Author Archive for lisan

Anything you can’t measure you can’t manage

Anything you can’t measure you can’t manage

Who said that?

Wilbur Ross, potential Secretary of Commerce in the Trump administration. This quote during his confirmation hearing is relevant as one of the agencies that will be under his purview is the Census Bureau.

He is also sympathetic to funding for the Census Bureau and presumably other statistical agencies:

In recent years, some Republicans in Congress have tried to restrict government data collection and to cut funding to the Census Bureau and other statistical agencies. Ross signaled he will fight those efforts. “It’s been hard getting the commitment for the appropriations that census really needs for its mission”

The Washington Post had an op-ed about the Census Bureau on February 20th and Wilbur Ross is mentioned in it:

IN HIS confirmation hearing last month, Wilbur Ross noted he may be the first secretary of commerce nominee who was once a U.S. census taker. Those skills could come in handy right about now: A recent report indicates the 2020 Census is in trouble.

Sources
What We Learned (And Didn’t) About Wilbur Ross At His Confirmation Hearing
Ben Casselman | FiveThirtyEight
January 18, 2017

Wilbur Ross’s first task should be saving the 2020 Census
Editorial Board | Washington Post
February 20, 2017

New word of the day: frugging

Do you know what frugging means? AAPOR condemned the Mainstream Media Accountability Survey for frugging:

AAPOR Statement on Trump/Pence Campaign Web Survey
American Association for Public Opinion Research (AAPOR)
February 23, 2017

A hint is in the title – campaign. But, otherwise, read the short statement and find out.

As to the ‘survey‘ it could also be criticized for all the double-barreled questions it has. Here’s Question 13:

survey question

The Forgotten Men Index

graph

The Economist has created an index based on the unemployment rate, labor force participation rate, and average hourly wages. The index compares the fortunes of white working class men (WWCM) to all men. It will be updated monthly. So far, the index stands at 100; it was at 62 in 1994.

Details – but not enough, are in the articles below:

Daily Chart: Tracking the fortunes of America’s white working-class men
The Data Team | The Economist
February 20, 2017

The forgotten men index: Tracking the fortunes of the white working-class
The Economist
February 18, 2017

It might be interesting to look at this at lower levels of geography (states, counties, etc.) based on the American Community Survey instead of the original sources, which aren’t necessarily suitable for sub-national geographies.

Rescuing Federal Data

ICPSR has a new portal that allows the public to upload valuable government resources for preservation and dissemination – DATA LUMOS. These government files are snapshots of the data as it existed on the day it was harvested. Obviously, a live version is preferable, but if data disappear from government websites, the last known version is preferable to nothing. And, via crowd sourcing, this harvesting effort can be shared among many.

DATA LUMOS focuses on preserving federal social science data, interpreted broadly. Pollution data from the EPA would be relevant as would be daily temperature data from NASA. The main decision point is whether users of these data would think to search for it at ICPSR.

DATA LUMOS Announcement
ICPSR Webinar on DATA LUMOS

This is not the only “save the data” organization. A few weeks earlier there was an Ann Arbor Data Rescue event, part of a national Data Refuge project and the Internet Archive’s End of Term Presidential Harvest:

Library participates in effort to preserve government data
Lynne Raughley | University Record (University of Michigan)
February 1, 2017

Here are two other similar events:

Saving Data: Preservation during Political Turmoil
Andrew Battista | Data Dispatch (NYU Data Services)
January 26, 2017

Rogue Scientists Race to Save Climate Data from Trump
Zoe Schlanger | Wired
January 19, 2017

More background on some of the larger collaborations driving this:

DataRefuge Project

DataRefuge is also an initiative committed to identifying, assessing, prioritizing, securing, and distributing reliable copies of federal climate and environmental data so that it remains available to researchers. Data collected as part of the #DataRefuge initiative will be stored in multiple, trusted locations to help ensure continued accessibility

End of Term Presidential Harvest 2016

This is a collaborative effort between the U.S. Government Publishing Office (GPO), the Library of Congress, the California Digital Library, the University of North Texas Libraries, Internet Archive, George Washington University Libraries, and Stanford University Libraries to harvest and preserve public U.S. Government websites at the conclusion of the current Presidential administration ending on January 20, 2017.

Note that this web harvest was done at the conclusion of other administrations, e.g., 2008 and 2012.

More on the “End of Term Presidential Harvest”
Harvesting Government History, One Web Page at a Time
Jim Dwyer | New York Times
December 1, 2016

Large portions of dot-gov have no mandate to be taken care of,” said Mark Phillips, a library dean at the University of North Texas, referring to government websites. “Nobody is really responsible for doing this.

Enter the End of Term Presidential Harvest 2016 — a volunteer, collaborative effort by a small group of university, government and nonprofit libraries to find and save valuable pages now on federal websites. The project began before the 2008 elections, when George W. Bush was serving his second term, and returned in 2012.

And, a few more interesting notes:

The EPA Just Posted a Mirror Website of the One Trump Plans to Censor
Matt Novak | Gizmodo.com
February 16, 2017

tweet

And, the Twitter poster above, is a one-man operation that rescues knowledge from the internet:

The Memory Hole

For instance, here are links to recently deleted items posted on the site:

The Education Department’s Deleted IDEA Website
The disappeared website about public education for disabled children still exists….

NASA’s Internal Counterintelligence Newsletter
Twenty-five issues of NASA’s newsletter about information security, terrorism, and spies.

Trump Deletions
A collection of online material deleted by Donald Trump, his campaign, and his transition team.

Muzzling Federal Scientists

Breaking news: scientists in selected federal agencies have been told to no longer update their websites with reports, factsheets, etc. Here’s a summary of the memo that the Environmental Protection Agency got:

The memo said there would be no press releases, social media posts or blog messages until further notice. It also asked for a list of external speaking engagements for staff and any planned webinars. It warned that listservs would be reviewed and that staff should “only send out critical messages, as messages can be shared broadly and end up in the press.”

Federal Workers Told To Halt External Communication In First Week Under Trump
Sam Stein | Huffington Post
January 24, 2017

Trump bans EPA employees from giving social media updates
Mallory Shellbourne | The Hill
January 24, 2017

Trump Administration Moves to Muzzle Scientists, Block Research
Union of Concerned Scientists | http://www.ucsusa.org/
January 24, 2017

USDA science researchers ordered to stop publishing news releases, other documents
Jose DelReal | Washington Post
January 24, 2017

Information lockdown hits Trump’s federal agencies
Andrew Restuccia, Alex Guillen and Nancy Cook | Politico
January 24, 2017

[Additional News]

Note, that this happened under the Harper government in Canada [summary]. And Canadian scientists warned US scientists of this possibility a month ago:

Canadian Scientists Warn U.S. Colleagues: Act Now to Protect Science under Trump
Dina Fine Maron | Scientific American
December 20, 2016

Federal Policy for Protection of Human Subjects

A final update to the regulations that govern Human Subjects research that have been in place since 1991 was released today. The original notice to update/change the Common Rule was proposed in Fall 2015 with an opportunity for reactions to it. And, plenty of folks made comments about some of the proposed revisions and at least some of these were dropped. The Science article below briefly discusses the controversial consent proposal, while the Bill of Health Blog gives a quick overview of what stays the same and what has changed. The Bill of Health blog also discusses some implementation issues, e.g., could the 115th Congress reject this, etc.

Update: U.S. abandons controversial consent proposal on using human research samples
Jocelyn Kaiser | Science
January 18, 2017

Final Common Rule Revisions Just Published
Holly Fernandez Lynch | Bill of Health Blog
January 18, 2017

Final rule enhances protections for research participants, modernizes oversight system
Press Release | Health and Human Services
January 18, 2017

Federal Policy for the Protection of Human Subjects
Document Number 2017-01058 | Federal Register
January 17, 2017

This rule is effective on January 19, 2018. The compliance date for this rule, except for Sec. __.114(b) (cooperative research), is January 19, 2018. The compliance date for Sec. __.114(b) (cooperative research) is January 20, 2020.

RAPID Research Community Alert
Dr. S. Jack Hu | Institutional Office for Human Subjects Research [University of Michigan]
January 30, 2017

Selected comments to the original proposed Common Rule
Researchers decry consent proposal
Jocelyn Kaiser | Science
May 20, 2016

More apportionment fun

The Census Bureau just released its 2016 Population Estimates. Let’s take a peek at what the Electoral College might look like in 2020 using the PSC Apportionment calculator. The easiest calculation is to just put the 2016 estimates into the calculator (remember to delete Washington, DC). In that scenario:

The losers: Illinois, Michigan, Minnesota and Pennsylvania
The winners: Florida, North Carolina, Oregon, and Texas

If we take the 2010 to 2016 growth rate and extend it to 2020, this is the scenario:

Losers: Alabama, Illinois, Michigan, Minnesota, New York, Ohio, Pennsylvania, Rhode Island, West Virginia
Winners: Arizona, Colorado, Florida (2), North Carolina, Oregon, Texas (3)

Notable in this scenario is Rhode Island losing an electoral seat. It is just slightly larger than Montana – 1,056,426 vs 1,042,520. But, Rhode Island has had an extra seat since 1990 when Montana lost 1 seat in the House of Representatives. Montana is on the cusp of getting that 2nd representative in the House – it just needs ~5,000 more people than our 2020 projection, which is not an unrealistic scenario.

One thing that is unrealistic about the previous scenario is the fast growth of North Dakota during this period. North Dakota has been the fastest growing state for the past 4 years, but its growth rate dropped to 37th fastest in 2016 due to the collapse of oil prices and thus the fracking industry. Thus, a more realistic estimate might be to use the 2015-2016 rate for the last 4 years of the decade. In that scenario, the results are exactly the same, except that now Montana needs only 550 more people to gain a 2nd seat in the House of Representatives.

Resources
Vintage Population Estimates (2016)
http://www.census.gov/programs-surveys/popest/data/tables.html
State Population Estimates (2016)
http://www.census.gov/data/tables/2016/demo/popest/state-total.html
PSC Apportionment Calculator
Representations apportioned to each state (1790 to 2010)
Congressional Apportionment Resource

Mapping Megaregions via ACS commuting data

Two researchers have created a map of megaregions in the US based on commuting data from the American Community Survey (ACS). The results are covered in both the popular press and in PLOS One. The latter provides more details about how they constructed the maps – it wasn’t just via a mapping program.

How 4 Million Commutes Shape America’s ‘Megaregions’
Laura Bliss | Atlantic: City Lab
December 7, 2016

An Economic Geography of the United States: From Commutes to Megaregions
Garrett Nelson and Alisdair Rae
PLOS One
November 30, 2016

If you are so inclined the authors have made their data available for replication via Figshare.

[Additional Media Coverage]

Great Tweetstorm: Most important year in Economics?

This is from the blogger @undercoverhistorian. We had a previous post on the site she maintains. Below is an interesting set of almost 50 tweets – some illustrated – where she defends 1952 as the most important year.

twitter feed
Click here for tweetstorm

Risk Visualization Theater

How to better communicate election forecasts — in one simple chart
Justin Gross | Monkey Cage blog [Washington Post]
November 29, 2016

Most folks were surprised by the results of the 2016 Presidential election and this was in part due to some of the rosy forecasts by some of the poll aggregators, like Huffington Post. But, even when a site had a forecast with a 30% chance of Trump winning, most people have trouble understanding that a Trump victory was possible. The explanation:

But certain representations of probability are more readily grasped than others. In particular, we have trouble understanding risk in terms of the “percent chance” but we do better when simple raw numbers of different outcomes are depicted visually.

Solution: Show the risk as a “Risk Visualization Theater.” Below are the representations of forecasts of victory for Trump via FiveThirtyEight, NYT Upshot, and Huffington Post Pollster. The filled theater seats (in black) represent the chance of a Trump victory. Clearly, the chance of that event happening don’t look so remote in the far left depiction, but look very unlikely as one moves to the right.

Risk via a theater