ state to state.

As our investigation with partners Pamplin Media Group and InvestigateWest found, those new laws created a paper trail of concussions. But the concussion incident reports mostly stay in school files. Few states compile the trove of information to help identify concussion risks and potentially fix them.

In some places, athletic associations have done more with the data. The Michigan High School Athletic Association requires member schools to report details about every concussion as it happens and publishes comprehensive data annually. A similar sports governing body in Texas has partnered with the Peter O’Donnell Jr. Brain Institute to create a database of student concussions. Large high schools in Texas are required to participate starting this fall. Public access will be limited, as the institute plans to share data only through periodic reports. Hawaii has tracked student concussion numbers statewide for more than a decade, but with limited public reporting.

Of note: No laws specifically address the long-term risk of repeated hits to the head, which currently is a major concern... Read More >

Working together 101: How academics and journalists can collaborate

[Cross-posted with Source]

Academics bring deep knowledge and expertise in their area of study, and journalists have mastered the art of telling stories. When academics and journalists collaborate, they can combine their skills to tackle complex questions and communicate the answers in ways that can have a huge impact on society. That’s why we’re working to foster these collaborations. At this year’s SRCCON, we facilitated a session about how to make this all happen.

We’re also excited to announce that we’re building a handbook to make it easier to work across disciplines, no matter your role. You can contribute examples and information about your experiences here!

A few basics about academia and journalism

Why should academics and journalists collaborate? What are some good examples of collaborations?

Sinduja Rangarajan: The Chicago Tribune’s collaboration with scientists at Columbia University to study fatal drug interactions is an example of a collaboration that made a big impact. Sam Roe, a veteran investigative journalist, brought together a team of scientists, academics and experts from the Columbia University Medical Center and University of Arizona medical school to answer questions about potentially fatal drug combinations that never had been studied before. His ideas, questions and knowledge of scientists and their work came from deep reporting that he’d done on the topic over many years. As a journalist, Roe was key to making this original research happen by bringing together academics from different institutions and managing this massive project with many partners over multiple years.

In another large-scale collaboration called the Open Policing Project,... Read More >

Redding confronts a deadly pattern: A history of wildfires and development in high-fire-risk areas

Wildfires are once again claiming homes and lives in a Northern California city that has pushed into wildfire-prone areas.

The Carr fire near Redding fits a deadly pattern, sharing traits of the firestorm that destroyed thousands of homes and killed dozens last year in Sonoma and Napa counties.

Since 1947, more than two dozen fires have burned in the same area as the Carr fire. Most of the area surrounding Redding has been rated as a very high fire hazard, according to the California Department of Forestry and Fire Protection.

Redding, like the Santa Rosa area, has also grown over the past few decades. The city has pushed into the Northern California wilderness, placing homes among oak trees and flammable chaparral outside of town. Many of those homes now wait in the fire’s path.

Credit: Eric Sagara/Reveal
Sources: Cal Fire, USGS, Google, U.S. Census Bureau

Already, dozens of buildings have been destroyed in the fire, which started on July 23 near Whiskeytown Lake, northwest of Redding. On Thursday night and Friday morning, the fire pushed into the city of Redding, which has a population of nearly 92,000.

The death toll so far includes at least two emergency responders, according to the Sacramento Bee. A private bulldozer operator died Thursday while working in an area where the fire was active. A Redding firefighter was also killed, authorities announced Friday morning.

Three more firefighters were hospitalized for burns to their ears, face and hands they received while protecting buildings from the 44,000-acre fire.

Fire officials estimated Friday morning that 65 buildings... Read More >

In 6 aerial images: How California wine country was primed for disaster

All eyes are on Northern California this week as devastating wildfires tear through wine country. At least 31 people have died and thousands more have been left homeless, mostly in Napa and Sonoma counties.

This infrared satellite image from Planet shows fire damage Tuesday near Napa’s Silverado Resort. Blackened areas represent scorched ground. Healthy vegetation, such as the resort’s golf course, is in red. We’ve added imagery from Google for comparison.

Even with thick smoke covering the area Wednesday, the hardest-hit fire zones were clearly visible.

But the conditions for the unfolding catastrophe were set long ago, both in wine country and throughout California.

Large portions of both Napa and Sonoma counties have been rated by the state Department of Forestry and Fire Protection as having a high chance of destructive fire based on the vegetation types and condition of the area. Nearly all the fires there are burning through areas with the most severe classification.

Most Northern American forests are supposed to burn at regular intervals – every few years or so. Northern California is no different, but many of the current fires are consuming trees left untouched by flame for decades.

The Tubbs and Nuns fires both began in areas that last burned in 1964. The Atlas Fire sparked in a region that hasn’t seen flames since 1981. CAL FIRE has no recorded fire history for the Partrick Fire to the south.

The flames did not stop at treelines. They moved into communities, coursing through streets, claiming house after house, building after building. These communities... Read More >

How we analyzed domestic terror incidents

To build our Homegrown Terror database, we obtained data from a variety of sources: the Congressional Research Service, the FBI, DT Analytics, The Heritage Foundation, the Investigative Project on Terrorism, New America, Mother Jones, the Muslim Public Affairs Council, the Southern Poverty Law Center, the Anti-Defamation League, the National Abortion Federation and the Animal Liberation Front’s own website. In addition, we set up news filters and searched journalism databases to scoop up missing incidents, using search terms such as “Islamist,” “sovereign citizen,” “Oath Keeper,” “ecoterrorism” and so on.

Through public databases and Freedom of Information Act requests, we collected primary court and law enforcement documents for almost every incident and carefully examined them, in combination with credible news coverage, to check whether each entry met the FBI criteria that define domestic terrorism. Incidents that met the definition were included, regardless of whether prosecutors filed terrorism charges. Some incidents inevitably involved judgment calls. To adjudicate those, we turned to a panel of experts: William Banks, director of the Institute for National Security and Counterterrorism at the Syracuse University College of Law; Bruce Hoffman, director of the Center for Security Studies at Georgetown University; and Daryl Johnson, former senior domestic terrorism analyst at the Department of Homeland Security’s Office of Intelligence and Analysis, now the owner and CEO of consulting firm DT Analytics. They provided guidance on whether to include or exclude borderline cases, such as those related to hate crimes, mental illness and confrontations with law enforcement.

In cases of mental illness, we... Read More >

Sounding out the border wall

You might know what the U.S.-Mexico border looks like – but do you know what it sounds like?

Using a map of the existing barriers between the U.S. and Mexico that our data team created, we worked together to sonically represent the presence – and absence – of border barriers in a few different ways.

First, to convert spatial data into sound, we determined when notes should be played.

We calculated the distance between the beginning of the border and where each segment of the fenced border began, as well as the length of that segment. We chose a speed at which we would move along the border (10 miles a second), then converted those distances to time. For example, a 10-mile-long segment of fence that begins 20 miles from the start of the border would start playing at 2 seconds, and the note would last for one second.

We also came up with different sounds to represent the various types of barriers along the border.

So go ahead, click play.

While you’re listening, picture yourself flying low over the U.S.-Mexico border, starting in San Diego and Tijuana, Mexico, and ending nearly 2,000 miles to the east near Brownsville, Texas, and Matamoros, Mexico.

The lower note/melody represents tall pedestrian fence, often 10 to 20 feet high. Higher-pitched piano plinks represent shorter fence designed to stop vehicles, but not people. An airy keyboard drone signals gaps in the fence.

You can hear a stripped-down version of this song in our episode “Up against the wall.” It relies on the... Read More >

1 2 3 5