Facebook Data Goes to Researchers

Facebook dataResearchers expected it a year and a half ago, but Facebook is finally giving researchers access to a lot of data. The data is about how users have shared information, including misinformation, about political events around the world.

The data released last month relates to URLs (38 million) that users shared publicly on Facebook between January 2017 and July 2019. Did they consider a linked site to be fake news or hate speech? Which links did they click or like or share?

Social scientists will also be able to connect that with some demographic information like age, gender, and location and political affinities. There are also concerns that there are distortions, or noise, that have been injected into the data. Why? Thankfully, because of differential privacy by data managers who have tried to ensure privacy.

This seems to echo the last U.S. Presidential election in 2016 when Facebook was hit with evidence that it had given political operatives unauthorized use of its data. In April 2018, they announced that they would turn over full access to information about its users with no strings attached - but to researchers.

It's the right thing to do but a tough thing for a company to do - turning over proprietary information. Previously, that data was only available for research that was either conducted in-house or required preapproval from Facebook.

Statistical Seduction

It is easy to be seduced by statistics. I know several friends who have websites and blogs and are rather obsessed with their web statistics. They are always checking to see how many hits the site gets or what pages or posts are most popular or what search terms are being used to find them.

Social media has encouraged this with Likes and Retweets and Reposts. Our smartphones love to send us notifications that someone has engaged with some piece of our content.

I got this alert last month about another blog of mine:

Your page is trending up
Your page clicks increased by more than 1,000% over the usual daily average of less than 1 click.
Possible explanations for this trend could be:

Modifications you did to your page's content.

Increased interest in a trending topic covered by the page.

Yes, their "possible explanations for this trend" are correct. It is a post about the winter solstice, so you might expect that interest would increase around mid-December. When a topic, such as "winter solstice," is trending generally, you will get more attention to your page. I also made some updates to the post and Google and other spiders scurrying around the web notice that. 

Top 30 of 130 countries that visited the site in the past year.


I have become less interested in the stats as the years have passed. In the early days of my blogging, I was much more interested in those hits, impressions, and visitors. Nowadays, I only check at year's end.

At the end of 2019, Serendipity35 added 3,595,439 hits to bring its total over the years to a rather daunting 104,596,905. In December 2019, we averaged 11,478 hits daily. It's more sobering to note that the average number of daily visits was 2722.

What I find more interesting in the analytics are things like which countries are visiting the site (see image) and the terms people searched that led them to the blog.


Tracking Your Health Data

fitness trackerThe Verge reported that in another move to gain more of our personal data, Google is teaming up with the nation’s second-largest health system, Ascension, in an effort it’s calling Project Nightingale.

Ascension will share the personal health data of tens of millions of patients with Google’s Cloud division in order to develop a search engine for medical records and other new artificial intelligence services for medical providers. That sounds helpful. But the announcement came right after Google announced it was buying the fitness tracker company Fitbit.

We could assume that Google is interested in selling this kind of hardware, but access to Fitbit user's personal data could be an even bigger and more profitable asset. (Fitbit data has already been used in some serious but not health manners - such as police investigations. )

Google is certainly not alone in wanting to gain this type of personal wellness data and do health care-tech collaborations. Apple would like to see its watch (similar to but more powerful than most fitness trackers) function as a medical-monitoring device. Its health data-sharing capabilities through Apple HealthKit are being enhanced.

All these companies will point out that the data they obtain is anonymized, but there are many examples of anonymized data being reversed so that it is far less anonymous. Are laws and policies ready for all this?

Farewell to Curbs and Other Unexpected Uses of Technology

curbDevin Liddell is Chief Futurist at Teague, a firm that specializes in design for transportation. He thinks about how technology and design will change mobility. An article on Geekwire.com that I saw via Amber MacArthur's newsletter discussed a few of those changes.

The one that surprised me the most was about curbs - that quite old and established way to separate the street from the sidewalk. In 19th century cities, they helped keep walkers from stepping in manure from horse-drawn carriages. But in the 21st century, Liddell says, “The curb as a fixed, rigid piece of infrastructure isn’t going to work.” He thinks there is a role for design in creating a more dynamic understanding of curbs. Nuanced with signage can change curb spaces from no parking to emergency-only to pay-by-the-hour parking.

A suburban curbside may not be an issue, but in cities and at airports, they are problem areas.

Liddell references Coord which is the urban planning spinout of Google/Alphabet’s Sidewalk Labs. Can you believe they have Open Curb Data that maps the use of city curbs. Self-described "Coord makes it easy to analyze, share, and collect curb data. Curbside management now includes better compliance, safety, and efficiency for communities of all sizes."

Curb data? Really?