The Centers for Disease Control have produced a great data visualization on the rising rate of drug-overdose fatalities in the United States over the past decade. Click on the Storypoints-- a great feature in Tableau-- to explore this endemic problem from different perspectives.
I had the opportunity recently to attend a solidarity rally in support of the Charleston 9 at First A.M.E. Church in Seattle. The room, though packed and sweltering, was filled with an overwhelming sense of a community unifying for change. It was powerful. Among the speakers was Dr. Sheley Secrest with the Seattle- King County NAACP. Dr. Secrest’s work focuses on economic development, and she spoke of a meeting at Boeing in the wake of significant recent layoffs. When she asked about data on who was being laid off – by race, hire date, and so on – she was told “Boeing doesn’t track that information.” For me, Dr. Secrest’s observation resonated as another example of data (or in this case, its absence) being used as a weapon. In The Current we recently reflected on the weaponization of data as part of a conversation with our friends and colleagues, Vu Le and Dr. Jondou Chen. Data can be weaponized, as well as the absence of data – and in some ways, this latter form is more insidious. […]
For the past few years our friend Vu Le of Rainier Valley Corps has been publishing a terrific blog called Nonprofit with Balls. If you don’t already know Vu, the title gives you a clue about his provocative ideas. A seasoned nonprofit leader, Vu has an unorthodox take on how the nonprofit world actually works—and lots of disruptive (in a good way) ideas about how it could work better. Vu recently posted a blog entry that got our attention here at TrueBearing: Weaponized data: How the obsession with data has been hurting marginalized communities. It’s a thought-provoking read for anyone involved in the nonprofit, public or grantmaking sectors, so all you unicorns out there go ahead and click on the link to read his post. I guarantee you’ll chuckle at least twice- and you’ll get the reference to unicorns. I’ll wait. Back already? OK. For those of you who didn’t bother to click the link, here is a 30,000-foot overview of Vu’s post: “Data can be used for good or for evil.” While acknowledging the power of skillfully used data and its benefits to both nonprofits and grantmakers, Vu nails ten distinct ways in which data can be—and too often has been—used to obscure rather than to illuminate, to diminish the richness of our understanding of nonprofit performance, and to maintain the power status quo in a way that marginalizes and sometimes even pathologizes entire communities. […]
Does your organization’s work involve health outcomes? Check out this awesome resource from Community Commons. Drawing from a number of reliable sources of health data (e.g., CDC, US Census, Department of Health and Human Services), this tool generates Community Health Needs Assessment reports that cover categories like demographics, social and economic factors, even health behaviors and outcomes (including what proportion of the population has specific health issues, and how each figure compares statewide and nationwide). The best part is that you can interact with the site to filter down to individual counties or even zip codes. […]
We’re always on the lookout for great sources of useful data, and particularly sources that visualize information in visually arresting ways. So we are really pleased to be able to share one such resource: the Opportunity Index. The Index is a website sponsored by the national coalition Opportunity Nation. Offering snapshots of the educational, economic and civic opportunities available across the country, the Index is capable of drilling down to the state and county levels to provide a remarkably nuanced picture. Fair warning: Do not click on the link to the Index unless you have some free time, because I promise that you will find it difficult to tear yourself away from this fascinating site! […]
Tying funding for social programs to their effectiveness seems like a no-brainer. Sadly, however, genuine evidence-based decision making in policy and budget priority-setting in federal social spending is all too rare. Evaluations are often required in federally-funded social programs; however, the standards of evidence have often been unclear or lacking altogether (Sorry, but using client satisfaction scales as your sole measure of success is a poor way to measure effectiveness!). On top of that, since performance is rarely considered in funding decisions, little incentive exists for programs to change and improve in response to evaluative feedback. An effort to rectify this situation began in the Bush II years— but even then, less than .2 percent of all nonmilitary discretionary programs were held to rigorous evaluation standards. Kind of takes your breath away, doesn’t it? That’s a particularly disturbing figure when you consider that, according to Ron Haskins in the NYT, “75 percent of programs or practices that are intended to help people do better at school or at work have little or no effect.” One way of interpreting this shocking figure is that in the absence of evidence-based decision making, a massive amount of funds are tied up in supporting ineffective programs that could be invested in promising alternatives. […]
We came across this article the other day and have to admit, it gave us a chuckle. Of course, we’re fans of data in all its many forms, but being based in downtown Seattle, we don’t get many chances to think about it in the context of cows – er, make that calving management. But the article provides fodder for a crunchy little morsel of usefulness. The author explains that there is an existing method for collecting data on dystocia (cows with difficult labor in delivering calves). It’s been around since 1978, and involves a five-point scale. The scale ranges from 1 – no problems in delivery, 2 – slight problem, 3 – assistance needed, 4 – considerable force required (we’re picturing an episode of James Herriot’s All Creatures Great and Small), to 5 – extreme difficulty or surgical intervention. The controversy involves a new proposed scale from 1 (no problems), 2 (one-person pull) to 3 (severe traction or surgery). On an intuitive level, this sounds nice and simple, right? Who needs those extra points on the scale anyway? Well, dairies do. Reducing the spectrum from five points down to three means that the resulting data will be less sensitive to what’s really going on. And that makes it less reliable for decision making. Moral of the story? Pay attention not only to the data being collected, but the metrics used to collect it. Simpler is not always better. […]
Good data is necessary – but not sufficient in itself – for evidence-based decision making. After all, you can’t make an evidence-based decision without evidence, right? That’s why it’s dismaying that the folks at the Census are considering removing a crucial set of items when the next Census comes around in 2020. […]
I’m still sifting through the aftermath of the Presidential election and the controversial accusations of bias in polling and predictions leveled against various pollsters on all sides of the political spectrum. For data geeks, the question of bias in polling is a source of endless fascination. As someone whose profession involves mostly non-political surveys, however, I zeroed in on a basic methodological question: Regardless of polling firm, what polling method showed the least bias in predicting the election? According to a detailed post by Nate Silver, the answer is clear, and should make anyone who relies on survey or polling data sit up and take notice: All things being equal, online surveys showed 40 percent less bias than live telephone interviewers, and an astonishing 72 percent less bias than automated telephone “robopolls.” […]
This graphic uses humor to communicate a widely held public view of research; that the reality isn’t all it appears to be (a good illustration of Erving Goffman’s ‘front stage – back stage’ role performance). […]
