Mistaken Data

3 posts
When data is not quite what it seems

FiveThirtyEight used a dataset on broadband as the basis for a couple of stories. The data appears to be flawed, which makes for a flawed analysis. From their post mortem: We should have been more careful in how we used the data to help guide where to report out our stories on inadequate internet, and we were reminded of an important lesson: that just because a data set comes from...

0 0
When the interesting pattern ends up just being computer byproduct

Good lesson here. Christian Laesser was playing around with receipt data and initially thought he had a fun pattern at hand. It looked like the shopper always put things in his or her cart in the same order every time. It turns out though that the order just came from the computer ordering items by category. It had nothing to do with shopping order. Familiarize yourself with your data source...

0 0
Misleading Medicaid funding with the baseline

The administration tweeted a chart that shows the Senate Republican health care bill increases Medicaid funding. The line moves up, so it must be true, right? Well, it depends on what you compare to. The original simply compares over time — against the past. Vox compared it against what spending would be under current law. Tags: baseline, Medicaid

0 0
Common statistical interpretation mistakes

Statistics is a game of subtleties, and you lose when you don’t pay attention to the details. Here are a handful of common mistakes when interpreting the numbers. In a nutshell: You get into trouble when you assume and ignore. Tags: causation, correlation, noise, pitfalls

0 0
Data distributed as clipart

Government data isn’t always the easiest to use with computers. Maybe it’s in PDF format. Maybe you have to go through a roundabout interface. Maybe you have to manually request files through an email address that may or may not work. However, this file that OpenElections received might take the cake. It’s a spreadsheet, but the numbers are clipart. City of Detroit produced a lookup tables for its absentee precincts...

0 0
Skittle disconnect

This is what happens when there is a disconnect between data and what it represents. So much wrong. This image says it all. Let's end the politically correct agenda that doesn't put America first. #trump2016 pic.twitter.com/9fHwog7ssN — Donald Trump Jr. (@DonaldJTrumpJr) September 19, 2016 I need to avoid social media for the next month. There is something upsetting every single time these days. Tags: human, Skittles

0 0
Sans human, Facebook’s Trending Topics algorithm faired poorly

Last week, Facebook announced that it was making the Trending Topics section more automated. More algorithm-based. Less person-based. On Monday, the section showed a fake news story at the top of the list for several hours. Nick Statt for the Verge on the human element: The changes instituted on Friday didn’t throw all of that away; Facebook has been slowly stripping away the human element of Trending Topics for months...

0 0
Guide to spotting data BS

As we delve deeper into election season, politicians will spit out more and more statistics to lend some factitude to their talking points. Some are real, and others will be less real. David Spiegelhalter for the Guardian provides a nine-point guide on how to sift out the latter. On estimates and margin of error: Next time you hear a politician boasting that unemployment has dropped by 30,000 over the previous...

0 0
Science formally retracts LaCour paper

Last week, graduate student Michael J. LaCour was in the news for allegedly making up data. The results were published in Science. LaCour's co-author Donald Green requested a retraction, but the paper stayed while the request was considered. Today, Science formally fulfilled the request. The reasons for retracting the paper are as follows: (i) Survey incentives were misrepresented. To encourage participation in the survey, respondents were claimed to have been...

0 0
Graduate student makes up data for fake findings

Last month, This American Life ran a story about research that asked if you could change people's mind about issues like same-sex marriage and abortion — with just a 22-minute conversation. The research was published in Science, but Donald Green asked the publication to retract the paper recently. It seems his co-author and UCLA graduate student, Michael LaCour, made up a lot of data. Green today told me if there...

0 0